commit ac08a0b6ffa1aed25680ee2fe884b33c717d9294
Author: zhangzf1119 <zzfloveown@163.com>
Date:   Sat Dec 13 23:00:09 2025 +0800

    I第一次提交

diff --git a/.DS_Store b/.DS_Store
new file mode 100644
index 0000000..e821bba
Binary files /dev/null and b/.DS_Store differ
diff --git a/.dockerignore b/.dockerignore
new file mode 100644
index 0000000..4e8c5cc
--- /dev/null
+++ b/.dockerignore
@@ -0,0 +1,41 @@
+# Maven
+# target/  # 注释掉，让Docker能访问JAR文件
+pom.xml
+
+# Git
+.git
+.gitignore
+
+# IDE
+.idea/
+.vscode/
+*.iml
+
+# OS
+.DS_Store
+Thumbs.db
+
+# Logs
+logs/
+*.log
+
+# Temporary files
+*.tmp
+*.swp
+
+# Test files
+test.txt
+
+# Upload and extract directories (will be created at runtime)
+uploads/
+extracts/
+
+# Documentation
+HELP.md
+SMART_QA_README.md
+README.md
+
+# Docker
+Dockerfile
+docker-compose.yml
+.dockerignore
diff --git a/.gitattributes b/.gitattributes
new file mode 100644
index 0000000..3b41682
--- /dev/null
+++ b/.gitattributes
@@ -0,0 +1,2 @@
+/mvnw text eol=lf
+*.cmd text eol=crlf
diff --git a/.gitignore b/.gitignore
new file mode 100644
index 0000000..667aaef
--- /dev/null
+++ b/.gitignore
@@ -0,0 +1,33 @@
+HELP.md
+target/
+.mvn/wrapper/maven-wrapper.jar
+!**/src/main/**/target/
+!**/src/test/**/target/
+
+### STS ###
+.apt_generated
+.classpath
+.factorypath
+.project
+.settings
+.springBeans
+.sts4-cache
+
+### IntelliJ IDEA ###
+.idea
+*.iws
+*.iml
+*.ipr
+
+### NetBeans ###
+/nbproject/private/
+/nbbuild/
+/dist/
+/nbdist/
+/.nb-gradle/
+build/
+!**/src/main/**/build/
+!**/src/test/**/build/
+
+### VS Code ###
+.vscode/
diff --git a/.mvn/wrapper/maven-wrapper.properties b/.mvn/wrapper/maven-wrapper.properties
new file mode 100644
index 0000000..c0bcafe
--- /dev/null
+++ b/.mvn/wrapper/maven-wrapper.properties
@@ -0,0 +1,3 @@
+wrapperVersion=3.3.4
+distributionType=only-script
+distributionUrl=https://repo.maven.apache.org/maven2/org/apache/maven/apache-maven/3.9.11/apache-maven-3.9.11-bin.zip
diff --git a/3.3 同类项目经验.pdf b/3.3 同类项目经验.pdf
new file mode 100644
index 0000000..b63f88c
Binary files /dev/null and b/3.3 同类项目经验.pdf differ
diff --git a/3.3 同类项目经验_catalog_test.json b/3.3 同类项目经验_catalog_test.json
new file mode 100644
index 0000000..6cdd864
--- /dev/null
+++ b/3.3 同类项目经验_catalog_test.json	
@@ -0,0 +1,26 @@
+{
+  "source": "filename_fallback",
+  "total_pages": 39,
+  "flat_chapters": [
+    {
+      "chapterId": "chap-0",
+      "title": "3.3 同类项目经验",
+      "level": 1,
+      "page": "1",
+      "startPage": 1,
+      "endPage": 39,
+      "children": []
+    }
+  ],
+  "hierarchical_chapters": [
+    {
+      "chapterId": "chap-0",
+      "title": "3.3 同类项目经验",
+      "level": 1,
+      "page": "1",
+      "startPage": 1,
+      "endPage": 39,
+      "children": []
+    }
+  ]
+}
\ No newline at end of file
diff --git a/Dockerfile b/Dockerfile
new file mode 100644
index 0000000..f78f0b4
--- /dev/null
+++ b/Dockerfile
@@ -0,0 +1,46 @@
+# 使用Eclipse Temurin OpenJDK 17作为基础镜像（离线部署）
+FROM eclipse-temurin:17-jre-alpine
+
+# 设置工作目录
+WORKDIR /app
+
+# 设置时区为东八区
+ENV TZ=Asia/Shanghai
+RUN ln -snf /usr/share/zoneinfo/$TZ /etc/localtime && echo $TZ > /etc/timezone
+
+# 安装curl用于健康检查和7-zip用于文件解压（可选）
+RUN apk add --no-cache curl p7zip
+
+# 设置构建参数
+ARG VERSION=1.0.0
+ARG BUILD_TIME=unknown
+ENV APP_VERSION=${VERSION}
+ENV BUILD_TIME=${BUILD_TIME}
+
+# 复制Maven构建的JAR文件
+COPY target/*.jar app.jar
+
+# 创建非root用户 (Alpine Linux)
+RUN addgroup -g 1001 -S appuser && adduser -S -D -H -u 1001 -s /sbin/nologin -G appuser -g appuser appuser
+
+# 创建必要的目录
+RUN mkdir -p /app/uploads /app/extracts /app/logs /app/config
+
+# 设置目录权限
+RUN chown -R appuser:appuser /app
+
+# 切换到非root用户
+USER appuser
+
+# 暴露端口
+EXPOSE 8080
+
+# 设置JVM参数
+ENV JAVA_OPTS="-Xmx1g -Xms512m -XX:+UseG1GC -XX:+UseContainerSupport -XX:MaxRAMPercentage=75.0"
+
+# 健康检查
+HEALTHCHECK --interval=30s --timeout=3s --start-period=60s --retries=3 \
+  CMD curl -f http://localhost:8080/health || exit 1
+
+# 启动应用
+ENTRYPOINT ["sh", "-c", "java $JAVA_OPTS -jar app.jar"]
diff --git a/build-docker-bundle.sh b/build-docker-bundle.sh
new file mode 100644
index 0000000..a95328c
--- /dev/null
+++ b/build-docker-bundle.sh
@@ -0,0 +1,279 @@
+#!/usr/bin/env bash
+#
+# 构建离线 Docker 交付包：包含应用镜像、pgvector(PostgreSQL)、Redis 及统一的 docker-compose
+# 使用方式：
+#   chmod +x build-docker-bundle.sh
+#   ./build-docker-bundle.sh                # 使用默认版本号（时间戳）
+#   VERSION=v1.2.3 ./build-docker-bundle.sh # 指定版本号
+
+set -euo pipefail
+
+ROOT_DIR="$(cd "$(dirname "$0")" && pwd)"
+TIMESTAMP="$(date +%Y%m%d_%H%M%S)"
+APP_NAME="gdyd-zhpb-zgf"
+VERSION="${VERSION:-${TIMESTAMP}}"
+OUTPUT_DIR="${ROOT_DIR}/docker-bundle-${VERSION}"
+IMAGES_DIR="${OUTPUT_DIR}/images"
+CONFIG_DIR="${OUTPUT_DIR}/config"
+INIT_DIR="${OUTPUT_DIR}/pgvector-init"
+
+APP_IMAGE="${APP_NAME}-app:${VERSION}"
+PGVECTOR_IMAGE="${PGVECTOR_IMAGE:-pgvector/pgvector:pg16-trixie}"
+REDIS_IMAGE="${REDIS_IMAGE:-redis:7-alpine}"
+
+# 默认账号密码，可在执行脚本前通过环境变量覆盖
+POSTGRES_DB="${POSTGRES_DB:-vector_db}"
+POSTGRES_USER="${POSTGRES_USER:-vectoruser}"
+POSTGRES_PASSWORD="${POSTGRES_PASSWORD:-vectorpass}"
+POSTGRES_PORT="${POSTGRES_PORT:-5433}"
+
+REDIS_PASSWORD="${REDIS_PASSWORD:-TNh8jiy8WpEweGLZ}"
+REDIS_PORT="${REDIS_PORT:-26739}"
+
+APP_PORT="${APP_PORT:-8000}"
+
+print_step() {
+  printf '\n\033[1;34m==> %s\033[0m\n' "$1"
+}
+
+print_info() {
+  printf '   %s\n' "$1"
+}
+
+abort() {
+  printf '\033[1;31m[错误]\033[0m %s\n' "$1" >&2
+  exit 1
+}
+
+check_command() {
+  command -v "$1" >/dev/null 2>&1 || abort "未找到命令：$1，请先安装。"
+}
+
+APT_UPDATED=0
+SUDO_CMD=""
+if command -v sudo >/dev/null 2>&1 && [[ ${EUID:-0} -ne 0 ]]; then
+  SUDO_CMD="sudo"
+fi
+
+update_apt_once() {
+  if [[ ${APT_UPDATED} -eq 0 ]]; then
+    print_step "更新 apt 软件源"
+    ${SUDO_CMD} env DEBIAN_FRONTEND=noninteractive apt-get update -y
+    APT_UPDATED=1
+  fi
+}
+
+install_with_apt() {
+  local package="$1"
+  local friendly="$2"
+  if ! command -v apt-get >/dev/null 2>&1; then
+    abort "缺少 ${friendly}，但当前环境不支持自动安装（未检测到 apt-get）。"
+  fi
+  update_apt_once
+  print_step "安装 ${friendly}"
+  ${SUDO_CMD} env DEBIAN_FRONTEND=noninteractive apt-get install -y "${package}"
+}
+
+ensure_maven() {
+  if ! command -v mvn >/dev/null 2>&1; then
+    install_with_apt "maven" "Maven"
+  fi
+}
+
+ensure_java21() {
+  local major=""
+  if command -v java >/dev/null 2>&1; then
+    local version_line
+    version_line="$(java -version 2>&1 | head -n 1 | cut -d'"' -f2)"
+    major="${version_line%%.*}"
+    if [[ "${major}" == "1" ]]; then
+      major="$(echo "${version_line}" | cut -d'.' -f2)"
+    fi
+    if [[ -n "${major}" && "${major}" -ge 21 ]]; then
+      return
+    fi
+    print_info "检测到 Java 版本 ${major:-unknown}，低于 21，准备自动安装 OpenJDK 21。"
+  else
+    print_info "未检测到 Java，准备自动安装 OpenJDK 21。"
+  fi
+  install_with_apt "openjdk-21-jdk" "OpenJDK 21"
+}
+
+ensure_clean_dir() {
+  if [[ -d "$1" ]]; then
+    rm -rf "$1"
+  fi
+  mkdir -p "$1"
+}
+
+ensure_java21
+ensure_maven
+check_command docker
+
+print_step "编译 Spring Boot 应用"
+mvn -q -U -DskipTests clean package
+
+JAR_FILE="$(find "${ROOT_DIR}/target" -maxdepth 1 -type f -name "*.jar" ! -name "*-sources.jar" ! -name "*-javadoc.jar" | head -n 1)"
+[[ -n "${JAR_FILE}" ]] || abort "未找到可用的 JAR 文件，请检查 Maven 构建结果。"
+print_info "找到 JAR：${JAR_FILE}"
+
+print_step "构建应用镜像 ${APP_IMAGE}"
+docker build \
+  --build-arg VERSION="${VERSION}" \
+  --build-arg BUILD_TIME="$(date -u +"%Y-%m-%dT%H:%M:%SZ")" \
+  -t "${APP_IMAGE}" \
+  "${ROOT_DIR}"
+
+print_step "拉取依赖镜像"
+docker pull "${PGVECTOR_IMAGE}"
+docker pull "${REDIS_IMAGE}"
+
+print_step "生成离线交付包目录 ${OUTPUT_DIR}"
+ensure_clean_dir "${OUTPUT_DIR}"
+mkdir -p "${IMAGES_DIR}" "${CONFIG_DIR}" "${INIT_DIR}"
+
+if [[ -f "${ROOT_DIR}/src/main/resources/application.yml" ]]; then
+  cp "${ROOT_DIR}/src/main/resources/application.yml" "${CONFIG_DIR}/application.yml"
+  print_info "已复制默认 application.yml 到 ${CONFIG_DIR}/application.yml"
+else
+  print_info "未找到 src/main/resources/application.yml，可手动放置到 ${CONFIG_DIR}/application.yml"
+fi
+
+cat > "${OUTPUT_DIR}/.env" <<EOF
+# PostgreSQL (pgvector)
+POSTGRES_DB=${POSTGRES_DB}
+POSTGRES_USER=${POSTGRES_USER}
+POSTGRES_PASSWORD=${POSTGRES_PASSWORD}
+POSTGRES_PORT=${POSTGRES_PORT}
+
+# Redis
+REDIS_PASSWORD=${REDIS_PASSWORD}
+REDIS_PORT=${REDIS_PORT}
+
+# 应用
+APP_PORT=${APP_PORT}
+EOF
+
+cat > "${INIT_DIR}/001-create-vector-extension.sql" <<'EOF'
+CREATE EXTENSION IF NOT EXISTS vector;
+EOF
+
+cat > "${OUTPUT_DIR}/docker-compose.yml" <<EOF
+version: "3.8"
+
+services:
+  pgvector:
+    image: ${PGVECTOR_IMAGE}
+    container_name: ${APP_NAME}-pgvector
+    restart: unless-stopped
+    environment:
+      POSTGRES_DB: \${POSTGRES_DB}
+      POSTGRES_USER: \${POSTGRES_USER}
+      POSTGRES_PASSWORD: \${POSTGRES_PASSWORD}
+      PGDATA: /var/lib/postgresql/data/pgdata
+    ports:
+      - "\${POSTGRES_PORT}:5432"
+    volumes:
+      - pgvector_data:/var/lib/postgresql/data
+      - ./pgvector-init:/docker-entrypoint-initdb.d
+    healthcheck:
+      test: ["CMD-SHELL", "pg_isready -U \${POSTGRES_USER} -d \${POSTGRES_DB}"]
+      interval: 20s
+      timeout: 5s
+      retries: 5
+
+  redis:
+    image: ${REDIS_IMAGE}
+    container_name: ${APP_NAME}-redis
+    restart: unless-stopped
+    command: ["redis-server", "--appendonly", "yes", "--requirepass", "\${REDIS_PASSWORD}"]
+    ports:
+      - "\${REDIS_PORT}:6379"
+    volumes:
+      - redis_data:/data
+    healthcheck:
+      test: ["CMD", "redis-cli", "-a", "\${REDIS_PASSWORD}", "ping"]
+      interval: 20s
+      timeout: 5s
+      retries: 5
+
+  app:
+    image: ${APP_IMAGE}
+    container_name: ${APP_NAME}-app
+    restart: unless-stopped
+    depends_on:
+      pgvector:
+        condition: service_healthy
+      redis:
+        condition: service_healthy
+    ports:
+      - "\${APP_PORT}:8000"
+    environment:
+      SPRING_DATASOURCE_URL: jdbc:postgresql://pgvector:5432/\${POSTGRES_DB}
+      SPRING_DATASOURCE_USERNAME: \${POSTGRES_USER}
+      SPRING_DATASOURCE_PASSWORD: \${POSTGRES_PASSWORD}
+      SPRING_DATA_REDIS_HOST: redis
+      SPRING_DATA_REDIS_PORT: 6379
+      SPRING_DATA_REDIS_PASSWORD: \${REDIS_PASSWORD}
+      SERVER_PORT: 8000
+    volumes:
+      - ./config/application.yml:/app/config/application.yml:ro
+      - app_uploads:/app/uploads
+      - app_extracts:/app/extracts
+      - app_logs:/app/logs
+
+volumes:
+  pgvector_data:
+  redis_data:
+  app_uploads:
+  app_extracts:
+  app_logs:
+EOF
+
+cat > "${OUTPUT_DIR}/README.md" <<EOF
+# Docker 离线部署包（${APP_NAME} ${VERSION}）
+
+## 包含内容
+- 应用镜像：${APP_IMAGE}
+- pgvector 数据库镜像：${PGVECTOR_IMAGE}
+- Redis 缓存镜像：${REDIS_IMAGE}
+- docker-compose.yml / .env / config / pgvector-init 目录
+
+## 使用步骤
+1. 将整个目录拷贝到目标服务器并进入该目录：
+   \`\`\`
+   cd $(basename "${OUTPUT_DIR}")
+   \`\`\`
+2. 导入离线镜像：
+   \`\`\`
+   docker load -i images/app-${VERSION}.tar
+   docker load -i images/pgvector.tar
+   docker load -i images/redis.tar
+   \`\`\`
+3. 根据需要修改 \`.env\`、\`config/application.yml\`。
+4. 启动：
+   \`\`\`
+   docker compose up -d
+   \`\`\`
+5. 查看状态：
+   \`\`\`
+   docker compose ps
+   docker compose logs -f app
+   \`\`\`
+
+缺省账号密码可通过执行脚本前设置环境变量覆盖，如：
+\`\`\`
+POSTGRES_PASSWORD=StrongPass REDIS_PASSWORD=Secret ./build-docker-bundle.sh
+\`\`\`
+EOF
+
+print_step "保存离线镜像到 ${IMAGES_DIR}"
+docker save "${APP_IMAGE}" -o "${IMAGES_DIR}/app-${VERSION}.tar"
+docker save "${PGVECTOR_IMAGE}" -o "${IMAGES_DIR}/pgvector.tar"
+docker save "${REDIS_IMAGE}" -o "${IMAGES_DIR}/redis.tar"
+
+print_step "打包完成"
+print_info "输出目录：${OUTPUT_DIR}"
+print_info "镜像文件：${IMAGES_DIR}"
+print_info "可根据需要编辑 ${OUTPUT_DIR}/.env 与 ${CONFIG_DIR}/application.yml 后，执行 docker compose up -d 启动。"
+
diff --git a/config/application.yml.template b/config/application.yml.template
new file mode 100644
index 0000000..15a9c90
--- /dev/null
+++ b/config/application.yml.template
@@ -0,0 +1,145 @@
+server:
+  port: 8080
+
+spring:
+  application:
+    name: gdyd_zhpb_zgf
+  
+  # PostgreSQL 单数据源配置
+  datasource:
+      driver-class-name: org.postgresql.Driver
+      url: jdbc:postgresql://127.0.0.1:5432/gdyd_zhpb_zgf
+      username: postgres
+      password: 123456
+  
+  # JPA 配置
+  jpa:
+    database: postgresql
+    hibernate:
+      ddl-auto: update
+    show-sql: true
+    properties:
+      hibernate:
+        dialect: org.hibernate.dialect.PostgreSQLDialect
+        format_sql: true
+        jdbc.time_zone: Asia/Shanghai
+  
+  # Redis 配置
+  data:
+    redis:
+      host: 127.0.0.1
+      port: 6379
+      password: 
+      database: 0
+      timeout: 6000ms
+      lettuce:
+        pool:
+          max-active: 8
+          max-wait: -1ms
+          max-idle: 8
+          min-idle: 0
+  
+  # 文件上传配置
+  servlet:
+    multipart:
+      # 单个文件最大大小 (500MB)
+      max-file-size: 500MB
+      # 整个请求最大大小 (500MB)
+      max-request-size: 500MB
+      # 文件写入磁盘的阈值 (10MB，超过此大小会写入临时文件)
+      file-size-threshold: 10MB
+      # 启用multipart上传
+      enabled: true
+
+# SSO配置
+sso:
+  # ES接口基础地址
+  es-base-url: http://es-integration.es-uat-paas:8890
+  # 获取用户信息接口路径
+  getUserInfoPath: /bi/getUserInfoByTicket
+
+# 文件管理配置
+file:
+  # 文件上传存储路径
+  upload-path: ./uploads
+  # ZIP文件解压路径
+  extract-path: ./extracts
+  # 支持的文件类型
+  allowed-types: pdf,doc,docx,txt
+  # 是否只返回技术文件类型（PDF、DOC、DOCX、PPT、PPTX、TXT、JSON）
+  filter:
+    technical-only: true
+
+# 向量化服务配置
+vectorize:
+  # 向量化接口基础地址
+  base-url: http://localhost:8001
+  # 向量化接口路径
+  api-path: /vectorize
+  # 向量维度（从接口返回的dimension字段获取，默认896）
+  dimension: 896
+
+# 向量化模型阈值配置（按模型区分）
+vector-limits:
+  default:
+    max-chars: 300
+    max-tokens: 512
+  text-embedding-ada-002:
+    max-chars: 300
+    max-tokens: 512
+  embed-bge-large-zh-v1.5:
+    max-chars: 300
+    max-tokens: 512
+
+# Swagger/OpenAPI 配置
+springdoc:
+  api-docs:
+    path: /api-docs
+    enabled: true
+  swagger-ui:
+    path: /swagger-ui.html
+    enabled: true
+    tags-sorter: alpha
+    operations-sorter: alpha
+    try-it-out-enabled: true
+    display-request-duration: true
+  show-actuator: false
+  # 排除异常处理器包，避免兼容性问题
+  packages-to-exclude: com.zhpb.gdyd_zhpb_zgf.exception
+  # 不扫描ControllerAdvice类
+  use-management-port: false
+
+# Actuator 配置
+management:
+  endpoints:
+    web:
+      exposure:
+        include: health,info
+  endpoint:
+    health:
+      show-details: always
+  health:
+    indicators:
+      enabled: true
+
+# AI模型配置
+ai:
+  model-type: public  # public: 公网模型(DeepSeek), private: 私有模型
+
+# DeepSeek AI 配置 (公网模型)
+deepseek:
+  api-key: sk-62f7d771d03046fa8cf77a4adfb9048f  # 建议通过环境变量设置
+  api-url: https://api.deepseek.com/v1/chat/completions
+  model: deepseek-chat
+  temperature: 0.7
+  max-tokens: 4096
+
+# 私有大模型配置
+private-model:
+  url: http://IP:PORT/bigmodel_infer_gateway/v1/service  # 私有模型API地址
+  appid: ""  # 私有模型AppID
+  appkey: ""  # 私有模型AppKey
+  capabilityname: semantic0000000000000000  # 能力名称
+  model: JiuTian-75B-8K  # 私有模型名称
+  temperature: 0.1  # 温度参数
+  max-tokens: 1024  # 最大token数
diff --git a/docker-compose.yml b/docker-compose.yml
new file mode 100644
index 0000000..88ea648
--- /dev/null
+++ b/docker-compose.yml
@@ -0,0 +1,135 @@
+version: '3.8'
+
+services:
+  # MySQL数据库
+  mysql:
+    image: mysql:8.0
+    container_name: smart-qa-mysql
+    restart: unless-stopped
+    environment:
+      MYSQL_ROOT_PASSWORD: rootpassword
+      MYSQL_DATABASE: smart_qa
+      MYSQL_USER: appuser
+      MYSQL_PASSWORD: apppassword
+      MYSQL_CHARSET: utf8mb4
+      MYSQL_COLLATION: utf8mb4_unicode_ci
+    ports:
+      - "3306:3306"
+    volumes:
+      - mysql_data:/var/lib/mysql
+      - ./docker/mysql/init:/docker-entrypoint-initdb.d
+      - ./docker/mysql/conf.d:/etc/mysql/conf.d
+    networks:
+      - smart-qa-network
+    healthcheck:
+      test: ["CMD", "mysqladmin", "ping", "-h", "localhost"]
+      timeout: 20s
+      retries: 10
+
+  # PostgreSQL数据库 (用于向量存储)
+  postgres:
+    image: postgres:15
+    container_name: smart-qa-postgres
+    restart: unless-stopped
+    environment:
+      POSTGRES_DB: smart_qa_vector
+      POSTGRES_USER: vectoruser
+      POSTGRES_PASSWORD: vectorpassword
+      POSTGRES_INITDB_ARGS: "--encoding=UTF-8"
+    ports:
+      - "5432:5432"
+    volumes:
+      - postgres_data:/var/lib/postgresql/data
+      - ./docker/postgres/init:/docker-entrypoint-initdb.d
+    networks:
+      - smart-qa-network
+    healthcheck:
+      test: ["CMD-SHELL", "pg_isready -U vectoruser -d smart_qa_vector"]
+      interval: 30s
+      timeout: 10s
+      retries: 5
+
+  # Redis缓存
+  redis:
+    image: redis:7-alpine
+    container_name: smart-qa-redis
+    restart: unless-stopped
+    ports:
+      - "6379:6379"
+    volumes:
+      - redis_data:/data
+      - ./docker/redis/redis.conf:/etc/redis/redis.conf
+    command: redis-server /etc/redis/redis.conf
+    networks:
+      - smart-qa-network
+    healthcheck:
+      test: ["CMD", "redis-cli", "ping"]
+      interval: 30s
+      timeout: 10s
+      retries: 3
+
+  # 应用服务
+  app:
+    image: smart-qa-app:v202511191435
+    container_name: smart-qa-app
+    restart: unless-stopped
+    ports:
+      - "8080:8080"
+    environment:
+      # 数据库配置
+      - SPRING_DATASOURCE_MYSQL_URL=jdbc:mysql://mysql:3306/smart_qa?useSSL=false&allowPublicKeyRetrieval=true&serverTimezone=Asia/Shanghai
+      - SPRING_DATASOURCE_MYSQL_USERNAME=appuser
+      - SPRING_DATASOURCE_MYSQL_PASSWORD=apppassword
+
+      - SPRING_DATASOURCE_PGSQL_URL=jdbc:postgresql://postgres:5432/smart_qa_vector
+      - SPRING_DATASOURCE_PGSQL_USERNAME=vectoruser
+      - SPRING_DATASOURCE_PGSQL_PASSWORD=vectorpassword
+
+      # Redis配置
+      - SPRING_DATA_REDIS_HOST=redis
+      - SPRING_DATA_REDIS_PORT=6379
+
+      # 应用配置
+      - SERVER_PORT=8080
+      - LOGGING_LEVEL_COM_ZHPB=INFO
+      - FILE_UPLOAD_MAX_SIZE=100MB
+      - FILE_UPLOAD_MAX_REQUEST_SIZE=100MB
+
+      # 外部配置文件路径
+      - SPRING_CONFIG_LOCATION=file:/app/config/application.yml
+    volumes:
+      # 挂载配置文件
+      - ./config/application.yml:/app/config/application.yml:ro
+      # 挂载上传文件目录
+      - uploads_data:/app/uploads
+      # 挂载解压文件目录
+      - extracts_data:/app/extracts
+      # 挂载日志目录
+      - logs_data:/app/logs
+    depends_on:
+      mysql:
+        condition: service_healthy
+      postgres:
+        condition: service_healthy
+      redis:
+        condition: service_healthy
+    networks:
+      - smart-qa-network
+
+volumes:
+  mysql_data:
+    driver: local
+  postgres_data:
+    driver: local
+  redis_data:
+    driver: local
+  uploads_data:
+    driver: local
+  extracts_data:
+    driver: local
+  logs_data:
+    driver: local
+
+networks:
+  smart-qa-network:
+    driver: bridge
diff --git a/docker/mysql/conf.d/mysql.cnf b/docker/mysql/conf.d/mysql.cnf
new file mode 100644
index 0000000..500b0e6
--- /dev/null
+++ b/docker/mysql/conf.d/mysql.cnf
@@ -0,0 +1,29 @@
+[mysql]
+default-character-set=utf8mb4
+
+[mysqld]
+# 基本设置
+character-set-server=utf8mb4
+collation-server=utf8mb4_unicode_ci
+init_connect='SET NAMES utf8mb4'
+
+# 性能优化
+innodb_buffer_pool_size=256M
+innodb_log_file_size=64M
+innodb_flush_log_at_trx_commit=1
+innodb_flush_method=O_DIRECT
+
+# 连接设置
+max_connections=200
+wait_timeout=28800
+interactive_timeout=28800
+
+# 日志设置
+general_log=0
+slow_query_log=1
+slow_query_log_file=/var/lib/mysql/mysql-slow.log
+long_query_time=2
+
+# 其他设置
+sql_mode=STRICT_TRANS_TABLES,NO_ZERO_IN_DATE,NO_ZERO_DATE,ERROR_FOR_DIVISION_BY_ZERO,NO_ENGINE_SUBSTITUTION
+lower_case_table_names=1
diff --git a/docker/postgres/init/01-init-db.sql b/docker/postgres/init/01-init-db.sql
new file mode 100644
index 0000000..0e83fdb
--- /dev/null
+++ b/docker/postgres/init/01-init-db.sql
@@ -0,0 +1,11 @@
+-- PostgreSQL 初始化脚本
+-- 用于向量数据库的初始化
+
+-- 创建向量扩展（如果需要的话）
+-- CREATE EXTENSION IF NOT EXISTS vector;
+
+-- 设置默认编码
+SET client_encoding = 'UTF8';
+
+-- 创建必要的模式和权限
+-- 这里可以添加更多的初始化SQL
diff --git a/docker/redis/redis.conf b/docker/redis/redis.conf
new file mode 100644
index 0000000..7ac4c29
--- /dev/null
+++ b/docker/redis/redis.conf
@@ -0,0 +1,37 @@
+# Redis配置文件 - 适用于Docker环境
+
+# 网络配置
+bind 0.0.0.0
+port 6379
+timeout 0
+tcp-keepalive 300
+
+# 内存管理
+maxmemory 256mb
+maxmemory-policy allkeys-lru
+
+# 持久化
+save 900 1
+save 300 10
+save 60 10000
+
+# 日志
+loglevel notice
+logfile ""
+
+# 安全
+protected-mode no
+requirepass ""
+
+# 其他优化
+tcp-backlog 511
+databases 16
+
+# AOF持久化
+appendonly yes
+appendfilename "appendonly.aof"
+appendfsync everysec
+
+# 慢日志
+slowlog-log-slower-than 10000
+slowlog-max-len 128
diff --git a/images/.DS_Store b/images/.DS_Store
new file mode 100644
index 0000000..d65402c
Binary files /dev/null and b/images/.DS_Store differ
diff --git a/images/主页.jpeg b/images/主页.jpeg
new file mode 100644
index 0000000..56fc69e
Binary files /dev/null and b/images/主页.jpeg differ
diff --git a/images/智能问答对话页面.jpeg b/images/智能问答对话页面.jpeg
new file mode 100644
index 0000000..a290339
Binary files /dev/null and b/images/智能问答对话页面.jpeg differ
diff --git a/images/智能问答选择文件.jpeg b/images/智能问答选择文件.jpeg
new file mode 100644
index 0000000..2ffe1ca
Binary files /dev/null and b/images/智能问答选择文件.jpeg differ
diff --git a/images/智能问答页面.jpeg b/images/智能问答页面.jpeg
new file mode 100644
index 0000000..7450c20
Binary files /dev/null and b/images/智能问答页面.jpeg differ
diff --git a/images/概括总结.jpeg b/images/概括总结.jpeg
new file mode 100644
index 0000000..05cbebc
Binary files /dev/null and b/images/概括总结.jpeg differ
diff --git a/images/概括总结答案.jpeg b/images/概括总结答案.jpeg
new file mode 100644
index 0000000..ded3d7f
Binary files /dev/null and b/images/概括总结答案.jpeg differ
diff --git a/images/概括总结选择文件.jpeg b/images/概括总结选择文件.jpeg
new file mode 100644
index 0000000..1fb63a6
Binary files /dev/null and b/images/概括总结选择文件.jpeg differ
diff --git a/images/横向对比结果.jpeg b/images/横向对比结果.jpeg
new file mode 100644
index 0000000..bcb8a8a
Binary files /dev/null and b/images/横向对比结果.jpeg differ
diff --git a/images/横向对比页面.jpeg b/images/横向对比页面.jpeg
new file mode 100644
index 0000000..ee8356d
Binary files /dev/null and b/images/横向对比页面.jpeg differ
diff --git a/main(2).py b/main(2).py
new file mode 100644
index 0000000..5ed1153
--- /dev/null
+++ b/main(2).py
@@ -0,0 +1,282 @@
+import json
+import requests
+import base64
+import time
+from typing import Dict, Any, Optional, List
+
+
+class UmiOcrClient:
+    """
+    封装了 Umi-OCR HTTP API 的客户端。
+    提供了文档识别的完整异步流程，以及图片和二维码识别接口。
+    """
+
+    def __init__(self, base_url: str = "http://154.219.106.93:1224"):
+        """
+        初始化客户端。
+        :param base_url: Umi-OCR 服务的基地址。
+        """
+        self.base_url = base_url.rstrip('/')
+        self.headers_json = {"Content-Type": "application/json"}
+        print(f"UmiOcrClient 初始化。目标地址: {self.base_url}")
+
+    def _send_request(self, method: str, endpoint: str, data: Optional[Dict[str, Any]] = None,
+                      files: Optional[Dict[str, Any]] = None) -> Dict[str, Any]:
+        """
+        发送 HTTP 请求的通用内部方法。
+        """
+        url = self.base_url + endpoint
+
+        try:
+            if method.upper() == 'GET':
+                response = requests.get(url)
+            elif method.upper() == 'POST':
+                if files:
+                    # 使用 multipart/form-data for file upload (Step 1)
+                    response = requests.post(url, data=data, files=files, timeout=300)
+                else:
+                    # 使用 application/json for other POST requests
+                    data_str = json.dumps(data)
+                    response = requests.post(url, data=data_str, headers=self.headers_json, timeout=300)
+            else:
+                raise ValueError(f"不支持的 HTTP 方法: {method}")
+
+            # 检查 HTTP 状态码
+            response.raise_for_status()
+
+            # 解析响应 JSON
+            return response.json()
+
+        except requests.exceptions.RequestException as e:
+            print(f"请求 {url} 失败。")
+            raise Exception(f"请求错误: {e}")
+        except json.JSONDecodeError as e:
+            # 某些接口（如 /api/doc/clear）可能返回空响应体，这里需要容错
+            if response.status_code == 200 and not response.text:
+                return {"code": 100, "data": "Success (No content)"}
+            print(f"响应内容不是有效的 JSON。原始内容:\n{response.text}")
+            raise Exception(f"JSON 解析错误: {e}")
+
+    # ==========================================================================
+    # 📚 文档识别（PDF等文件）接口封装
+    # ==========================================================================
+
+    def doc_get_options(self) -> Dict[str, Any]:
+        """ 0. 准备工作：参数查询 (/api/doc/get_options) """
+        endpoint = "/api/doc/get_options"
+        return self._send_request("GET", endpoint)
+
+    def doc_upload(self, file_path: str, options: Optional[Dict[str, Any]] = None) -> str:
+        """ 1. 上传待识别文档 (/api/doc/upload)，返回任务ID """
+        print(f"🚀 上传文件: {file_path}")
+        endpoint = "/api/doc/upload"
+
+        # 准备 form-data 的 'json' 部分
+        json_data = json.dumps(options if options is not None else {})
+
+        # 准备 form-data 的 'file' 部分
+        files = {'file': open(file_path, 'rb')}
+        data = {'json': json_data}
+
+        try:
+            response = self._send_request("POST", endpoint, data=data, files=files)
+
+            if response.get('code') == 100:
+                return response.get('data')  # 任务ID
+            else:
+                raise Exception(f"文件上传失败: {response.get('data')}")
+        finally:
+            # 确保文件流关闭
+            if 'file' in files and files['file'].closed == False:
+                files['file'].close()
+
+    def doc_query_result(self, task_id: str, is_data: bool = False, is_unread: bool = True,
+                         data_format: str = "dict") -> Dict[str, Any]:
+        """ 2. 查询任务状态 (/api/doc/result) """
+        endpoint = "/api/doc/result"
+        request_data = {
+            "id": task_id,
+            "is_data": is_data,
+            "is_unread": is_unread,
+            "format": data_format,
+        }
+        return self._send_request("POST", endpoint, request_data)
+
+    def doc_get_download_link(self, task_id: str, file_types: List[str] = ["pdfLayered"], ignore_blank: bool = True) -> \
+    Dict[str, str]:
+        """ 3. 获取结果下载链接 (/api/doc/download) """
+        endpoint = "/api/doc/download"
+        request_data = {
+            "id": task_id,
+            "file_types": file_types,
+            "ignore_blank": ignore_blank,
+        }
+        response = self._send_request("POST", endpoint, request_data)
+
+        if response.get('code') == 100:
+            return {
+                "download_url": self.base_url + response.get('data'),  # 完整的下载链接
+                "file_name": response.get('name')
+            }
+        else:
+            raise Exception(f"生成下载链接失败: {response.get('data')}")
+
+    def doc_clear_task(self, task_id: str) -> bool:
+        """ 5. 任务清理 (/api/doc/clear/<id>) """
+        endpoint = f"/api/doc/clear/{task_id}"
+        response = self._send_request("GET", endpoint)
+
+        if response.get('code') == 100:
+            print(f"🧹 任务ID {task_id} 清理成功。")
+            return True
+        else:
+            print(f"⚠️ 任务ID {task_id} 清理失败或不存在: {response.get('data')}")
+            return False
+
+    # --- 高级同步流程函数 ---
+
+    def run_doc_ocr_sync(self, file_path: str, options: Optional[Dict[str, Any]] = None,
+                         download_file_types: List[str] = ["pdfLayered"], interval: int = 5) -> Dict[str, Any]:
+        """
+        同步执行整个文档OCR流程 (上传 -> 轮询 -> 下载链接 -> 清理)。
+        """
+        task_id = None
+        try:
+            # 1. 上传文件，获取任务ID
+            task_id = self.doc_upload(file_path, options)
+            print(f"📝 任务启动成功，任务ID: {task_id}")
+
+            # 2. 轮询任务状态
+            print("⏳ 正在轮询任务状态...")
+            while True:
+                status_res = self.doc_query_result(task_id, is_data=False)
+
+                is_done = status_res.get('is_done', False)
+                state = status_res.get('state', 'unknown')
+                processed = status_res.get('processed_count', 0)
+                total = status_res.get('pages_count', '?')
+
+                print(f"   - 状态: {state.upper()} ({processed}/{total} 页已处理)")
+
+                if is_done:
+                    if state == "success":
+                        print("✅ 任务完成！")
+                        break
+                    else:
+                        message = status_res.get('message', '未知错误')
+                        raise Exception(f"任务执行失败 (State: {state}): {message}")
+
+                time.sleep(interval)  # 休息指定间隔时间后继续查询
+
+            # 3. 获取下载链接
+            download_info = self.doc_get_download_link(task_id, download_file_types)
+            print(f"🔗 下载链接生成成功。文件: {download_info['file_name']}")
+
+            # 4. 获取识别文本（可选步骤，通过查询获取）
+            text_result = self.doc_query_result(task_id, is_data=True, data_format="text").get('data')
+
+            return {
+                "code": 100,
+                "task_id": task_id,
+                "download_info": download_info,
+                "recognized_text": text_result,
+            }
+
+        except Exception as e:
+            # 捕捉所有错误并报告
+            raise Exception(f"文档识别流程失败: {e}")
+
+        finally:
+            # 5. 任务清理 (无论成功失败都清理)
+            if task_id:
+                self.doc_clear_task(task_id)
+
+    # ==========================================================================
+    # 📸 其他接口（保留自上次封装）
+    # ==========================================================================
+
+    def ocr_base64_image(self, base64_image: str, options: Optional[Dict[str, Any]] = None) -> Dict[str, Any]:
+        """ 图片OCR：Base64 识别接口 (/api/ocr) """
+        print("\n--- 调用图片OCR Base64 识别接口 (/api/ocr) ---")
+        endpoint = "/api/ocr"
+        request_data = {
+            "base64": base64_image,
+            "options": options if options is not None else {}
+        }
+        return self._send_request("POST", endpoint, request_data)
+
+    def recognize_qrcode_base64(self, base64_image: str) -> Dict[str, Any]:
+        """ 【推测接口】二维码：Base64 识别 (/api/qrcode/recognize) """
+        print("\n--- 调用二维码 Base64 识别接口（推测: /api/qrcode/recognize）---")
+        endpoint = "/api/qrcode/recognize"
+        request_data = {"base64": base64_image}
+        return self._send_request("POST", endpoint, request_data)
+
+    def generate_qrcode(self, text: str, options: Optional[Dict[str, Any]] = None) -> Dict[str, Any]:
+        """ 【推测接口】二维码：从文本生成图片 (/api/qrcode/generate) """
+        print("\n--- 调用二维码生成接口（推测: /api/qrcode/generate）---")
+        endpoint = "/api/qrcode/generate"
+        request_data = {
+            "text": text,
+            "options": options if options is not None else {}
+        }
+        return self._send_request("POST", endpoint, request_data)
+
+
+# ==============================================================================
+# 示例用法
+# ==============================================================================
+
+if __name__ == '__main__':
+    # 请根据您的实际情况修改以下参数
+    TEST_PDF_PATH = r"D:\daima\java\gdyd_zhpb_zgf\第2章 方案建议书.pdf"
+
+    # 实例化客户端
+    client = UmiOcrClient()
+
+    # --- 演示文档识别流程 ---
+    print("\n" + "=" * 50)
+    print("      演示：同步文档识别流程 (Step 1-5)")
+    print("=" * 50)
+
+    try:
+        # 尝试使用 run_doc_ocr_sync 接口
+        # 假设我们想要中文识别，并生成双层PDF和纯文本txt
+        doc_options = {
+            "ocr.language": "models/config_chinese.txt",
+            "pageRangeEnd": 1  # 限制只识别前5页
+        }
+
+        # ⚠️ 注意: 运行此段代码前，请确保您的目录下存在 TEST_PDF_PATH 所指的文件
+        final_result = client.run_doc_ocr_sync(
+            file_path=TEST_PDF_PATH,
+            options=doc_options,
+            download_file_types=["pdfLayered", "txtPlain"],
+            interval=3  # 轮询间隔3秒
+        )
+
+        print("\n🎉 整个文档识别任务执行完毕。")
+        print("--- 总结结果 ---")
+        print(f"任务ID: {final_result['task_id']}")
+        print(f"下载文件名: {final_result['download_info']['file_name']}")
+        print(f"下载链接: {final_result['download_info']['download_url']}")
+        print(f"部分识别文本 (前100字): \n{final_result['recognized_text']}...")
+
+    except Exception as e:
+        print(f"\n❌ 文档识别流程中断/失败。错误信息: {e}")
+        print("提示：请确认 Umi-OCR 是否运行在目标 IP/端口，并且文件路径正确。")
+
+    print("\n" + "=" * 50)
+    print("      演示：图片 Base64 识别")
+    print("=" * 50)
+
+    # 演示图片识别 (与上次封装相同，确认接口可用)
+    EXAMPLE_BASE64 = "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"
+    try:
+        ocr_result = client.ocr_base64_image(
+            base64_image=EXAMPLE_BASE64,
+            options={"data.format": "text"}
+        )
+        print("✅ Base64 识别成功。识别结果:", ocr_result.get('data'))
+    except Exception as e:
+        print(f"❌ Base64 识别失败。")
\ No newline at end of file
diff --git a/mvnw b/mvnw
new file mode 100755
index 0000000..bd8896b
--- /dev/null
+++ b/mvnw
@@ -0,0 +1,295 @@
+#!/bin/sh
+# ----------------------------------------------------------------------------
+# Licensed to the Apache Software Foundation (ASF) under one
+# or more contributor license agreements.  See the NOTICE file
+# distributed with this work for additional information
+# regarding copyright ownership.  The ASF licenses this file
+# to you under the Apache License, Version 2.0 (the
+# "License"); you may not use this file except in compliance
+# with the License.  You may obtain a copy of the License at
+#
+#    http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing,
+# software distributed under the License is distributed on an
+# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+# KIND, either express or implied.  See the License for the
+# specific language governing permissions and limitations
+# under the License.
+# ----------------------------------------------------------------------------
+
+# ----------------------------------------------------------------------------
+# Apache Maven Wrapper startup batch script, version 3.3.4
+#
+# Optional ENV vars
+# -----------------
+#   JAVA_HOME - location of a JDK home dir, required when download maven via java source
+#   MVNW_REPOURL - repo url base for downloading maven distribution
+#   MVNW_USERNAME/MVNW_PASSWORD - user and password for downloading maven
+#   MVNW_VERBOSE - true: enable verbose log; debug: trace the mvnw script; others: silence the output
+# ----------------------------------------------------------------------------
+
+set -euf
+[ "${MVNW_VERBOSE-}" != debug ] || set -x
+
+# OS specific support.
+native_path() { printf %s\\n "$1"; }
+case "$(uname)" in
+CYGWIN* | MINGW*)
+  [ -z "${JAVA_HOME-}" ] || JAVA_HOME="$(cygpath --unix "$JAVA_HOME")"
+  native_path() { cygpath --path --windows "$1"; }
+  ;;
+esac
+
+# set JAVACMD and JAVACCMD
+set_java_home() {
+  # For Cygwin and MinGW, ensure paths are in Unix format before anything is touched
+  if [ -n "${JAVA_HOME-}" ]; then
+    if [ -x "$JAVA_HOME/jre/sh/java" ]; then
+      # IBM's JDK on AIX uses strange locations for the executables
+      JAVACMD="$JAVA_HOME/jre/sh/java"
+      JAVACCMD="$JAVA_HOME/jre/sh/javac"
+    else
+      JAVACMD="$JAVA_HOME/bin/java"
+      JAVACCMD="$JAVA_HOME/bin/javac"
+
+      if [ ! -x "$JAVACMD" ] || [ ! -x "$JAVACCMD" ]; then
+        echo "The JAVA_HOME environment variable is not defined correctly, so mvnw cannot run." >&2
+        echo "JAVA_HOME is set to \"$JAVA_HOME\", but \"\$JAVA_HOME/bin/java\" or \"\$JAVA_HOME/bin/javac\" does not exist." >&2
+        return 1
+      fi
+    fi
+  else
+    JAVACMD="$(
+      'set' +e
+      'unset' -f command 2>/dev/null
+      'command' -v java
+    )" || :
+    JAVACCMD="$(
+      'set' +e
+      'unset' -f command 2>/dev/null
+      'command' -v javac
+    )" || :
+
+    if [ ! -x "${JAVACMD-}" ] || [ ! -x "${JAVACCMD-}" ]; then
+      echo "The java/javac command does not exist in PATH nor is JAVA_HOME set, so mvnw cannot run." >&2
+      return 1
+    fi
+  fi
+}
+
+# hash string like Java String::hashCode
+hash_string() {
+  str="${1:-}" h=0
+  while [ -n "$str" ]; do
+    char="${str%"${str#?}"}"
+    h=$(((h * 31 + $(LC_CTYPE=C printf %d "'$char")) % 4294967296))
+    str="${str#?}"
+  done
+  printf %x\\n $h
+}
+
+verbose() { :; }
+[ "${MVNW_VERBOSE-}" != true ] || verbose() { printf %s\\n "${1-}"; }
+
+die() {
+  printf %s\\n "$1" >&2
+  exit 1
+}
+
+trim() {
+  # MWRAPPER-139:
+  #   Trims trailing and leading whitespace, carriage returns, tabs, and linefeeds.
+  #   Needed for removing poorly interpreted newline sequences when running in more
+  #   exotic environments such as mingw bash on Windows.
+  printf "%s" "${1}" | tr -d '[:space:]'
+}
+
+scriptDir="$(dirname "$0")"
+scriptName="$(basename "$0")"
+
+# parse distributionUrl and optional distributionSha256Sum, requires .mvn/wrapper/maven-wrapper.properties
+while IFS="=" read -r key value; do
+  case "${key-}" in
+  distributionUrl) distributionUrl=$(trim "${value-}") ;;
+  distributionSha256Sum) distributionSha256Sum=$(trim "${value-}") ;;
+  esac
+done <"$scriptDir/.mvn/wrapper/maven-wrapper.properties"
+[ -n "${distributionUrl-}" ] || die "cannot read distributionUrl property in $scriptDir/.mvn/wrapper/maven-wrapper.properties"
+
+case "${distributionUrl##*/}" in
+maven-mvnd-*bin.*)
+  MVN_CMD=mvnd.sh _MVNW_REPO_PATTERN=/maven/mvnd/
+  case "${PROCESSOR_ARCHITECTURE-}${PROCESSOR_ARCHITEW6432-}:$(uname -a)" in
+  *AMD64:CYGWIN* | *AMD64:MINGW*) distributionPlatform=windows-amd64 ;;
+  :Darwin*x86_64) distributionPlatform=darwin-amd64 ;;
+  :Darwin*arm64) distributionPlatform=darwin-aarch64 ;;
+  :Linux*x86_64*) distributionPlatform=linux-amd64 ;;
+  *)
+    echo "Cannot detect native platform for mvnd on $(uname)-$(uname -m), use pure java version" >&2
+    distributionPlatform=linux-amd64
+    ;;
+  esac
+  distributionUrl="${distributionUrl%-bin.*}-$distributionPlatform.zip"
+  ;;
+maven-mvnd-*) MVN_CMD=mvnd.sh _MVNW_REPO_PATTERN=/maven/mvnd/ ;;
+*) MVN_CMD="mvn${scriptName#mvnw}" _MVNW_REPO_PATTERN=/org/apache/maven/ ;;
+esac
+
+# apply MVNW_REPOURL and calculate MAVEN_HOME
+# maven home pattern: ~/.m2/wrapper/dists/{apache-maven-<version>,maven-mvnd-<version>-<platform>}/<hash>
+[ -z "${MVNW_REPOURL-}" ] || distributionUrl="$MVNW_REPOURL$_MVNW_REPO_PATTERN${distributionUrl#*"$_MVNW_REPO_PATTERN"}"
+distributionUrlName="${distributionUrl##*/}"
+distributionUrlNameMain="${distributionUrlName%.*}"
+distributionUrlNameMain="${distributionUrlNameMain%-bin}"
+MAVEN_USER_HOME="${MAVEN_USER_HOME:-${HOME}/.m2}"
+MAVEN_HOME="${MAVEN_USER_HOME}/wrapper/dists/${distributionUrlNameMain-}/$(hash_string "$distributionUrl")"
+
+exec_maven() {
+  unset MVNW_VERBOSE MVNW_USERNAME MVNW_PASSWORD MVNW_REPOURL || :
+  exec "$MAVEN_HOME/bin/$MVN_CMD" "$@" || die "cannot exec $MAVEN_HOME/bin/$MVN_CMD"
+}
+
+if [ -d "$MAVEN_HOME" ]; then
+  verbose "found existing MAVEN_HOME at $MAVEN_HOME"
+  exec_maven "$@"
+fi
+
+case "${distributionUrl-}" in
+*?-bin.zip | *?maven-mvnd-?*-?*.zip) ;;
+*) die "distributionUrl is not valid, must match *-bin.zip or maven-mvnd-*.zip, but found '${distributionUrl-}'" ;;
+esac
+
+# prepare tmp dir
+if TMP_DOWNLOAD_DIR="$(mktemp -d)" && [ -d "$TMP_DOWNLOAD_DIR" ]; then
+  clean() { rm -rf -- "$TMP_DOWNLOAD_DIR"; }
+  trap clean HUP INT TERM EXIT
+else
+  die "cannot create temp dir"
+fi
+
+mkdir -p -- "${MAVEN_HOME%/*}"
+
+# Download and Install Apache Maven
+verbose "Couldn't find MAVEN_HOME, downloading and installing it ..."
+verbose "Downloading from: $distributionUrl"
+verbose "Downloading to: $TMP_DOWNLOAD_DIR/$distributionUrlName"
+
+# select .zip or .tar.gz
+if ! command -v unzip >/dev/null; then
+  distributionUrl="${distributionUrl%.zip}.tar.gz"
+  distributionUrlName="${distributionUrl##*/}"
+fi
+
+# verbose opt
+__MVNW_QUIET_WGET=--quiet __MVNW_QUIET_CURL=--silent __MVNW_QUIET_UNZIP=-q __MVNW_QUIET_TAR=''
+[ "${MVNW_VERBOSE-}" != true ] || __MVNW_QUIET_WGET='' __MVNW_QUIET_CURL='' __MVNW_QUIET_UNZIP='' __MVNW_QUIET_TAR=v
+
+# normalize http auth
+case "${MVNW_PASSWORD:+has-password}" in
+'') MVNW_USERNAME='' MVNW_PASSWORD='' ;;
+has-password) [ -n "${MVNW_USERNAME-}" ] || MVNW_USERNAME='' MVNW_PASSWORD='' ;;
+esac
+
+if [ -z "${MVNW_USERNAME-}" ] && command -v wget >/dev/null; then
+  verbose "Found wget ... using wget"
+  wget ${__MVNW_QUIET_WGET:+"$__MVNW_QUIET_WGET"} "$distributionUrl" -O "$TMP_DOWNLOAD_DIR/$distributionUrlName" || die "wget: Failed to fetch $distributionUrl"
+elif [ -z "${MVNW_USERNAME-}" ] && command -v curl >/dev/null; then
+  verbose "Found curl ... using curl"
+  curl ${__MVNW_QUIET_CURL:+"$__MVNW_QUIET_CURL"} -f -L -o "$TMP_DOWNLOAD_DIR/$distributionUrlName" "$distributionUrl" || die "curl: Failed to fetch $distributionUrl"
+elif set_java_home; then
+  verbose "Falling back to use Java to download"
+  javaSource="$TMP_DOWNLOAD_DIR/Downloader.java"
+  targetZip="$TMP_DOWNLOAD_DIR/$distributionUrlName"
+  cat >"$javaSource" <<-END
+	public class Downloader extends java.net.Authenticator
+	{
+	  protected java.net.PasswordAuthentication getPasswordAuthentication()
+	  {
+	    return new java.net.PasswordAuthentication( System.getenv( "MVNW_USERNAME" ), System.getenv( "MVNW_PASSWORD" ).toCharArray() );
+	  }
+	  public static void main( String[] args ) throws Exception
+	  {
+	    setDefault( new Downloader() );
+	    java.nio.file.Files.copy( java.net.URI.create( args[0] ).toURL().openStream(), java.nio.file.Paths.get( args[1] ).toAbsolutePath().normalize() );
+	  }
+	}
+	END
+  # For Cygwin/MinGW, switch paths to Windows format before running javac and java
+  verbose " - Compiling Downloader.java ..."
+  "$(native_path "$JAVACCMD")" "$(native_path "$javaSource")" || die "Failed to compile Downloader.java"
+  verbose " - Running Downloader.java ..."
+  "$(native_path "$JAVACMD")" -cp "$(native_path "$TMP_DOWNLOAD_DIR")" Downloader "$distributionUrl" "$(native_path "$targetZip")"
+fi
+
+# If specified, validate the SHA-256 sum of the Maven distribution zip file
+if [ -n "${distributionSha256Sum-}" ]; then
+  distributionSha256Result=false
+  if [ "$MVN_CMD" = mvnd.sh ]; then
+    echo "Checksum validation is not supported for maven-mvnd." >&2
+    echo "Please disable validation by removing 'distributionSha256Sum' from your maven-wrapper.properties." >&2
+    exit 1
+  elif command -v sha256sum >/dev/null; then
+    if echo "$distributionSha256Sum  $TMP_DOWNLOAD_DIR/$distributionUrlName" | sha256sum -c - >/dev/null 2>&1; then
+      distributionSha256Result=true
+    fi
+  elif command -v shasum >/dev/null; then
+    if echo "$distributionSha256Sum  $TMP_DOWNLOAD_DIR/$distributionUrlName" | shasum -a 256 -c >/dev/null 2>&1; then
+      distributionSha256Result=true
+    fi
+  else
+    echo "Checksum validation was requested but neither 'sha256sum' or 'shasum' are available." >&2
+    echo "Please install either command, or disable validation by removing 'distributionSha256Sum' from your maven-wrapper.properties." >&2
+    exit 1
+  fi
+  if [ $distributionSha256Result = false ]; then
+    echo "Error: Failed to validate Maven distribution SHA-256, your Maven distribution might be compromised." >&2
+    echo "If you updated your Maven version, you need to update the specified distributionSha256Sum property." >&2
+    exit 1
+  fi
+fi
+
+# unzip and move
+if command -v unzip >/dev/null; then
+  unzip ${__MVNW_QUIET_UNZIP:+"$__MVNW_QUIET_UNZIP"} "$TMP_DOWNLOAD_DIR/$distributionUrlName" -d "$TMP_DOWNLOAD_DIR" || die "failed to unzip"
+else
+  tar xzf${__MVNW_QUIET_TAR:+"$__MVNW_QUIET_TAR"} "$TMP_DOWNLOAD_DIR/$distributionUrlName" -C "$TMP_DOWNLOAD_DIR" || die "failed to untar"
+fi
+
+# Find the actual extracted directory name (handles snapshots where filename != directory name)
+actualDistributionDir=""
+
+# First try the expected directory name (for regular distributions)
+if [ -d "$TMP_DOWNLOAD_DIR/$distributionUrlNameMain" ]; then
+  if [ -f "$TMP_DOWNLOAD_DIR/$distributionUrlNameMain/bin/$MVN_CMD" ]; then
+    actualDistributionDir="$distributionUrlNameMain"
+  fi
+fi
+
+# If not found, search for any directory with the Maven executable (for snapshots)
+if [ -z "$actualDistributionDir" ]; then
+  # enable globbing to iterate over items
+  set +f
+  for dir in "$TMP_DOWNLOAD_DIR"/*; do
+    if [ -d "$dir" ]; then
+      if [ -f "$dir/bin/$MVN_CMD" ]; then
+        actualDistributionDir="$(basename "$dir")"
+        break
+      fi
+    fi
+  done
+  set -f
+fi
+
+if [ -z "$actualDistributionDir" ]; then
+  verbose "Contents of $TMP_DOWNLOAD_DIR:"
+  verbose "$(ls -la "$TMP_DOWNLOAD_DIR")"
+  die "Could not find Maven distribution directory in extracted archive"
+fi
+
+verbose "Found extracted Maven distribution directory: $actualDistributionDir"
+printf %s\\n "$distributionUrl" >"$TMP_DOWNLOAD_DIR/$actualDistributionDir/mvnw.url"
+mv -- "$TMP_DOWNLOAD_DIR/$actualDistributionDir" "$MAVEN_HOME" || [ -d "$MAVEN_HOME" ] || die "fail to move MAVEN_HOME"
+
+clean || :
+exec_maven "$@"
diff --git a/mvnw.cmd b/mvnw.cmd
new file mode 100644
index 0000000..92450f9
--- /dev/null
+++ b/mvnw.cmd
@@ -0,0 +1,189 @@
+<# : batch portion
+@REM ----------------------------------------------------------------------------
+@REM Licensed to the Apache Software Foundation (ASF) under one
+@REM or more contributor license agreements.  See the NOTICE file
+@REM distributed with this work for additional information
+@REM regarding copyright ownership.  The ASF licenses this file
+@REM to you under the Apache License, Version 2.0 (the
+@REM "License"); you may not use this file except in compliance
+@REM with the License.  You may obtain a copy of the License at
+@REM
+@REM    http://www.apache.org/licenses/LICENSE-2.0
+@REM
+@REM Unless required by applicable law or agreed to in writing,
+@REM software distributed under the License is distributed on an
+@REM "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+@REM KIND, either express or implied.  See the License for the
+@REM specific language governing permissions and limitations
+@REM under the License.
+@REM ----------------------------------------------------------------------------
+
+@REM ----------------------------------------------------------------------------
+@REM Apache Maven Wrapper startup batch script, version 3.3.4
+@REM
+@REM Optional ENV vars
+@REM   MVNW_REPOURL - repo url base for downloading maven distribution
+@REM   MVNW_USERNAME/MVNW_PASSWORD - user and password for downloading maven
+@REM   MVNW_VERBOSE - true: enable verbose log; others: silence the output
+@REM ----------------------------------------------------------------------------
+
+@IF "%__MVNW_ARG0_NAME__%"=="" (SET __MVNW_ARG0_NAME__=%~nx0)
+@SET __MVNW_CMD__=
+@SET __MVNW_ERROR__=
+@SET __MVNW_PSMODULEP_SAVE=%PSModulePath%
+@SET PSModulePath=
+@FOR /F "usebackq tokens=1* delims==" %%A IN (`powershell -noprofile "& {$scriptDir='%~dp0'; $script='%__MVNW_ARG0_NAME__%'; icm -ScriptBlock ([Scriptblock]::Create((Get-Content -Raw '%~f0'))) -NoNewScope}"`) DO @(
+  IF "%%A"=="MVN_CMD" (set __MVNW_CMD__=%%B) ELSE IF "%%B"=="" (echo %%A) ELSE (echo %%A=%%B)
+)
+@SET PSModulePath=%__MVNW_PSMODULEP_SAVE%
+@SET __MVNW_PSMODULEP_SAVE=
+@SET __MVNW_ARG0_NAME__=
+@SET MVNW_USERNAME=
+@SET MVNW_PASSWORD=
+@IF NOT "%__MVNW_CMD__%"=="" ("%__MVNW_CMD__%" %*)
+@echo Cannot start maven from wrapper >&2 && exit /b 1
+@GOTO :EOF
+: end batch / begin powershell #>
+
+$ErrorActionPreference = "Stop"
+if ($env:MVNW_VERBOSE -eq "true") {
+  $VerbosePreference = "Continue"
+}
+
+# calculate distributionUrl, requires .mvn/wrapper/maven-wrapper.properties
+$distributionUrl = (Get-Content -Raw "$scriptDir/.mvn/wrapper/maven-wrapper.properties" | ConvertFrom-StringData).distributionUrl
+if (!$distributionUrl) {
+  Write-Error "cannot read distributionUrl property in $scriptDir/.mvn/wrapper/maven-wrapper.properties"
+}
+
+switch -wildcard -casesensitive ( $($distributionUrl -replace '^.*/','') ) {
+  "maven-mvnd-*" {
+    $USE_MVND = $true
+    $distributionUrl = $distributionUrl -replace '-bin\.[^.]*$',"-windows-amd64.zip"
+    $MVN_CMD = "mvnd.cmd"
+    break
+  }
+  default {
+    $USE_MVND = $false
+    $MVN_CMD = $script -replace '^mvnw','mvn'
+    break
+  }
+}
+
+# apply MVNW_REPOURL and calculate MAVEN_HOME
+# maven home pattern: ~/.m2/wrapper/dists/{apache-maven-<version>,maven-mvnd-<version>-<platform>}/<hash>
+if ($env:MVNW_REPOURL) {
+  $MVNW_REPO_PATTERN = if ($USE_MVND -eq $False) { "/org/apache/maven/" } else { "/maven/mvnd/" }
+  $distributionUrl = "$env:MVNW_REPOURL$MVNW_REPO_PATTERN$($distributionUrl -replace "^.*$MVNW_REPO_PATTERN",'')"
+}
+$distributionUrlName = $distributionUrl -replace '^.*/',''
+$distributionUrlNameMain = $distributionUrlName -replace '\.[^.]*$','' -replace '-bin$',''
+
+$MAVEN_M2_PATH = "$HOME/.m2"
+if ($env:MAVEN_USER_HOME) {
+  $MAVEN_M2_PATH = "$env:MAVEN_USER_HOME"
+}
+
+if (-not (Test-Path -Path $MAVEN_M2_PATH)) {
+    New-Item -Path $MAVEN_M2_PATH -ItemType Directory | Out-Null
+}
+
+$MAVEN_WRAPPER_DISTS = $null
+if ((Get-Item $MAVEN_M2_PATH).Target[0] -eq $null) {
+  $MAVEN_WRAPPER_DISTS = "$MAVEN_M2_PATH/wrapper/dists"
+} else {
+  $MAVEN_WRAPPER_DISTS = (Get-Item $MAVEN_M2_PATH).Target[0] + "/wrapper/dists"
+}
+
+$MAVEN_HOME_PARENT = "$MAVEN_WRAPPER_DISTS/$distributionUrlNameMain"
+$MAVEN_HOME_NAME = ([System.Security.Cryptography.SHA256]::Create().ComputeHash([byte[]][char[]]$distributionUrl) | ForEach-Object {$_.ToString("x2")}) -join ''
+$MAVEN_HOME = "$MAVEN_HOME_PARENT/$MAVEN_HOME_NAME"
+
+if (Test-Path -Path "$MAVEN_HOME" -PathType Container) {
+  Write-Verbose "found existing MAVEN_HOME at $MAVEN_HOME"
+  Write-Output "MVN_CMD=$MAVEN_HOME/bin/$MVN_CMD"
+  exit $?
+}
+
+if (! $distributionUrlNameMain -or ($distributionUrlName -eq $distributionUrlNameMain)) {
+  Write-Error "distributionUrl is not valid, must end with *-bin.zip, but found $distributionUrl"
+}
+
+# prepare tmp dir
+$TMP_DOWNLOAD_DIR_HOLDER = New-TemporaryFile
+$TMP_DOWNLOAD_DIR = New-Item -Itemtype Directory -Path "$TMP_DOWNLOAD_DIR_HOLDER.dir"
+$TMP_DOWNLOAD_DIR_HOLDER.Delete() | Out-Null
+trap {
+  if ($TMP_DOWNLOAD_DIR.Exists) {
+    try { Remove-Item $TMP_DOWNLOAD_DIR -Recurse -Force | Out-Null }
+    catch { Write-Warning "Cannot remove $TMP_DOWNLOAD_DIR" }
+  }
+}
+
+New-Item -Itemtype Directory -Path "$MAVEN_HOME_PARENT" -Force | Out-Null
+
+# Download and Install Apache Maven
+Write-Verbose "Couldn't find MAVEN_HOME, downloading and installing it ..."
+Write-Verbose "Downloading from: $distributionUrl"
+Write-Verbose "Downloading to: $TMP_DOWNLOAD_DIR/$distributionUrlName"
+
+$webclient = New-Object System.Net.WebClient
+if ($env:MVNW_USERNAME -and $env:MVNW_PASSWORD) {
+  $webclient.Credentials = New-Object System.Net.NetworkCredential($env:MVNW_USERNAME, $env:MVNW_PASSWORD)
+}
+[Net.ServicePointManager]::SecurityProtocol = [Net.SecurityProtocolType]::Tls12
+$webclient.DownloadFile($distributionUrl, "$TMP_DOWNLOAD_DIR/$distributionUrlName") | Out-Null
+
+# If specified, validate the SHA-256 sum of the Maven distribution zip file
+$distributionSha256Sum = (Get-Content -Raw "$scriptDir/.mvn/wrapper/maven-wrapper.properties" | ConvertFrom-StringData).distributionSha256Sum
+if ($distributionSha256Sum) {
+  if ($USE_MVND) {
+    Write-Error "Checksum validation is not supported for maven-mvnd. `nPlease disable validation by removing 'distributionSha256Sum' from your maven-wrapper.properties."
+  }
+  Import-Module $PSHOME\Modules\Microsoft.PowerShell.Utility -Function Get-FileHash
+  if ((Get-FileHash "$TMP_DOWNLOAD_DIR/$distributionUrlName" -Algorithm SHA256).Hash.ToLower() -ne $distributionSha256Sum) {
+    Write-Error "Error: Failed to validate Maven distribution SHA-256, your Maven distribution might be compromised. If you updated your Maven version, you need to update the specified distributionSha256Sum property."
+  }
+}
+
+# unzip and move
+Expand-Archive "$TMP_DOWNLOAD_DIR/$distributionUrlName" -DestinationPath "$TMP_DOWNLOAD_DIR" | Out-Null
+
+# Find the actual extracted directory name (handles snapshots where filename != directory name)
+$actualDistributionDir = ""
+
+# First try the expected directory name (for regular distributions)
+$expectedPath = Join-Path "$TMP_DOWNLOAD_DIR" "$distributionUrlNameMain"
+$expectedMvnPath = Join-Path "$expectedPath" "bin/$MVN_CMD"
+if ((Test-Path -Path $expectedPath -PathType Container) -and (Test-Path -Path $expectedMvnPath -PathType Leaf)) {
+  $actualDistributionDir = $distributionUrlNameMain
+}
+
+# If not found, search for any directory with the Maven executable (for snapshots)
+if (!$actualDistributionDir) {
+  Get-ChildItem -Path "$TMP_DOWNLOAD_DIR" -Directory | ForEach-Object {
+    $testPath = Join-Path $_.FullName "bin/$MVN_CMD"
+    if (Test-Path -Path $testPath -PathType Leaf) {
+      $actualDistributionDir = $_.Name
+    }
+  }
+}
+
+if (!$actualDistributionDir) {
+  Write-Error "Could not find Maven distribution directory in extracted archive"
+}
+
+Write-Verbose "Found extracted Maven distribution directory: $actualDistributionDir"
+Rename-Item -Path "$TMP_DOWNLOAD_DIR/$actualDistributionDir" -NewName $MAVEN_HOME_NAME | Out-Null
+try {
+  Move-Item -Path "$TMP_DOWNLOAD_DIR/$MAVEN_HOME_NAME" -Destination $MAVEN_HOME_PARENT | Out-Null
+} catch {
+  if (! (Test-Path -Path "$MAVEN_HOME" -PathType Container)) {
+    Write-Error "fail to move MAVEN_HOME"
+  }
+} finally {
+  try { Remove-Item $TMP_DOWNLOAD_DIR -Recurse -Force | Out-Null }
+  catch { Write-Warning "Cannot remove $TMP_DOWNLOAD_DIR" }
+}
+
+Write-Output "MVN_CMD=$MAVEN_HOME/bin/$MVN_CMD"
diff --git a/pom.xml b/pom.xml
new file mode 100644
index 0000000..124d36a
--- /dev/null
+++ b/pom.xml
@@ -0,0 +1,171 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<project xmlns="http://maven.apache.org/POM/4.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
+         xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 https://maven.apache.org/xsd/maven-4.0.0.xsd">
+    <modelVersion>4.0.0</modelVersion>
+    <parent>
+        <groupId>org.springframework.boot</groupId>
+        <artifactId>spring-boot-starter-parent</artifactId>
+        <version>3.5.7</version>
+        <relativePath/> <!-- lookup parent from repository -->
+    </parent>
+    <groupId>com.zhpb</groupId>
+    <artifactId>gdyd_zhpb_zgf</artifactId>
+    <version>0.0.1-SNAPSHOT</version>
+    <name>gdyd_zhpb_zgf</name>
+    <description>gdyd_zhpb_zgf</description>
+    <url/>
+    <licenses>
+        <license/>
+    </licenses>
+    <developers>
+        <developer/>
+    </developers>
+    <scm>
+        <connection/>
+        <developerConnection/>
+        <tag/>
+        <url/>
+    </scm>
+    <properties>
+        <java.version>17</java.version>
+        <maven.compiler.source>17</maven.compiler.source>
+        <maven.compiler.target>17</maven.compiler.target>
+    </properties>
+    <dependencies>
+        <dependency>
+            <groupId>org.springframework.boot</groupId>
+            <artifactId>spring-boot-starter-web</artifactId>
+        </dependency>
+
+        <!-- MySQL 数据源 -->
+        <dependency>
+            <groupId>com.mysql</groupId>
+            <artifactId>mysql-connector-j</artifactId>
+        </dependency>
+        <dependency>
+            <groupId>org.springframework.boot</groupId>
+            <artifactId>spring-boot-starter-data-jpa</artifactId>
+        </dependency>
+
+        <!-- PostgreSQL 数据源 -->
+        <dependency>
+            <groupId>org.postgresql</groupId>
+            <artifactId>postgresql</artifactId>
+        </dependency>
+
+        <!-- Redis -->
+        <dependency>
+            <groupId>org.springframework.boot</groupId>
+            <artifactId>spring-boot-starter-data-redis</artifactId>
+        </dependency>
+
+        <!-- Lombok -->
+        <dependency>
+            <groupId>org.projectlombok</groupId>
+            <artifactId>lombok</artifactId>
+            <optional>true</optional>
+        </dependency>
+
+        <!-- Fastjson2 -->
+        <dependency>
+            <groupId>com.alibaba.fastjson2</groupId>
+            <artifactId>fastjson2</artifactId>
+            <version>2.0.43</version>
+        </dependency>
+
+        <!-- Apache HttpComponents -->
+        <dependency>
+            <groupId>org.apache.httpcomponents.client5</groupId>
+            <artifactId>httpclient5</artifactId>
+        </dependency>
+
+        <!-- JWT -->
+        <dependency>
+            <groupId>io.jsonwebtoken</groupId>
+            <artifactId>jjwt-api</artifactId>
+            <version>0.12.3</version>
+        </dependency>
+        <dependency>
+            <groupId>io.jsonwebtoken</groupId>
+            <artifactId>jjwt-impl</artifactId>
+            <version>0.12.3</version>
+            <scope>runtime</scope>
+        </dependency>
+        <dependency>
+            <groupId>io.jsonwebtoken</groupId>
+            <artifactId>jjwt-jackson</artifactId>
+            <version>0.12.3</version>
+            <scope>runtime</scope>
+        </dependency>
+
+        <!-- Swagger/OpenAPI 3 文档 -->
+        <dependency>
+            <groupId>org.springdoc</groupId>
+            <artifactId>springdoc-openapi-starter-webmvc-ui</artifactId>
+            <version>2.6.0</version>
+        </dependency>
+
+        <!-- Spring Boot Actuator 健康检查 -->
+        <dependency>
+            <groupId>org.springframework.boot</groupId>
+            <artifactId>spring-boot-starter-actuator</artifactId>
+        </dependency>
+
+        <!-- Apache PDFBox for PDF reading -->
+        <dependency>
+            <groupId>org.apache.pdfbox</groupId>
+            <artifactId>pdfbox</artifactId>
+            <version>3.0.3</version>
+        </dependency>
+
+        <!-- Apache Commons Compress -->
+        <dependency>
+            <groupId>org.apache.commons</groupId>
+            <artifactId>commons-compress</artifactId>
+            <version>1.20</version>
+        </dependency>
+
+        <!-- Tabula-Java for PDF table extraction -->
+        <dependency>
+            <groupId>technology.tabula</groupId>
+            <artifactId>tabula</artifactId>
+            <version>1.0.5</version>
+            <exclusions>
+                <exclusion>
+                    <groupId>org.slf4j</groupId>
+                    <artifactId>slf4j-simple</artifactId>
+                </exclusion>
+                <exclusion>
+                    <groupId>commons-logging</groupId>
+                    <artifactId>commons-logging</artifactId>
+                </exclusion>
+            </exclusions>
+        </dependency>
+
+        <dependency>
+            <groupId>org.springframework.boot</groupId>
+            <artifactId>spring-boot-starter-test</artifactId>
+            <scope>test</scope>
+        </dependency>
+    </dependencies>
+
+    <build>
+        <plugins>
+            <plugin>
+                <groupId>org.apache.maven.plugins</groupId>
+                <artifactId>maven-compiler-plugin</artifactId>
+                <version>3.13.0</version>
+                <configuration>
+                    <source>21</source>
+                    <target>21</target>
+                    <encoding>UTF-8</encoding>
+                </configuration>
+            </plugin>
+            <plugin>
+                <groupId>org.springframework.boot</groupId>
+                <artifactId>spring-boot-maven-plugin</artifactId>
+            </plugin>
+        </plugins>
+    </build>
+
+</project>
diff --git a/product_manual.html b/product_manual.html
new file mode 100644
index 0000000..4ee7bb9
--- /dev/null
+++ b/product_manual.html
@@ -0,0 +1,653 @@
+<!DOCTYPE html>
+<html lang="zh-CN">
+
+<head>
+    <meta charset="UTF-8">
+    <meta name="viewport" content="width=device-width, initial-scale=1.0">
+    <title>智慧评标-主观分助手 | 产品手册</title>
+    <!-- 引入 Inter 字体，提升阅读体验 -->
+    <link rel="preconnect" href="https://fonts.googleapis.com">
+    <link rel="preconnect" href="https://fonts.gstatic.com" crossorigin>
+    <link href="https://fonts.googleapis.com/css2?family=Inter:wght@300;400;500;600;700&display=swap" rel="stylesheet">
+    <style>
+        :root {
+            --primary-color: #2563eb;
+            --primary-dark: #1d4ed8;
+            --secondary-color: #4f46e5;
+            --text-color: #334155;
+            --heading-color: #0f172a;
+            --bg-color: #f8fafc;
+            --card-bg: #ffffff;
+            --sidebar-width: 280px;
+            --header-height: 64px;
+            --border-color: #e2e8f0;
+            --success-color: #10b981;
+            --warning-color: #f59e0b;
+        }
+
+        * {
+            margin: 0;
+            padding: 0;
+            box-sizing: border-box;
+        }
+
+        html {
+            scroll-behavior: smooth;
+            scroll-padding-top: var(--header-height);
+        }
+
+        body {
+            font-family: 'Inter', -apple-system, BlinkMacSystemFont, "Segoe UI", Roboto, "Helvetica Neue", Arial, sans-serif;
+            color: var(--text-color);
+            background-color: var(--bg-color);
+            display: flex;
+            min-height: 100vh;
+            line-height: 1.6;
+            -webkit-font-smoothing: antialiased;
+        }
+
+        /* Sidebar Styles */
+        .sidebar {
+            width: var(--sidebar-width);
+            background: linear-gradient(180deg, #0f172a 0%, #1e293b 100%);
+            color: white;
+            position: fixed;
+            height: 100vh;
+            overflow-y: auto;
+            border-right: 1px solid rgba(255, 255, 255, 0.1);
+            z-index: 50;
+            transition: transform 0.3s cubic-bezier(0.4, 0, 0.2, 1);
+        }
+
+        .logo {
+            padding: 24px;
+            font-size: 1.25rem;
+            font-weight: 700;
+            border-bottom: 1px solid rgba(255, 255, 255, 0.1);
+            margin-bottom: 10px;
+            display: flex;
+            align-items: center;
+            gap: 12px;
+            color: #fff;
+            letter-spacing: -0.5px;
+        }
+
+        .nav-links {
+            list-style: none;
+            padding: 10px 0;
+        }
+
+        .nav-link {
+            display: flex;
+            align-items: center;
+            padding: 12px 24px;
+            color: #94a3b8;
+            text-decoration: none;
+            transition: all 0.2s;
+            font-size: 0.95rem;
+            font-weight: 500;
+            border-left: 3px solid transparent;
+        }
+
+        .nav-link:hover {
+            color: #fff;
+            background-color: rgba(255, 255, 255, 0.05);
+        }
+
+        .nav-link.active {
+            color: #fff;
+            background-color: rgba(59, 130, 246, 0.1);
+            border-left-color: var(--primary-color);
+        }
+
+        /* Nav Group Title */
+        .nav-group-title {
+            padding: 8px 24px;
+            font-size: 0.75rem;
+            text-transform: uppercase;
+            color: #64748b;
+            font-weight: 600;
+            margin-top: 16px;
+        }
+
+        /* Main Content Styles */
+        .main-content {
+            margin-left: var(--sidebar-width);
+            flex: 1;
+            background-color: var(--bg-color);
+            min-height: 100vh;
+            display: flex;
+            flex-direction: column;
+        }
+
+        .header {
+            background-color: rgba(255, 255, 255, 0.9);
+            backdrop-filter: blur(10px);
+            height: var(--header-height);
+            border-bottom: 1px solid var(--border-color);
+            display: flex;
+            align-items: center;
+            padding: 0 40px;
+            position: sticky;
+            top: 0;
+            z-index: 40;
+            justify-content: space-between;
+        }
+
+        .header h3 {
+            color: var(--text-color);
+            font-size: 1rem;
+            font-weight: 500;
+        }
+
+        .content-wrapper {
+            padding: 40px;
+            max-width: 1000px;
+            margin: 0 auto;
+            width: 100%;
+        }
+
+        section {
+            background: var(--card-bg);
+            border-radius: 16px;
+            padding: 48px;
+            margin-bottom: 32px;
+            box-shadow: 0 4px 6px -1px rgba(0, 0, 0, 0.05), 0 2px 4px -1px rgba(0, 0, 0, 0.03);
+            border: 1px solid var(--border-color);
+        }
+
+        h1 {
+            font-size: 2.5rem;
+            margin-bottom: 1.5rem;
+            color: var(--heading-color);
+            font-weight: 800;
+            letter-spacing: -1px;
+            line-height: 1.2;
+        }
+
+        h2 {
+            font-size: 1.875rem;
+            margin-bottom: 1.5rem;
+            color: var(--heading-color);
+            font-weight: 700;
+            letter-spacing: -0.5px;
+            padding-bottom: 16px;
+            border-bottom: 2px solid var(--border-color);
+            margin-top: 0;
+        }
+
+        h3 {
+            font-size: 1.35rem;
+            margin: 2rem 0 1rem;
+            color: var(--heading-color);
+            font-weight: 600;
+            display: flex;
+            align-items: center;
+            gap: 10px;
+        }
+
+        h4 {
+            font-size: 1.125rem;
+            margin-bottom: 0.75rem;
+            margin-top: 1.5rem;
+            color: var(--heading-color);
+            font-weight: 600;
+        }
+
+        p {
+            margin-bottom: 1.25rem;
+            color: var(--text-color);
+            font-size: 1.05rem;
+        }
+
+        ul,
+        ol {
+            margin-left: 24px;
+            margin-bottom: 1.5rem;
+            color: var(--text-color);
+        }
+
+        li {
+            margin-bottom: 0.75rem;
+        }
+
+        /* Images */
+        .img-container {
+            margin: 24px 0 32px;
+            border-radius: 12px;
+            overflow: hidden;
+            border: 1px solid var(--border-color);
+            box-shadow: 0 10px 15px -3px rgba(0, 0, 0, 0.1);
+        }
+
+        .img-container img {
+            width: 100%;
+            height: auto;
+            display: block;
+        }
+
+        .img-caption {
+            text-align: center;
+            font-size: 0.9rem;
+            color: #64748b;
+            margin-top: 8px;
+            font-style: italic;
+        }
+
+        /* Tables */
+        table {
+            width: 100%;
+            border-collapse: collapse;
+            margin-bottom: 1.5rem;
+            font-size: 0.95rem;
+        }
+
+        th,
+        td {
+            padding: 12px 16px;
+            border: 1px solid var(--border-color);
+            text-align: left;
+        }
+
+        th {
+            background-color: #f1f5f9;
+            font-weight: 600;
+            color: var(--heading-color);
+        }
+
+        td {
+            color: var(--text-color);
+        }
+
+        /* Feature Cards */
+        .feature-grid {
+            display: grid;
+            grid-template-columns: repeat(auto-fit, minmax(280px, 1fr));
+            gap: 24px;
+            margin-top: 24px;
+        }
+
+        .feature-card {
+            background: #fff;
+            padding: 24px;
+            border-radius: 12px;
+            border: 1px solid var(--border-color);
+            transition: transform 0.2s;
+        }
+
+        .feature-card:hover {
+            transform: translateY(-2px);
+            border-color: var(--primary-color);
+        }
+
+        .feature-card h4 {
+            margin-top: 0;
+            color: var(--primary-color);
+        }
+
+        /* Step List */
+        .step-list {
+            counter-reset: step;
+            list-style: none;
+            margin: 0;
+        }
+
+        .step-list li {
+            position: relative;
+            padding-left: 40px;
+            margin-bottom: 16px;
+        }
+
+        .step-list li::before {
+            counter-increment: step;
+            content: counter(step);
+            position: absolute;
+            left: 0;
+            top: 2px;
+            width: 24px;
+            height: 24px;
+            background-color: var(--primary-color);
+            color: white;
+            border-radius: 50%;
+            text-align: center;
+            line-height: 24px;
+            font-size: 0.85rem;
+            font-weight: bold;
+        }
+
+        /* Badge */
+        .badge {
+            display: inline-flex;
+            align-items: center;
+            padding: 4px 12px;
+            border-radius: 9999px;
+            font-size: 0.75rem;
+            font-weight: 700;
+            background-color: #dbeafe;
+            color: var(--primary-color);
+            vertical-align: middle;
+            margin-left: 12px;
+        }
+
+        /* Responsive */
+        @media (max-width: 768px) {
+            .sidebar {
+                transform: translateX(-100%);
+            }
+
+            .main-content {
+                margin-left: 0;
+            }
+
+            .header {
+                padding: 0 20px;
+            }
+
+            .content-wrapper {
+                padding: 20px;
+            }
+        }
+    </style>
+</head>
+
+<body>
+
+    <aside class="sidebar">
+        <div class="logo">
+            <svg width="24" height="24" viewBox="0 0 24 24" fill="none" xmlns="http://www.w3.org/2000/svg">
+                <path d="M12 2L2 7L12 12L22 7L12 2Z" stroke="#3b82f6" stroke-width="2" stroke-linecap="round"
+                    stroke-linejoin="round" />
+                <path d="M2 17L12 22L22 17" stroke="#3b82f6" stroke-width="2" stroke-linecap="round"
+                    stroke-linejoin="round" />
+                <path d="M2 12L12 17L22 12" stroke="#3b82f6" stroke-width="2" stroke-linecap="round"
+                    stroke-linejoin="round" />
+            </svg>
+            智慧评标助手
+        </div>
+        <ul class="nav-links">
+            <li class="nav-item"><a href="#intro" class="nav-link active">产品概述</a></li>
+
+            <div class="nav-group-title">核心功能</div>
+            <li class="nav-item"><a href="#start" class="nav-link">快速入门</a></li>
+            <li class="nav-item"><a href="#summary" class="nav-link">概括总结</a></li>
+            <li class="nav-item"><a href="#qa" class="nav-link">智能问答</a></li>
+            <li class="nav-item"><a href="#comparison" class="nav-link">横向对比</a></li>
+
+            <div class="nav-group-title">其他</div>
+            <li class="nav-item"><a href="#review" class="nav-link">评审与评分</a></li>
+            <li class="nav-item"><a href="#faq" class="nav-link">常见问题</a></li>
+            <li class="nav-item"><a href="#support" class="nav-link">联系方式</a></li>
+        </ul>
+    </aside>
+
+    <main class="main-content">
+        <header class="header">
+            <h3>文档中心 / 用户手册 v1.0</h3>
+            <div style="font-size: 0.9em; color: #64748b;">最后更新: 2025-08-15</div>
+        </header>
+
+        <div class="content-wrapper">
+            <!-- 产品概述 -->
+            <section id="intro">
+                <h1>主观分评审助手 <span class="badge">v1.0</span></h1>
+
+                <!-- 产品介绍视频 -->
+                <div class="img-container" style="border:none; box-shadow:none; background:transparent;">
+                    <video controls width="100%"
+                        style="border-radius: 12px; box-shadow: 0 10px 15px -3px rgba(0, 0, 0, 0.1);">
+                        <source src="./主观分助手.mp4" type="video/mp4">
+                        您的浏览器不支持视频播放，请下载视频观看。
+                    </video>
+                    <div class="img-caption">产品功能演示视频</div>
+                </div>
+
+                <div class="img-container">
+                    <img src="./images/主页.jpeg" alt="产品主页展示">
+                    <div class="img-caption">产品主界面：清晰的卡片式布局与专业配色</div>
+                </div>
+
+                <h3>1.2 产品简介</h3>
+                <p>主观分评审助手v1.0是深度集成于智慧评审系统的AI辅助工具，基于LLM（大型语言模型）和RAG（检索增强生成）技术，提供<strong>智能问答</strong>、<strong>概括总结</strong>、<strong>横向对比</strong>三大核心功能。产品旨在通过智能化手段，帮助用户快速定位投标文件关键信息、建立方案整体认知、实现多供应商方案标准化对比，最终提升评审效率、保障评审公平性。
+                </p>
+
+                <h3>1.3 适用范围</h3>
+                <p>本手册适用于使用智慧评审系统进行招投标项目评审的专家，涵盖技术、财务、服务等各类主观分评审场景。</p>
+
+                <h3>1.4 环境要求</h3>
+                <ul>
+                    <li><strong>运行环境</strong>：需在智慧评审系统内使用，支持主流浏览器（Chrome 90+、Edge 90+、Firefox 88+）；</li>
+                    <li><strong>权限要求</strong>：需具备对应招投标项目的评审权限（由系统管理员分配）；</li>
+                    <li><strong>文档要求</strong>：仅支持可正常读取的PDF、Word格式投标文件，暂不支持加密、手写版、图片密集型文档。</li>
+                </ul>
+            </section>
+
+            <!-- 快速入门 -->
+            <section id="start">
+                <h2>二、核心功能快速入门</h2>
+                <p>产品全程嵌入智慧评审系统，以“评审助手”按钮为核心入口，操作流程遵循“准备-概览-质询-对比-结论”逻辑。</p>
+
+                <div class="feature-grid">
+                    <div class="feature-card">
+                        <h4>Step 1: 启动</h4>
+                        <p>登录系统，在主观分评审项（如服务方案）旁点击【评审助手】按钮。</p>
+                    </div>
+                    <div class="feature-card">
+                        <h4>Step 2: 概览</h4>
+                        <p>使用【概括总结】功能，快速了解各供应商方案框架与核心内容。</p>
+                    </div>
+                    <div class="feature-card">
+                        <h4>Step 3: 质询</h4>
+                        <p>通过【智能问答】功能，精准检索细节，支持跨文档对比提问。</p>
+                    </div>
+                    <div class="feature-card">
+                        <h4>Step 4: 对比</h4>
+                        <p>使用【横向对比】生成多维度对比矩阵，辅助打分。</p>
+                    </div>
+                </div>
+            </section>
+
+            <!-- 概括总结 -->
+            <section id="summary">
+                <h2>3.2 概括总结功能操作</h2>
+                <p>自动提取投标文件的层级目录，生成各章节核心摘要，帮助用户快速建立对方案的整体认知，替代传统逐页阅读。</p>
+
+                <div class="img-container">
+                    <img src="./images/概括总结.jpeg" alt="概括总结界面">
+                    <div class="img-caption">目录提取与章节智能摘要</div>
+                </div>
+
+                <div class="img-container">
+                    <img src="./images/概括总结答案.jpeg" alt="概括总结详细内容">
+                    <div class="img-caption">生成的详细章节总结，支持点击原文</div>
+                </div>
+
+                <h3>操作步骤</h3>
+                <ul class="step-list">
+                    <li>在评审助手主界面，点击【概括总结】按钮，进入功能界面；</li>
+                    <li>系统自动加载当前评审项对应的所有供应商投标文件，左侧显示“供应商列表”；</li>
+                    <li>选择目标供应商，系统自动提取该供应商方案的树状目录（含章节标题及页码）；</li>
+                    <li>点击目录任意章节，右侧“摘要展示区”实时显示该章节AI摘要；</li>
+                    <li>点击摘要中“查看原文”链接，可直达投标文件对应页码预览。</li>
+                </ul>
+
+                <div class="feature-card" style="background:#fff7ed; border-color:#fed7aa; margin-top:20px;">
+                    <h4 style="color:#c2410c">⚠️ 注意事项</h4>
+                    <p style="color:#9a3412; margin-bottom:0">摘要通常30秒内完成。若投标文件有更新，请点击【刷新】按钮重新提取。</p>
+                </div>
+            </section>
+
+            <!-- 智能问答 -->
+            <section id="qa">
+                <h2>3.3 智能问答功能操作</h2>
+                <p>支持自然语言提问，基于检索增强生成(RAG)技术，精准检索相关信息并生成答案，附带原文溯源。</p>
+
+                <div class="img-container">
+                    <img src="./images/智能问答对话页面.jpeg" alt="智能问答对话界面">
+                    <div class="img-caption">支持多轮对话与原文溯源</div>
+                </div>
+
+                <div class="img-container">
+                    <img src="./images/智能问答选择文件.jpeg" alt="多文档选择">
+                    <div class="img-caption">支持同时勾选多个文档进行对比提问</div>
+                </div>
+
+                <h3>操作步骤</h3>
+                <ul class="step-list">
+                    <li>点击【文档问答】按钮，进入功能界面；</li>
+                    <li>通过“文件选择器”勾选需要提问的投标文件（单选深入挖掘，多选横向对比）；</li>
+                    <li>输入自然语言问题（如“A公司自研时序数据库的性能指标是什么？”）；</li>
+                    <li>点击发送，系统生成答案并标注来源（如 <code>来源：P28，3.1章节</code>）；</li>
+                    <li>点击来源标注可直接跳转原文高亮查看。</li>
+                </ul>
+            </section>
+
+            <!-- 横向对比 -->
+            <section id="comparison">
+                <h2>3.4 横向对比功能操作</h2>
+                <p>支持选择多供应商、多维度进行方案对比，系统生成标准化对比矩阵，直观呈现优劣差异。</p>
+
+                <div class="img-container">
+                    <img src="./images/横向对比页面.jpeg" alt="横向对比配置">
+                    <div class="img-caption">Step 1: 配置对比维度与厂商</div>
+                </div>
+
+                <div class="img-container">
+                    <img src="./images/横向对比结果.jpeg" alt="横向对比结果矩阵">
+                    <div class="img-caption">Step 2: 查看多维度对比矩阵结果</div>
+                </div>
+
+                <h3>核心流程</h3>
+                <div class="feature-grid">
+                    <div>
+                        <h4>1. 配置 (P4页面)</h4>
+                        <ul>
+                            <li>选择预设评审模块或添加“自定义维度”</li>
+                            <li>勾选需对比的2家及以上供应商</li>
+                            <li>点击【开始解析对比】启动异步任务</li>
+                        </ul>
+                    </div>
+                    <div>
+                        <h4>2. 查看 (P5页面)</h4>
+                        <ul>
+                            <li>任务完成后自动展示对比矩阵</li>
+                            <li>每行代表一个维度，每列代表一家厂商</li>
+                            <li>若结果不满意，支持单维度【重新生成】</li>
+                        </ul>
+                    </div>
+                </div>
+            </section>
+
+            <!-- 评审与评分 -->
+            <section id="review">
+                <h2>3.5 评审意见填写与评分</h2>
+                <p>完成概览、质询、对比后，可直接在评审助手侧边栏下方的“评审意见”输入区填写评审结论。填写完成后，点击输入区下方【保存】按钮，意见会同步至智慧评审系统的对应区域。</p>
+            </section>
+
+            <!-- 常见问题 -->
+            <section id="faq">
+                <h2>四、常见问题与解决方法</h2>
+                <table>
+                    <thead>
+                        <tr>
+                            <th width="25%">常见问题</th>
+                            <th width="35%">可能原因</th>
+                            <th width="40%">解决方法</th>
+                        </tr>
+                    </thead>
+                    <tbody>
+                        <tr>
+                            <td>点击“评审助手”无响应</td>
+                            <td>浏览器缓存异常；无权限</td>
+                            <td>刷新页面或清除缓存；联系管理员确认权限</td>
+                        </tr>
+                        <tr>
+                            <td>无法生成摘要</td>
+                            <td>文档加密；文档过大</td>
+                            <td>联系招标方获取非加密文档；耐心等待处理</td>
+                        </tr>
+                        <tr>
+                            <td>问答不准确或无来源</td>
+                            <td>问题模糊；文档无相关信息</td>
+                            <td>优化问题表述；确认文档相关性</td>
+                        </tr>
+                        <tr>
+                            <td>横向对比时间过长</td>
+                            <td>对比厂商/维度过多</td>
+                            <td>建议每次对比厂商≤5家，维度≤8个</td>
+                        </tr>
+                    </tbody>
+                </table>
+            </section>
+
+            <!-- 联系方式 -->
+            <section id="support">
+                <h2>五、联系方式与支持</h2>
+                <div class="feature-grid">
+                    <div class="feature-card">
+                        <h4>📞 技术支持</h4>
+                        <p>电话：400-XXXX-XXXX<br><span style="font-size:0.9em; color:#64748b">(工作日 9:00-18:00)</span></p>
+                    </div>
+                    <div class="feature-card">
+                        <h4>📧 邮箱支持</h4>
+                        <p>support@xxx.com<br><span style="font-size:0.9em; color:#64748b">(24小时内回复)</span></p>
+                    </div>
+                    <div class="feature-card">
+                        <h4>💬 系统反馈</h4>
+                        <p>登录系统 -> 【帮助中心】 -> 【问题反馈】提交工单。</p>
+                    </div>
+                </div>
+
+                <div style="margin-top: 24px; font-size: 0.9em; color: #94a3b8; text-align: center;">
+                    注：文档部分辅助内容可能由AI生成，请以实际评审要求为准。
+                </div>
+            </section>
+        </div>
+    </main>
+
+    <script>
+        // 平滑滚动与导航高亮
+        document.querySelectorAll('a[href^="#"]').forEach(anchor => {
+            anchor.addEventListener('click', function (e) {
+                e.preventDefault();
+                const targetId = this.getAttribute('href');
+                const targetSection = document.querySelector(targetId);
+
+                if (targetSection) {
+                    targetSection.scrollIntoView({
+                        behavior: 'smooth',
+                        block: 'start'
+                    });
+
+                    history.pushState(null, null, targetId);
+                }
+            });
+        });
+
+        // 滚动监听 (Scroll Spy)
+        const observerOptions = {
+            root: null,
+            rootMargin: '-20% 0px -60% 0px',
+            threshold: 0
+        };
+
+        const observer = new IntersectionObserver((entries) => {
+            entries.forEach(entry => {
+                if (entry.isIntersecting) {
+                    const id = entry.target.getAttribute('id');
+                    document.querySelectorAll('.nav-link').forEach(link => {
+                        link.classList.remove('active');
+                        if (link.getAttribute('href') === '#' + id) {
+                            link.classList.add('active');
+                        }
+                    });
+                }
+            });
+        }, observerOptions);
+
+        document.querySelectorAll('section').forEach(section => {
+            observer.observe(section);
+        });
+    </script>
+</body>
+
+</html>
\ No newline at end of file
diff --git a/sql/migration_add_catalog_extraction_status.sql b/sql/migration_add_catalog_extraction_status.sql
new file mode 100644
index 0000000..fe82296
--- /dev/null
+++ b/sql/migration_add_catalog_extraction_status.sql
@@ -0,0 +1,30 @@
+-- 添加目录摘取状态相关字段到 t_file 表
+-- 执行时间: 2025-12-02
+
+-- 添加目录摘取状态字段
+ALTER TABLE t_file ADD COLUMN IF NOT EXISTS catalog_extraction_status VARCHAR(32) DEFAULT 'not_started';
+
+-- 添加目录摘取重试次数字段
+ALTER TABLE t_file ADD COLUMN IF NOT EXISTS catalog_extraction_retry_count INT DEFAULT 0 NOT NULL;
+
+-- 添加目录摘取完成时间字段
+ALTER TABLE t_file ADD COLUMN IF NOT EXISTS catalog_extraction_time TIMESTAMP;
+
+-- 添加字段注释
+COMMENT ON COLUMN t_file.catalog_extraction_status IS '目录摘取状态：not_started-未开始, processing-处理中, completed-已完成';
+COMMENT ON COLUMN t_file.catalog_extraction_retry_count IS '目录摘取重试次数（最多重试3次）';
+COMMENT ON COLUMN t_file.catalog_extraction_time IS '目录摘取完成时间';
+
+-- 为已有数据设置初始状态
+-- 如果已经有目录记录，设置为 completed
+UPDATE t_file f
+SET catalog_extraction_status = 'completed',
+    catalog_extraction_time = f.update_time
+WHERE EXISTS (
+    SELECT 1 FROM t_file_directory fd 
+    WHERE fd.file_id = f.file_id
+)
+AND (catalog_extraction_status IS NULL OR catalog_extraction_status = 'not_started');
+
+-- 创建索引以提高查询性能
+CREATE INDEX IF NOT EXISTS idx_file_catalog_status ON t_file(catalog_extraction_status);
diff --git a/src/main/java/com/zhpb/.DS_Store b/src/main/java/com/zhpb/.DS_Store
new file mode 100644
index 0000000..4673513
Binary files /dev/null and b/src/main/java/com/zhpb/.DS_Store differ
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/.DS_Store b/src/main/java/com/zhpb/gdyd_zhpb_zgf/.DS_Store
new file mode 100644
index 0000000..6111cc2
Binary files /dev/null and b/src/main/java/com/zhpb/gdyd_zhpb_zgf/.DS_Store differ
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/GdydZhpbZgfApplication.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/GdydZhpbZgfApplication.java
new file mode 100644
index 0000000..97b3688
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/GdydZhpbZgfApplication.java
@@ -0,0 +1,29 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf
+ * @ClassName: GdydZhpbZgfApplication
+ * @Description: Spring Boot应用主启动类，启用异步支持
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf;
+
+import org.springframework.boot.SpringApplication;
+import org.springframework.boot.autoconfigure.SpringBootApplication;
+import org.springframework.scheduling.annotation.EnableAsync;
+import org.springframework.scheduling.annotation.EnableScheduling;
+
+@SpringBootApplication
+@EnableAsync
+@EnableScheduling
+public class GdydZhpbZgfApplication {
+
+    public static void main(String[] args) {
+        System.out.println("http://127.0.0.1:8080");
+        SpringApplication.run(GdydZhpbZgfApplication.class, args);
+
+    }
+
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/AIModelConfig.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/AIModelConfig.java
new file mode 100644
index 0000000..4a2164b
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/AIModelConfig.java
@@ -0,0 +1,43 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.config
+ * @ClassName: AIModelConfig
+ * @Description: AI模型配置类
+ * @Author: 张志锋
+ * @Date: 2025-11-05
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.config;
+
+import com.zhpb.gdyd_zhpb_zgf.service.AIModelService;
+import com.zhpb.gdyd_zhpb_zgf.service.impl.DeepSeekServiceImpl;
+import com.zhpb.gdyd_zhpb_zgf.service.impl.PrivateModelServiceImpl;
+import org.springframework.beans.factory.annotation.Value;
+import org.springframework.context.annotation.Bean;
+import org.springframework.context.annotation.Configuration;
+import org.springframework.context.annotation.Primary;
+
+/**
+ * AI模型配置类
+ */
+@Configuration
+public class AIModelConfig {
+
+    @Value("${ai.model-type:public}")
+    private String modelType;
+
+    /**
+     * 根据配置选择使用的AI模型服务
+     */
+    @Bean
+    @Primary
+    public AIModelService aiModelService(DeepSeekServiceImpl deepSeekService,
+                                       PrivateModelServiceImpl privateModelService) {
+        if ("private".equalsIgnoreCase(modelType)) {
+            return privateModelService;
+        } else {
+            return deepSeekService;
+        }
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/AsyncConfig.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/AsyncConfig.java
new file mode 100644
index 0000000..1a064ce
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/AsyncConfig.java
@@ -0,0 +1,98 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.config
+ * @ClassName: AsyncConfig
+ * @Description: 异步任务线程池配置
+ * @Author: 张志锋
+ * @Date: 2025-12-11
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.config;
+
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.context.annotation.Bean;
+import org.springframework.context.annotation.Configuration;
+import org.springframework.scheduling.annotation.EnableAsync;
+import org.springframework.scheduling.concurrent.ThreadPoolTaskExecutor;
+
+import java.util.concurrent.Executor;
+import java.util.concurrent.ThreadPoolExecutor;
+
+/**
+ * 异步任务线程池配置
+ * 为AI模型调用和异步处理任务提供专用的线程池
+ */
+@Configuration
+@EnableAsync
+public class AsyncConfig {
+
+    private static final Logger logger = LoggerFactory.getLogger(AsyncConfig.class);
+
+    /**
+     * AI任务专用线程池
+     * 用于章节总结、横向对比等AI相关的异步任务
+     */
+    @Bean(name = "aiTaskExecutor")
+    public Executor aiTaskExecutor() {
+        ThreadPoolTaskExecutor executor = new ThreadPoolTaskExecutor();
+        executor.setCorePoolSize(3);      // 核心线程数
+        executor.setMaxPoolSize(6);       // 最大线程数
+        executor.setQueueCapacity(50);    // 队列容量
+        executor.setThreadNamePrefix("ai-task-");
+        executor.setKeepAliveSeconds(60);
+        
+        // 设置拒绝策略：由调用者线程执行（确保任务不会丢失）
+        executor.setRejectedExecutionHandler(new ThreadPoolExecutor.CallerRunsPolicy());
+        
+        executor.initialize();
+        logger.info("AI任务线程池初始化完成: coreSize={}, maxSize={}, queueCapacity={}", 
+            executor.getCorePoolSize(), executor.getMaxPoolSize(), 50);
+        return executor;
+    }
+
+    /**
+     * 目录抽取专用线程池
+     * 用于文件目录抽取等IO密集型任务
+     */
+    @Bean(name = "directoryTaskExecutor")
+    public Executor directoryTaskExecutor() {
+        ThreadPoolTaskExecutor executor = new ThreadPoolTaskExecutor();
+        executor.setCorePoolSize(2);      // 核心线程数
+        executor.setMaxPoolSize(4);       // 最大线程数
+        executor.setQueueCapacity(100);   // 队列容量
+        executor.setThreadNamePrefix("dir-task-");
+        executor.setKeepAliveSeconds(60);
+        
+        // 设置拒绝策略：由调用者线程执行
+        executor.setRejectedExecutionHandler(new ThreadPoolExecutor.CallerRunsPolicy());
+        
+        executor.initialize();
+        logger.info("目录抽取线程池初始化完成: coreSize={}, maxSize={}, queueCapacity={}", 
+            executor.getCorePoolSize(), executor.getMaxPoolSize(), 100);
+        return executor;
+    }
+
+    /**
+     * 对比分析专用线程池
+     * 用于供应商横向对比等分析任务
+     */
+    @Bean(name = "comparisonTaskExecutor")
+    public Executor comparisonTaskExecutor() {
+        ThreadPoolTaskExecutor executor = new ThreadPoolTaskExecutor();
+        executor.setCorePoolSize(2);      // 核心线程数
+        executor.setMaxPoolSize(4);       // 最大线程数
+        executor.setQueueCapacity(30);    // 队列容量
+        executor.setThreadNamePrefix("comp-task-");
+        executor.setKeepAliveSeconds(60);
+        
+        // 设置拒绝策略：由调用者线程执行
+        executor.setRejectedExecutionHandler(new ThreadPoolExecutor.CallerRunsPolicy());
+        
+        executor.initialize();
+        logger.info("对比分析线程池初始化完成: coreSize={}, maxSize={}, queueCapacity={}", 
+            executor.getCorePoolSize(), executor.getMaxPoolSize(), 30);
+        return executor;
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/DataInitializer.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/DataInitializer.java
new file mode 100644
index 0000000..de14d16
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/DataInitializer.java
@@ -0,0 +1,225 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.config
+ * @ClassName: DataInitializer
+ * @Description: 应用启动时初始化数据
+ * @Author: 张志锋
+ * @Date: 2025-11-03
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.config;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.File;
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.PromptTemplate;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.FileRepository;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.PromptTemplateRepository;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.boot.CommandLineRunner;
+import org.springframework.stereotype.Component;
+
+import java.util.ArrayList;
+import java.util.Arrays;
+import java.util.List;
+
+/**
+ * 数据初始化器
+ */
+@Component
+public class DataInitializer implements CommandLineRunner {
+
+    private static final Logger logger = LoggerFactory.getLogger(DataInitializer.class);
+
+    @Autowired
+    private PromptTemplateRepository promptTemplateRepository;
+
+    @Autowired
+    private FileRepository fileRepository;
+
+    @Override
+    public void run(String... args) throws Exception {
+        logger.info("开始初始化系统数据...");
+
+        // 初始化提示词模板
+        initializePromptTemplates();
+
+        logger.info("系统数据初始化完成");
+    }
+
+    /**
+     * 初始化提示词模板
+     */
+    private void initializePromptTemplates() {
+        logger.info("检查是否需要初始化提示词模板...");
+
+        // 检查是否已有各种类型的模板
+        long qaCount = promptTemplateRepository.countByTemplateType("qa");
+        long summaryCount = promptTemplateRepository.countByTemplateType("summary");
+        long comparisonCount = promptTemplateRepository.countByTemplateType("comparison");
+
+        if (qaCount > 0 && summaryCount > 0 && comparisonCount > 0) {
+            logger.info("所有提示词模板已存在 (QA: {}, Summary: {}, Comparison: {}), 跳过初始化", qaCount, summaryCount, comparisonCount);
+            return;
+        }
+
+        logger.info("开始初始化提示词模板数据 (QA: {}, Summary: {}, Comparison: {})...", qaCount, summaryCount, comparisonCount);
+
+        List<PromptTemplate> templatesToAdd = new ArrayList<>();
+
+        // 如果没有QA模板，添加QA模板
+        if (qaCount == 0) {
+            templatesToAdd.addAll(Arrays.asList(
+            createTemplate("tender-qa-general-001", "通用评标问答", 0,
+                "你是一个专业的招标评审专家助手。请基于以下提供的招标文件内容，准确、客观地回答评审专家的问题。\n\n招标文件内容：\n{document_content}\n\n评审问题：{query}\n\n请遵循以下原则：\n1. 严格基于招标文件内容回答，不添加外部知识或个人观点\n2. 引用具体条款、页码或章节，便于核实\n3. 对于技术参数、商务条件要准确表述\n4. 如有多个相关条款，应进行归纳整理\n5. 对于文件未明确的内容，应明确说明\"招标文件中未明确规定\"\n6. 保持专业、客观、中立的态度\n\n评审意见：",
+                "通用评标问答提示词，适用于各类招标项目的评审问答"),
+
+            createTemplate("tender-qa-technical-002", "技术评审问答", 1,
+                "你是一个专业的技术评审专家。请基于以下技术规格书和相关技术文档，回答技术评审相关问题。\n\n技术文档内容：\n{document_content}\n\n技术问题：{query}\n\n技术评审要点：\n1. 严格按照技术规格书要求进行评审\n2. 重点关注技术参数、性能指标、质量标准\n3. 识别技术风险和潜在问题\n4. 评估技术方案的可行性和先进性\n5. 检查技术指标是否满足最低要求\n6. 对于技术偏离应明确指出影响程度\n\n技术评审意见：\n- 符合性分析：\n- 技术风险评估：\n- 建议意见：",
+                "技术评审专用提示词，重点关注技术参数、性能指标等技术评审内容"),
+
+            createTemplate("tender-qa-commercial-003", "商务评审问答", 2,
+                "你是一个专业的商务评审专家。请基于以下商务文件和报价文件，回答商务评审相关问题。\n\n商务文件内容：\n{document_content}\n\n商务问题：{query}\n\n商务评审要点：\n1. 严格按照招标文件商务要求进行评审\n2. 重点关注报价合理性、商务条件、付款方式\n3. 检查商务偏差及其影响\n4. 评估供应商的商务实力和信誉\n5. 审核商务条款的合规性和可行性\n6. 计算商务得分并给出评分理由\n\n商务评审意见：\n- 报价分析：\n- 商务条件评估：\n- 合规性检查：\n- 评分建议：",
+                "商务评审专用提示词，重点关注报价、商务条件、付款方式等商务评审内容"),
+
+            createTemplate("tender-qa-compliance-004", "合规性检查问答", 3,
+                "你是一个专业的招标合规性审查专家。请基于招标文件和相关法律法规，回答合规性相关问题。\n\n招标文件及法律法规内容：\n{document_content}\n\n合规性问题：{query}\n\n合规性审查要点：\n1. 检查投标文件是否符合招标文件要求\n2. 审核供应商资质和相关证书\n3. 验证投标保证金、履约保证金等保证措施\n4. 审查投标文件完整性和有效性\n5. 识别潜在的违规行为和风险\n6. 确认投标有效期和投标截止时间\n\n合规性审查意见：\n- 资格审查：\n- 文件完整性：\n- 保证措施：\n- 违规风险：\n- 审查结论：",
+                "合规性检查专用提示词，重点关注投标文件的合规性和有效性审查"),
+
+            createTemplate("tender-qa-comparison-005", "供应商对比问答", 4,
+                "你是一个专业的供应商对比分析专家。请基于多个供应商的投标文件，进行对比分析并回答相关问题。\n\n供应商投标文件内容：\n{document_content}\n\n对比分析问题：{query}\n\n供应商对比要点：\n1. 对比各供应商的技术方案差异\n2. 分析报价合理性和竞争力\n3. 评估供应商实力和服务能力\n4. 识别各供应商的优势和劣势\n5. 综合评价供应商的综合竞争力\n6. 提出供应商选择建议\n\n供应商对比分析：\n- 技术方案对比：\n- 商务条件对比：\n- 供应商实力评估：\n- 综合评分建议：\n- 推荐意见：",
+                "供应商对比分析专用提示词，用于多个供应商投标文件的对比评审")
+            ));
+        }
+
+        // 如果没有Summary模板，添加Summary模板
+        if (summaryCount == 0) {
+            templatesToAdd.add(
+                createSummaryTemplate("default-summary-template-001", "默认文档摘要提示词", 10,
+                    "章节标题：{title}\n\n请直接返回该章节的总结内容，不要任何开场白、结尾语或其他多余文字。不使用markdown格式。不说'好的，这是...'之类的话。只返回纯文本总结内容，300字以内，可以换行。总结要准确、客观、简洁，突出核心要点。",
+                    "默认的文档摘要提示词模板，用于指导AI如何对文档章节进行结构化摘要")
+            );
+        }
+
+        // 如果没有Comparison模板，添加Comparison模板
+        if (comparisonCount == 0) {
+            templatesToAdd.add(
+                createComparisonTemplate("default-comparison-template-001", "默认供应商对比提示词", 11,
+                    "请对以下供应商在{comparison_dimension}维度的表现进行对比分析。\n\n供应商信息：\n{supplier_info}\n\n请直接返回对比分析结果，不要任何开场白、结尾语或其他多余文字。不使用markdown格式。只返回纯文本分析内容，500字以内，可以换行。分析要求：客观、准确、突出差异和特点。",
+                    "默认的供应商对比分析提示词模板，用于指导AI如何进行供应商间的横向对比")
+            );
+        }
+
+        try {
+            if (!templatesToAdd.isEmpty()) {
+                promptTemplateRepository.saveAll(templatesToAdd);
+                logger.info("成功初始化 {} 个提示词模板", templatesToAdd.size());
+            } else {
+                logger.info("没有需要初始化的提示词模板");
+            }
+        } catch (Exception e) {
+            logger.error("初始化提示词模板失败", e);
+        }
+
+        // 初始化测试文件数据
+        initializeTestFiles();
+    }
+
+    /**
+     * 初始化测试文件数据
+     */
+    private void initializeTestFiles() {
+        logger.info("开始初始化测试文件数据...");
+
+        List<File> testFiles = Arrays.asList(
+            createTestFile("file-001", "技术方案.pdf", "project-001", "supplier-001", 2048576L),
+            createTestFile("file-002", "商务文件.pdf", "project-001", "supplier-002", 1536000L),
+            createTestFile("file-003", "报价清单.xlsx", "project-001", "supplier-001", 512000L),
+            createTestFile("file-004", "技术参数表.doc", "project-001", "supplier-002", 1024000L),
+            createTestFile("file-005", "合规性承诺书.pdf", "project-001", "supplier-003", 768000L)
+        );
+
+        try {
+            fileRepository.saveAll(testFiles);
+            logger.info("成功初始化 {} 个测试文件", testFiles.size());
+        } catch (Exception e) {
+            logger.error("初始化测试文件失败", e);
+        }
+    }
+
+    /**
+     * 创建QA提示词模板对象
+     */
+    private PromptTemplate createTemplate(String templateId, String templateName, int sortOrder,
+                                        String promptContent, String description) {
+        PromptTemplate template = new PromptTemplate();
+        template.setTemplateId(templateId);
+        template.setTemplateName(templateName);
+        template.setTemplateType("qa");
+        template.setPromptContent(promptContent);
+        template.setDescription(description);
+        template.setIsActive(true);
+        template.setSortOrder(sortOrder);
+        template.setCreatedBy("system");
+        template.setUpdatedBy("system");
+        return template;
+    }
+
+    /**
+     * 创建文档摘要提示词模板对象
+     */
+    private PromptTemplate createSummaryTemplate(String templateId, String templateName, int sortOrder,
+                                               String promptContent, String description) {
+        PromptTemplate template = new PromptTemplate();
+        template.setTemplateId(templateId);
+        template.setTemplateName(templateName);
+        template.setTemplateType("summary");
+        template.setPromptContent(promptContent);
+        template.setDescription(description);
+        template.setIsActive(true);
+        template.setSortOrder(sortOrder);
+        template.setCreatedBy("system");
+        template.setUpdatedBy("system");
+        return template;
+    }
+
+    /**
+     * 创建供应商对比分析提示词模板对象
+     */
+    private PromptTemplate createComparisonTemplate(String templateId, String templateName, int sortOrder,
+                                                  String promptContent, String description) {
+        PromptTemplate template = new PromptTemplate();
+        template.setTemplateId(templateId);
+        template.setTemplateName(templateName);
+        template.setTemplateType("comparison");
+        template.setPromptContent(promptContent);
+        template.setDescription(description);
+        template.setIsActive(true);
+        template.setSortOrder(sortOrder);
+        template.setCreatedBy("system");
+        template.setUpdatedBy("system");
+        return template;
+    }
+
+    /**
+     * 创建测试文件对象
+     */
+    private File createTestFile(String fileId, String fileName, String projectId, String supplierId, Long fileSize) {
+        File file = new File();
+        file.setFileId(fileId);
+        file.setFileName(fileName);
+        file.setProjectId(projectId);
+        file.setSupplierId(supplierId);
+        file.setFileSize(fileSize);
+        file.setFileExtension(fileName.substring(fileName.lastIndexOf('.') + 1));
+        file.setFilePath("/test/" + fileName); // 模拟路径
+        file.setFileMd5("test-md5-" + fileId); // 模拟MD5
+        file.setStatus("uploaded");
+        file.setIsVectorized(true); // 设为已向量化，方便测试
+        file.setVectorizationStatus("completed");
+        file.setVectorizationCount(1);
+        file.setDescription("测试文件 - " + fileName);
+        return file;
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/DataSourceConfig.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/DataSourceConfig.java
new file mode 100644
index 0000000..a44d44c
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/DataSourceConfig.java
@@ -0,0 +1,240 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.config
+ * @ClassName: DataSourceConfig
+ * @Description: 数据源配置类，配置MySQL和PostgreSQL双数据源，支持数据库自动创建
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.config;
+
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.boot.jdbc.DataSourceBuilder;
+import org.springframework.context.annotation.Bean;
+import org.springframework.context.annotation.Configuration;
+import org.springframework.context.annotation.Primary;
+import org.springframework.core.env.Environment;
+import org.springframework.data.jpa.repository.config.EnableJpaRepositories;
+import org.springframework.orm.jpa.JpaTransactionManager;
+import org.springframework.orm.jpa.LocalContainerEntityManagerFactoryBean;
+import org.springframework.orm.jpa.vendor.HibernateJpaVendorAdapter;
+import org.springframework.transaction.PlatformTransactionManager;
+import org.springframework.transaction.annotation.EnableTransactionManagement;
+
+import javax.sql.DataSource;
+import java.util.Properties;
+
+@Configuration
+@EnableTransactionManagement
+@EnableJpaRepositories(basePackages = "com.zhpb.gdyd_zhpb_zgf.repository")
+public class DataSourceConfig {
+
+    private static final Logger LOGGER = LoggerFactory.getLogger(DataSourceConfig.class);
+
+    private final Environment environment;
+
+    public DataSourceConfig(Environment environment) {
+        this.environment = environment;
+    }
+
+    @Bean
+    @Primary
+    public DataSource dataSource() {
+        ensurePostgreSQLDatabase();
+
+        DataSource dataSource = DataSourceBuilder.create()
+                .driverClassName(environment.getProperty("spring.datasource.driver-class-name"))
+                .url(environment.getProperty("spring.datasource.url"))
+                .username(environment.getProperty("spring.datasource.username"))
+                .password(environment.getProperty("spring.datasource.password"))
+                .build();
+
+        ensurePostgreSQLVectorExtension();
+        return dataSource;
+    }
+
+    @Bean
+    @Primary
+    public LocalContainerEntityManagerFactoryBean entityManagerFactory(DataSource dataSource) {
+        LocalContainerEntityManagerFactoryBean em = new LocalContainerEntityManagerFactoryBean();
+        em.setDataSource(dataSource);
+        em.setPackagesToScan("com.zhpb.gdyd_zhpb_zgf.entity");
+
+        HibernateJpaVendorAdapter vendorAdapter = new HibernateJpaVendorAdapter();
+        vendorAdapter.setShowSql(Boolean.parseBoolean(environment.getProperty("spring.jpa.show-sql", "false")));
+        em.setJpaVendorAdapter(vendorAdapter);
+
+        Properties properties = new Properties();
+        properties.setProperty("hibernate.hbm2ddl.auto",
+                environment.getProperty("spring.jpa.hibernate.ddl-auto", "update"));
+        properties.setProperty("hibernate.dialect",
+                environment.getProperty("spring.jpa.properties.hibernate.dialect",
+                        "org.hibernate.dialect.PostgreSQLDialect"));
+        properties.setProperty("hibernate.format_sql",
+                environment.getProperty("spring.jpa.properties.hibernate.format_sql", "false"));
+        properties.setProperty("hibernate.jdbc.time_zone",
+                environment.getProperty("spring.jpa.properties.hibernate.jdbc.time_zone", "Asia/Shanghai"));
+        em.setJpaProperties(properties);
+
+        return em;
+    }
+
+    @Bean
+    @Primary
+    public PlatformTransactionManager transactionManager(
+            LocalContainerEntityManagerFactoryBean entityManagerFactory) {
+        return new JpaTransactionManager(entityManagerFactory.getObject());
+    }
+
+    private void ensurePostgreSQLVectorExtension() {
+        try {
+            String url = environment.getProperty("spring.datasource.url");
+            String username = environment.getProperty("spring.datasource.username", "postgres");
+            String password = environment.getProperty("spring.datasource.password", "123456");
+
+            if (url == null) {
+                return;
+            }
+
+            try (java.sql.Connection conn = java.sql.DriverManager.getConnection(url, username, password);
+                 java.sql.Statement stmt = conn.createStatement()) {
+
+                stmt.execute("CREATE EXTENSION IF NOT EXISTS vector");
+
+                try (java.sql.ResultSet rs = stmt.executeQuery(
+                        "SELECT EXISTS (SELECT FROM information_schema.tables WHERE table_name = 't_pdf_vector')")) {
+                    if (rs.next() && !rs.getBoolean(1)) {
+                        stmt.execute("CREATE TABLE t_pdf_vector (" +
+                                "id BIGSERIAL PRIMARY KEY, " +
+                                "file_path VARCHAR(1000) NOT NULL, " +
+                                "file_name VARCHAR(500) NOT NULL, " +
+                                "content TEXT, " +
+                                "vector vector, " +
+                                "dimension INTEGER NOT NULL, " +
+                                "file_id VARCHAR(64), " +
+                                "project_id VARCHAR(100), " +
+                                "supplier_id VARCHAR(64), " +
+                                "file_type_id BIGINT, " +
+                                "create_time TIMESTAMP(6) NOT NULL DEFAULT NOW(), " +
+                                "update_time TIMESTAMP(6) DEFAULT NOW()" +
+                                ")");
+                        LOGGER.info("PostgreSQL t_pdf_vector表已创建");
+                    } else {
+                        addMissingColumnsIfNeeded(stmt);
+                    }
+                }
+
+                try (java.sql.ResultSet rs = stmt.executeQuery(
+                        "SELECT EXISTS (SELECT FROM information_schema.tables WHERE table_name = 't_product')")) {
+                    if (rs.next() && !rs.getBoolean(1)) {
+                        stmt.execute("CREATE TABLE t_product (" +
+                                "id BIGSERIAL PRIMARY KEY, " +
+                                "name VARCHAR(255), " +
+                                "price DOUBLE PRECISION" +
+                                ")");
+                        LOGGER.info("PostgreSQL t_product表已创建");
+                    }
+                }
+
+                LOGGER.debug("PostgreSQL pgvector扩展和表初始化完成");
+            }
+        } catch (Exception e) {
+            LOGGER.warn("创建PostgreSQL pgvector扩展或表失败: {}", e.getMessage());
+        }
+    }
+
+    private void addMissingColumnsIfNeeded(java.sql.Statement stmt) {
+        try {
+            try (java.sql.ResultSet rs = stmt.executeQuery(
+                    "SELECT EXISTS (SELECT 1 FROM information_schema.columns " +
+                            "WHERE table_schema = 'public' AND table_name = 't_pdf_vector' AND column_name = 'file_id')")) {
+                if (rs.next() && !rs.getBoolean(1)) {
+                    stmt.execute("ALTER TABLE t_pdf_vector ADD COLUMN file_id VARCHAR(64)");
+                    LOGGER.info("已添加file_id列到t_pdf_vector表");
+                }
+            }
+
+            try (java.sql.ResultSet rs = stmt.executeQuery(
+                    "SELECT EXISTS (SELECT 1 FROM information_schema.columns " +
+                            "WHERE table_schema = 'public' AND table_name = 't_pdf_vector' AND column_name = 'supplier_id')")) {
+                if (rs.next() && !rs.getBoolean(1)) {
+                    stmt.execute("ALTER TABLE t_pdf_vector ADD COLUMN supplier_id VARCHAR(64)");
+                    LOGGER.info("已添加supplier_id列到t_pdf_vector表");
+                }
+            }
+
+            try (java.sql.ResultSet rs = stmt.executeQuery(
+                    "SELECT EXISTS (SELECT 1 FROM information_schema.columns " +
+                            "WHERE table_schema = 'public' AND table_name = 't_pdf_vector' AND column_name = 'file_type_id')")) {
+                if (rs.next() && !rs.getBoolean(1)) {
+                    stmt.execute("ALTER TABLE t_pdf_vector ADD COLUMN file_type_id BIGINT");
+                    LOGGER.info("已添加file_type_id列到t_pdf_vector表");
+                }
+            }
+
+            LOGGER.info("t_pdf_vector表字段检查完成");
+        } catch (Exception e) {
+            LOGGER.error("添加缺失列失败: {}", e.getMessage(), e);
+        }
+    }
+
+    private void ensurePostgreSQLDatabase() {
+        try {
+            String url = environment.getProperty("spring.datasource.url");
+            String username = environment.getProperty("spring.datasource.username", "postgres");
+            String password = environment.getProperty("spring.datasource.password");
+
+            if (url == null || password == null) {
+                return;
+            }
+
+            String databaseName = extractDatabaseName(url);
+            if (databaseName == null) {
+                return;
+            }
+
+            String defaultUrl = url.replace("/" + databaseName, "/postgres");
+
+            try (java.sql.Connection conn = java.sql.DriverManager.getConnection(defaultUrl, username, password)) {
+                String checkSql = "SELECT 1 FROM pg_database WHERE datname = ?";
+                try (java.sql.PreparedStatement stmt = conn.prepareStatement(checkSql)) {
+                    stmt.setString(1, databaseName);
+                    try (java.sql.ResultSet rs = stmt.executeQuery()) {
+                        if (!rs.next()) {
+                            conn.setAutoCommit(true);
+                            String createSql = "CREATE DATABASE \"" + databaseName + "\"";
+                            try (java.sql.Statement createStmt = conn.createStatement()) {
+                                createStmt.execute(createSql);
+                            }
+                        }
+                    }
+                }
+            }
+        } catch (Exception e) {
+            LOGGER.warn("自动创建PostgreSQL数据库失败: {}", e.getMessage());
+        }
+    }
+
+    private String extractDatabaseName(String url) {
+        if (url == null) {
+            return null;
+        }
+        try {
+            int lastSlash = url.lastIndexOf('/');
+            if (lastSlash > 0 && lastSlash < url.length() - 1) {
+                String dbPart = url.substring(lastSlash + 1);
+                int questionMark = dbPart.indexOf('?');
+                if (questionMark > 0) {
+                    return dbPart.substring(0, questionMark);
+                }
+                return dbPart;
+            }
+        } catch (Exception ignored) {
+        }
+        return null;
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/DatabaseInitializer.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/DatabaseInitializer.java
new file mode 100644
index 0000000..dbccc81
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/DatabaseInitializer.java
@@ -0,0 +1,120 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.config
+ * @ClassName: DatabaseInitializer
+ * @Description: 数据库初始化器，用于在应用启动时创建数据库（已迁移到DataSourceConfig）
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.config;
+
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.beans.factory.annotation.Value;
+import org.springframework.boot.autoconfigure.condition.ConditionalOnProperty;
+import org.springframework.stereotype.Component;
+
+import jakarta.annotation.PostConstruct;
+import java.sql.Connection;
+import java.sql.DriverManager;
+import java.sql.PreparedStatement;
+import java.sql.ResultSet;
+import java.sql.Statement;
+
+/**
+ * 数据库初始化配置类
+ * 用于自动创建不存在的数据库
+ * 在数据源Bean创建之前执行初始化
+ */
+@Component
+@ConditionalOnProperty(name = "spring.datasource.url")
+public class DatabaseInitializer {
+
+    private static final Logger logger = LoggerFactory.getLogger(DatabaseInitializer.class);
+
+    @Value("${spring.datasource.url}")
+    private String pgsqlUrl;
+
+    @Value("${spring.datasource.username:postgres}")
+    private String pgsqlUsername;
+
+    @Value("${spring.datasource.password}")
+    private String pgsqlPassword;
+
+    /**
+     * 应用启动后自动初始化数据库
+     * 使用@PostConstruct确保在Bean初始化时立即执行
+     */
+    @PostConstruct
+    public void initializeDatabase() {
+        try {
+            // 解析数据库名称
+            String databaseName = extractDatabaseName(pgsqlUrl);
+            if (databaseName == null) {
+                logger.warn("无法从PostgreSQL URL中解析数据库名称，跳过自动创建");
+                return;
+            }
+
+            // 连接到默认的postgres数据库
+            String defaultUrl = pgsqlUrl.replace("/" + databaseName, "/postgres");
+            
+            logger.info("正在检查PostgreSQL数据库 '{}' 是否存在...", databaseName);
+            
+            try (Connection conn = DriverManager.getConnection(defaultUrl, pgsqlUsername, pgsqlPassword)) {
+                // 检查数据库是否存在
+                String checkSql = "SELECT 1 FROM pg_database WHERE datname = ?";
+                try (PreparedStatement stmt = conn.prepareStatement(checkSql)) {
+                    stmt.setString(1, databaseName);
+                    try (ResultSet rs = stmt.executeQuery()) {
+                        if (!rs.next()) {
+                            // 数据库不存在，创建它
+                            logger.info("数据库 '{}' 不存在，正在创建...", databaseName);
+                            
+                            // PostgreSQL不允许在事务中执行CREATE DATABASE
+                            // CREATE DATABASE不能使用PreparedStatement，需要使用Statement
+                            conn.setAutoCommit(true);
+                            String createSql = "CREATE DATABASE \"" + databaseName + "\"";
+                            try (Statement createStmt = conn.createStatement()) {
+                                createStmt.execute(createSql);
+                                logger.info("数据库 '{}' 创建成功！", databaseName);
+                            }
+                        } else {
+                            logger.info("数据库 '{}' 已存在，跳过创建", databaseName);
+                        }
+                    }
+                }
+            }
+        } catch (Exception e) {
+            logger.warn("自动创建PostgreSQL数据库失败: {}", e.getMessage());
+            logger.debug("详细信息", e);
+        }
+    }
+
+    /**
+     * 从JDBC URL中提取数据库名称
+     * 
+     * @param url JDBC URL，例如：jdbc:postgresql://localhost:5432/database_name
+     * @return 数据库名称
+     */
+    private String extractDatabaseName(String url) {
+        try {
+            // 格式：jdbc:postgresql://host:port/database_name
+            int lastSlash = url.lastIndexOf('/');
+            if (lastSlash > 0 && lastSlash < url.length() - 1) {
+                String dbPart = url.substring(lastSlash + 1);
+                // 移除可能的查询参数
+                int questionMark = dbPart.indexOf('?');
+                if (questionMark > 0) {
+                    return dbPart.substring(0, questionMark);
+                }
+                return dbPart;
+            }
+        } catch (Exception e) {
+            logger.error("解析数据库名称失败", e);
+        }
+        return null;
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/DatabaseLogAppender.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/DatabaseLogAppender.java
new file mode 100644
index 0000000..1786ade
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/DatabaseLogAppender.java
@@ -0,0 +1,177 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.config
+ * @ClassName: DatabaseLogAppender
+ * @Description: Logback自定义Appender，将所有日志输出保存到数据库
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.config;
+
+import ch.qos.logback.classic.spi.ILoggingEvent;
+import ch.qos.logback.core.UnsynchronizedAppenderBase;
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.SysLog;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.SysLogRepository;
+import org.springframework.context.ApplicationContext;
+
+import java.time.LocalDateTime;
+
+/**
+ * 数据库日志Appender
+ */
+public class DatabaseLogAppender extends UnsynchronizedAppenderBase<ILoggingEvent> {
+
+    private static ApplicationContext applicationContext;
+    private static SysLogRepository sysLogRepository;
+
+    public static void setApplicationContext(ApplicationContext context) {
+        DatabaseLogAppender.applicationContext = context;
+        if (context != null) {
+            try {
+                sysLogRepository = context.getBean(SysLogRepository.class);
+            } catch (Exception e) {
+                // Bean未准备好，稍后重试
+            }
+        }
+    }
+
+    @Override
+    protected void append(ILoggingEvent event) {
+        // 过滤掉不需要记录的日志
+        if (shouldSkip(event)) {
+            return;
+        }
+        
+        if (sysLogRepository == null && applicationContext != null) {
+            try {
+                sysLogRepository = applicationContext.getBean(SysLogRepository.class);
+            } catch (Exception e) {
+                // Bean未准备好，忽略
+                return;
+            }
+        }
+
+        if (sysLogRepository == null) {
+            return;
+        }
+
+        try {
+            SysLog sysLog = new SysLog();
+            sysLog.setLogLevel(event.getLevel().toString());
+            sysLog.setLoggerName(event.getLoggerName());
+            
+            // 格式化日志消息
+            String formattedMessage = event.getFormattedMessage();
+            // 将多行内容合并为一行
+            String singleLineContent = formattedMessage.replace("\n", " ").replace("\r", " ").trim();
+            sysLog.setLogContent(singleLineContent);
+            sysLog.setFullMessage(getFullMessage(event).replace("\n", " ").replace("\r", " "));
+            
+            // 异常堆栈（保留多行格式）
+            if (event.getThrowableProxy() != null) {
+                sysLog.setExceptionStack(getStackTrace(event));
+            }
+            
+            sysLog.setThreadName(event.getThreadName());
+            sysLog.setCreateTime(LocalDateTime.now());
+            sysLog.setApplicationName("gdyd_zhpb_zgf");
+            
+            // 获取调用位置信息
+            if (event.getCallerData() != null && event.getCallerData().length > 0) {
+                StackTraceElement caller = event.getCallerData()[0];
+                sysLog.setClassName(caller.getClassName());
+                sysLog.setMethodName(caller.getMethodName());
+                sysLog.setLineNumber(caller.getLineNumber());
+            }
+
+            // 异步保存（使用新线程，避免阻塞）
+            new Thread(() -> {
+                try {
+                    sysLogRepository.save(sysLog);
+                } catch (Exception e) {
+                    // 忽略保存失败，避免循环日志
+                }
+            }).start();
+        } catch (Exception e) {
+            // 忽略异常，避免影响正常日志输出
+        }
+    }
+
+    /**
+     * 判断是否应该跳过该日志
+     */
+    private boolean shouldSkip(ILoggingEvent event) {
+        String loggerName = event.getLoggerName().toLowerCase();
+        String message = event.getFormattedMessage().toLowerCase();
+        
+        // 跳过SQL相关日志
+        if (loggerName.contains("org.hibernate.sql") || 
+            loggerName.contains("org.hibernate.type") ||
+            message.contains("hibernate:") ||
+            message.contains("select ") ||
+            message.contains("insert ") ||
+            message.contains("update ") ||
+            message.contains("delete ") ||
+            message.contains("create table") ||
+            message.contains("alter table")) {
+            return true;
+        }
+        
+        // 跳过Spring Boot启动相关日志
+        if (loggerName.contains("org.springframework.boot") ||
+            loggerName.contains("org.springframework.context") ||
+            loggerName.contains("org.springframework.beans") ||
+            loggerName.contains("org.hibernate.boot") ||
+            loggerName.contains("org.hibernate.jpa") ||
+            loggerName.contains("org.springframework.orm") ||
+            message.contains("started application") ||
+            message.contains("beanfactory") ||
+            message.contains("obtaining singleton") ||
+            message.contains("autowired annotation") ||
+            message.contains("initialized jpa")) {
+            return true;
+        }
+        
+        // 跳过Hibernate初始化日志
+        if ((loggerName.contains("hibernate") || loggerName.contains("org.hibernate")) &&
+            (message.contains("dialect") || 
+             message.contains("sessionfactory") ||
+             message.contains("entitymanagerfactory") ||
+             message.contains("hibernate is in classpath"))) {
+            return true;
+        }
+        
+        return false;
+    }
+
+    private String getFullMessage(ILoggingEvent event) {
+        StringBuilder sb = new StringBuilder();
+        sb.append("[").append(event.getLevel()).append("] ");
+        sb.append("[").append(event.getThreadName()).append("] ");
+        sb.append(event.getLoggerName()).append(" - ");
+        sb.append(event.getFormattedMessage());
+        return sb.toString();
+    }
+
+    private String getStackTrace(ILoggingEvent event) {
+        if (event.getThrowableProxy() == null) {
+            return null;
+        }
+        
+        StringBuilder sb = new StringBuilder();
+        sb.append(event.getThrowableProxy().getClassName()).append(": ")
+          .append(event.getThrowableProxy().getMessage()).append("\n");
+        
+        if (event.getThrowableProxy().getStackTraceElementProxyArray() != null) {
+            for (int i = 0; i < event.getThrowableProxy().getStackTraceElementProxyArray().length; i++) {
+                sb.append("  at ").append(event.getThrowableProxy().getStackTraceElementProxyArray()[i])
+                  .append("\n");
+            }
+        }
+        
+        return sb.toString();
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/DatabaseMigration.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/DatabaseMigration.java
new file mode 100644
index 0000000..0dde955
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/DatabaseMigration.java
@@ -0,0 +1,143 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.config
+ * @ClassName: DatabaseMigration
+ * @Description: 数据库迁移配置 - 启动时自动添加缺失的字段
+ * @Author: 张志锋
+ * @Date: 2025-12-02
+ * @Version: 1.0
+ */
+package com.zhpb.gdyd_zhpb_zgf.config;
+
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.boot.ApplicationArguments;
+import org.springframework.boot.ApplicationRunner;
+import org.springframework.jdbc.core.JdbcTemplate;
+import org.springframework.stereotype.Component;
+
+/**
+ * 数据库迁移 - 自动添加缺失的字段
+ */
+@Component
+public class DatabaseMigration implements ApplicationRunner {
+
+    private static final Logger logger = LoggerFactory.getLogger(DatabaseMigration.class);
+
+    @Autowired
+    private JdbcTemplate jdbcTemplate;
+
+    @Override
+    public void run(ApplicationArguments args) {
+        logger.info("开始执行数据库迁移检查...");
+        
+        try {
+            // 添加目录摘取相关字段
+            addCatalogExtractionFields();
+            
+            logger.info("数据库迁移检查完成");
+        } catch (Exception e) {
+            logger.error("数据库迁移失败", e);
+            // 不抛出异常，避免影响应用启动
+        }
+    }
+
+    /**
+     * 添加目录摘取相关字段
+     */
+    private void addCatalogExtractionFields() {
+        try {
+            // 检查并添加 catalog_extraction_status 字段
+            if (!columnExists("t_file", "catalog_extraction_status")) {
+                logger.info("添加字段: t_file.catalog_extraction_status");
+                jdbcTemplate.execute(
+                    "ALTER TABLE t_file ADD COLUMN catalog_extraction_status VARCHAR(32) DEFAULT 'processing'"
+                );
+                logger.info("字段添加成功: catalog_extraction_status");
+            } else {
+                logger.info("字段已存在: catalog_extraction_status");
+            }
+
+            // 检查并添加 catalog_extraction_retry_count 字段
+            if (!columnExists("t_file", "catalog_extraction_retry_count")) {
+                logger.info("添加字段: t_file.catalog_extraction_retry_count");
+                jdbcTemplate.execute(
+                    "ALTER TABLE t_file ADD COLUMN catalog_extraction_retry_count INT DEFAULT 0 NOT NULL"
+                );
+                logger.info("字段添加成功: catalog_extraction_retry_count");
+            } else {
+                logger.info("字段已存在: catalog_extraction_retry_count");
+            }
+
+            // 检查并添加 catalog_extraction_time 字段
+            if (!columnExists("t_file", "catalog_extraction_time")) {
+                logger.info("添加字段: t_file.catalog_extraction_time");
+                jdbcTemplate.execute(
+                    "ALTER TABLE t_file ADD COLUMN catalog_extraction_time TIMESTAMP"
+                );
+                logger.info("字段添加成功: catalog_extraction_time");
+            } else {
+                logger.info("字段已存在: catalog_extraction_time");
+            }
+
+            // 添加字段注释
+            try {
+                jdbcTemplate.execute(
+                    "COMMENT ON COLUMN t_file.catalog_extraction_status IS '目录摘取状态：not_started-未开始, processing-处理中, completed-已完成'"
+                );
+                jdbcTemplate.execute(
+                    "COMMENT ON COLUMN t_file.catalog_extraction_retry_count IS '目录摘取重试次数（最多重试3次）'"
+                );
+                jdbcTemplate.execute(
+                    "COMMENT ON COLUMN t_file.catalog_extraction_time IS '目录摘取完成时间'"
+                );
+                logger.info("字段注释添加成功");
+            } catch (Exception e) {
+                logger.warn("添加字段注释失败（可忽略）: {}", e.getMessage());
+            }
+
+            // 为已有数据设置初始状态
+            try {
+                int updatedRows = jdbcTemplate.update(
+                    "UPDATE t_file f SET catalog_extraction_status = 'completed', " +
+                    "catalog_extraction_time = f.update_time " +
+                    "WHERE EXISTS (SELECT 1 FROM t_file_directory fd WHERE fd.file_id = f.file_id) " +
+                    "AND (catalog_extraction_status IS NULL OR catalog_extraction_status = 'not_started')"
+                );
+                logger.info("更新已有数据状态，影响行数: {}", updatedRows);
+            } catch (Exception e) {
+                logger.warn("更新已有数据状态失败（可忽略）: {}", e.getMessage());
+            }
+
+            // 创建索引
+            try {
+                jdbcTemplate.execute(
+                    "CREATE INDEX IF NOT EXISTS idx_file_catalog_status ON t_file(catalog_extraction_status)"
+                );
+                logger.info("索引创建成功: idx_file_catalog_status");
+            } catch (Exception e) {
+                logger.warn("创建索引失败（可忽略）: {}", e.getMessage());
+            }
+
+        } catch (Exception e) {
+            logger.error("添加目录摘取字段失败", e);
+            throw new RuntimeException("数据库迁移失败: " + e.getMessage(), e);
+        }
+    }
+
+    /**
+     * 检查字段是否存在
+     */
+    private boolean columnExists(String tableName, String columnName) {
+        try {
+            String sql = "SELECT COUNT(*) FROM information_schema.columns " +
+                        "WHERE table_name = ? AND column_name = ?";
+            Integer count = jdbcTemplate.queryForObject(sql, Integer.class, tableName, columnName);
+            return count != null && count > 0;
+        } catch (Exception e) {
+            logger.warn("检查字段是否存在时出错: {}", e.getMessage());
+            return false;
+        }
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/LogConfig.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/LogConfig.java
new file mode 100644
index 0000000..dfb27b5
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/LogConfig.java
@@ -0,0 +1,34 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.config
+ * @ClassName: LogConfig
+ * @Description: 日志配置类，初始化数据库日志Appender
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.config;
+
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.context.ApplicationContext;
+import org.springframework.context.ApplicationListener;
+import org.springframework.context.event.ContextRefreshedEvent;
+import org.springframework.stereotype.Component;
+
+/**
+ * 日志配置类
+ */
+@Component
+public class LogConfig implements ApplicationListener<ContextRefreshedEvent> {
+
+    @Autowired
+    private ApplicationContext applicationContext;
+
+    @Override
+    public void onApplicationEvent(ContextRefreshedEvent event) {
+        // 初始化数据库日志Appender（通过logback-spring.xml配置的Appender会自动调用setApplicationContext）
+        DatabaseLogAppender.setApplicationContext(applicationContext);
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/OpenApiConfig.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/OpenApiConfig.java
new file mode 100644
index 0000000..ef389fd
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/OpenApiConfig.java
@@ -0,0 +1,53 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.config
+ * @ClassName: OpenApiConfig
+ * @Description: Swagger/OpenAPI配置类，配置API文档的标题、描述和分组
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.config;
+
+import io.swagger.v3.oas.models.OpenAPI;
+import io.swagger.v3.oas.models.info.Contact;
+import io.swagger.v3.oas.models.info.Info;
+import io.swagger.v3.oas.models.info.License;
+import org.springdoc.core.models.GroupedOpenApi;
+import org.springframework.context.annotation.Bean;
+import org.springframework.context.annotation.Configuration;
+
+/**
+ * Swagger/OpenAPI 配置类
+ */
+@Configuration
+public class OpenApiConfig {
+
+    @Bean
+    public OpenAPI customOpenAPI() {
+        return new OpenAPI()
+                .info(new Info()
+                        .title("智慧评标-主观分助手 API 文档")
+                        .version("1.0.0")
+                        .description("智慧评标-主观分助手接口文档，支持单点登录（SSO）通过ES接口验证ticket并生成JWT Token")
+                        .contact(new Contact()
+                                .name("开发团队")
+                                .email("dev@example.com"))
+                        .license(new License()
+                                .name("Apache 2.0")
+                                .url("https://www.apache.org/licenses/LICENSE-2.0.html")));
+    }
+
+    /**
+     * 配置API分组
+     */
+    @Bean
+    public GroupedOpenApi publicApi() {
+        return GroupedOpenApi.builder()
+                .group("default")
+                .pathsToMatch("/api/**")
+                .build();
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/PostgreSQLHealthIndicator.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/PostgreSQLHealthIndicator.java
new file mode 100644
index 0000000..2006aa5
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/PostgreSQLHealthIndicator.java
@@ -0,0 +1,66 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.config
+ * @ClassName: PostgreSQLHealthIndicator
+ * @Description: PostgreSQL健康检查指示器，用于Actuator健康检查
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.config;
+
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.boot.actuate.health.Health;
+import org.springframework.boot.actuate.health.HealthIndicator;
+import org.springframework.stereotype.Component;
+
+import javax.sql.DataSource;
+import java.sql.Connection;
+import java.sql.SQLException;
+
+/**
+ * PostgreSQL 数据库健康检查指示器
+ */
+@Component("pgsqlHealthIndicator")
+public class PostgreSQLHealthIndicator implements HealthIndicator {
+
+    @Autowired
+    private DataSource dataSource;
+
+    @Override
+    public Health health() {
+        try (Connection connection = dataSource.getConnection()) {
+            if (connection.isValid(1)) {
+                // 可以执行一个简单的查询来验证连接
+                try (java.sql.Statement stmt = connection.createStatement();
+                     java.sql.ResultSet rs = stmt.executeQuery("SELECT version()")) {
+                    if (rs.next()) {
+                        String version = rs.getString(1);
+                        return Health.up()
+                                .withDetail("database", "PostgreSQL")
+                                .withDetail("status", "连接正常")
+                                .withDetail("version", version)
+                                .build();
+                    }
+                }
+                return Health.up()
+                        .withDetail("database", "PostgreSQL")
+                        .withDetail("status", "连接正常")
+                        .build();
+            } else {
+                return Health.down()
+                        .withDetail("database", "PostgreSQL")
+                        .withDetail("status", "连接无效")
+                        .build();
+            }
+        } catch (SQLException e) {
+            return Health.down()
+                    .withDetail("database", "PostgreSQL")
+                    .withDetail("status", "连接失败")
+                    .withDetail("error", e.getMessage())
+                    .build();
+        }
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/PostgreSQLVectorInitializer.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/PostgreSQLVectorInitializer.java
new file mode 100644
index 0000000..08c66bb
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/PostgreSQLVectorInitializer.java
@@ -0,0 +1,59 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.config
+ * @ClassName: PostgreSQLVectorInitializer
+ * @Description: PostgreSQL向量扩展初始化器，确保pgvector扩展已安装
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.config;
+
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.beans.factory.annotation.Value;
+import org.springframework.boot.autoconfigure.condition.ConditionalOnProperty;
+import org.springframework.stereotype.Component;
+
+import jakarta.annotation.PostConstruct;
+import java.sql.Connection;
+import java.sql.Statement;
+
+/**
+ * PostgreSQL向量扩展初始化器
+ */
+@Component
+@ConditionalOnProperty(name = "vectorize.enabled", havingValue = "true", matchIfMissing = true)
+public class PostgreSQLVectorInitializer {
+
+    private static final Logger logger = LoggerFactory.getLogger(PostgreSQLVectorInitializer.class);
+
+    @Value("${spring.datasource.url:jdbc:postgresql://127.0.0.1:5432/gdyd_zhpb_zgf}")
+    private String pgsqlUrl;
+
+    @Value("${spring.datasource.username:postgres}")
+    private String pgsqlUsername;
+
+    @Value("${spring.datasource.password:123456}")
+    private String pgsqlPassword;
+
+    @PostConstruct
+    public void init() {
+        try {
+            // 连接到目标数据库并创建扩展
+            try (Connection conn = java.sql.DriverManager.getConnection(pgsqlUrl, pgsqlUsername, pgsqlPassword)) {
+                try (Statement stmt = conn.createStatement()) {
+                    // 创建pgvector扩展
+                    stmt.execute("CREATE EXTENSION IF NOT EXISTS vector");
+                    logger.info("PostgreSQL pgvector扩展初始化成功");
+                }
+            }
+        } catch (Exception e) {
+            logger.warn("PostgreSQL pgvector扩展初始化失败，可能未安装pgvector扩展: {}", e.getMessage());
+            logger.warn("请手动在PostgreSQL中执行: CREATE EXTENSION IF NOT EXISTS vector;");
+            logger.warn("如果pgvector未安装，请参考: https://github.com/pgvector/pgvector");
+        }
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/RedisConfig.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/RedisConfig.java
new file mode 100644
index 0000000..9400893
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/RedisConfig.java
@@ -0,0 +1,43 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.config
+ * @ClassName: RedisConfig
+ * @Description: Redis配置类，配置Redis连接和序列化方式
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.config;
+
+import org.springframework.context.annotation.Bean;
+import org.springframework.context.annotation.Configuration;
+import org.springframework.data.redis.connection.RedisConnectionFactory;
+import org.springframework.data.redis.core.RedisTemplate;
+import org.springframework.data.redis.serializer.GenericJackson2JsonRedisSerializer;
+import org.springframework.data.redis.serializer.StringRedisSerializer;
+
+@Configuration
+public class RedisConfig {
+
+    /**
+     * Redis Template 配置
+     */
+    @Bean
+    public RedisTemplate<String, Object> redisTemplate(RedisConnectionFactory connectionFactory) {
+        RedisTemplate<String, Object> template = new RedisTemplate<>();
+        template.setConnectionFactory(connectionFactory);
+
+        // 设置 Key 的序列化方式
+        template.setKeySerializer(new StringRedisSerializer());
+        template.setHashKeySerializer(new StringRedisSerializer());
+
+        // 设置 Value 的序列化方式
+        template.setValueSerializer(new GenericJackson2JsonRedisSerializer());
+        template.setHashValueSerializer(new GenericJackson2JsonRedisSerializer());
+
+        template.afterPropertiesSet();
+        return template;
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/RequestResponseWrapperConfig.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/RequestResponseWrapperConfig.java
new file mode 100644
index 0000000..47d6a10
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/RequestResponseWrapperConfig.java
@@ -0,0 +1,68 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.config
+ * @ClassName: RequestResponseWrapperConfig
+ * @Description: 请求响应包装器配置类，包装HttpServletRequest和HttpServletResponse以支持内容缓存
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.config;
+
+import jakarta.servlet.Filter;
+import jakarta.servlet.FilterChain;
+import jakarta.servlet.FilterConfig;
+import jakarta.servlet.ServletException;
+import jakarta.servlet.ServletRequest;
+import jakarta.servlet.ServletResponse;
+import jakarta.servlet.http.HttpServletRequest;
+import jakarta.servlet.http.HttpServletResponse;
+import org.springframework.stereotype.Component;
+import org.springframework.web.util.ContentCachingRequestWrapper;
+import org.springframework.web.util.ContentCachingResponseWrapper;
+
+import java.io.IOException;
+
+/**
+ * 请求和响应包装器配置
+ * 用于在拦截器中读取请求体和响应体
+ */
+@Component
+public class RequestResponseWrapperConfig implements Filter {
+
+    @Override
+    public void init(FilterConfig filterConfig) throws ServletException {
+        // 不需要初始化
+    }
+
+    @Override
+    public void doFilter(ServletRequest request, ServletResponse response, FilterChain chain)
+            throws IOException, ServletException {
+        HttpServletRequest httpRequest = (HttpServletRequest) request;
+        HttpServletResponse httpResponse = (HttpServletResponse) response;
+
+        String uri = httpRequest.getRequestURI();
+        if (uri.contains("/api/v1/files/stream/pdf") || uri.contains("/api/v1/files/preview/pdf")) {
+            chain.doFilter(request, response);
+            return;
+        }
+
+        // 包装请求和响应，以便可以多次读取内容
+        ContentCachingRequestWrapper wrappedRequest = new ContentCachingRequestWrapper(httpRequest);
+        ContentCachingResponseWrapper wrappedResponse = new ContentCachingResponseWrapper(httpResponse);
+
+        try {
+            chain.doFilter(wrappedRequest, wrappedResponse);
+        } finally {
+            // 确保响应内容被写入
+            wrappedResponse.copyBodyToResponse();
+        }
+    }
+
+    @Override
+    public void destroy() {
+        // 不需要清理
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/SystemOutInterceptor.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/SystemOutInterceptor.java
new file mode 100644
index 0000000..3b72b80
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/SystemOutInterceptor.java
@@ -0,0 +1,193 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.config
+ * @ClassName: SystemOutInterceptor
+ * @Description: 拦截System.out.println输出，保存到数据库
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.config;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.SysLog;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.SysLogRepository;
+import jakarta.annotation.PostConstruct;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.beans.factory.annotation.Value;
+import org.springframework.stereotype.Component;
+
+import java.io.ByteArrayOutputStream;
+import java.io.PrintStream;
+import java.nio.charset.Charset;
+import java.time.LocalDateTime;
+
+/**
+ * System.out拦截器
+ */
+@Component
+public class SystemOutInterceptor {
+
+    private static final Logger logger = LoggerFactory.getLogger(SystemOutInterceptor.class);
+
+    @Autowired
+    private SysLogRepository sysLogRepository;
+
+    @Value("${spring.application.name:gdyd_zhpb_zgf}")
+    private String applicationName;
+
+    @PostConstruct
+    public void init() {
+        // 创建拦截输出流
+        try {
+            PrintStream originalOut = System.out;
+            PrintStream originalErr = System.err;
+            
+            System.setOut(new PrintStream(new DatabasePrintStream(originalOut, "OUT", false), true, "UTF-8"));
+            System.setErr(new PrintStream(new DatabasePrintStream(originalErr, "ERR", true), true, "UTF-8"));
+            logger.info("System.out和System.err拦截器已启动");
+        } catch (Exception e) {
+            logger.error("启动System.out拦截器失败", e);
+        }
+    }
+
+    /**
+     * 自定义打印流，同时输出到原流和数据库
+     */
+    private class DatabasePrintStream extends ByteArrayOutputStream {
+        private final PrintStream original;
+        private final String type;
+        private final boolean isError;
+
+        public DatabasePrintStream(PrintStream original, String type, boolean isError) {
+            this.original = original;
+            this.type = type;
+            this.isError = isError;
+        }
+
+        @Override
+        public synchronized void write(byte[] b, int off, int len) {
+            super.write(b, off, len);
+            original.write(b, off, len);
+        }
+
+        @Override
+        public synchronized void write(int b) {
+            super.write(b);
+            original.write(b);
+        }
+
+        @Override
+        public synchronized void flush() {
+            try {
+                super.flush();
+                original.flush();
+                saveToDatabase();
+            } catch (Exception e) {
+                // 忽略异常，确保不影响正常输出
+            }
+        }
+
+        private void saveToDatabase() {
+            try {
+                String content = new String(toByteArray(), Charset.forName("UTF-8"));
+                if (content != null && !content.trim().isEmpty()) {
+                    // 清除缓冲区
+                    reset();
+
+                    // 过滤掉不需要记录的日志
+                    if (shouldSkip(content)) {
+                        return;
+                    }
+
+                    // 将多行内容合并为一行（替换换行符为空格）
+                    String singleLine = content.replace("\n", " ").replace("\r", " ").trim();
+                    
+                    if (singleLine.isEmpty()) {
+                        return;
+                    }
+
+                    SysLog sysLog = new SysLog();
+                    sysLog.setLogLevel(isError ? "ERROR" : "OUT");
+                    sysLog.setLoggerName("SYSTEM_" + type);
+                    sysLog.setLogContent(singleLine);
+                    sysLog.setFullMessage("[" + type + "] " + singleLine);
+                    sysLog.setThreadName(Thread.currentThread().getName());
+                    sysLog.setCreateTime(LocalDateTime.now());
+                    sysLog.setApplicationName(applicationName);
+
+                    // 异步保存
+                    final SysLog finalLog = sysLog;
+                    new Thread(() -> {
+                        try {
+                            sysLogRepository.save(finalLog);
+                        } catch (Exception e) {
+                            // 忽略保存失败，避免循环
+                        }
+                    }).start();
+                }
+            } catch (Exception e) {
+                // 忽略异常，确保不影响正常输出
+            }
+        }
+
+        /**
+         * 判断是否应该跳过该日志
+         */
+        private boolean shouldSkip(String content) {
+            if (content == null || content.trim().isEmpty()) {
+                return true;
+            }
+            
+            String lowerContent = content.toLowerCase();
+            
+            // 跳过SQL语句
+            if (lowerContent.contains("hibernate:") || 
+                lowerContent.contains("select ") || 
+                lowerContent.contains("insert ") || 
+                lowerContent.contains("update ") || 
+                lowerContent.contains("delete ") ||
+                lowerContent.contains("create table") ||
+                lowerContent.contains("alter table") ||
+                lowerContent.contains("drop table")) {
+                return true;
+            }
+            
+            // 跳过Spring Boot启动相关日志
+            if (lowerContent.contains("started gdydzhpbzgfapplication") ||
+                lowerContent.contains("spring application startup") ||
+                lowerContent.contains("beanfactory") ||
+                lowerContent.contains("obtaining singleton bean") ||
+                lowerContent.contains("autowired annotation") ||
+                lowerContent.contains("initialized jpa entitymanagerfactory") ||
+                lowerContent.contains("hibernate is in classpath") ||
+                lowerContent.contains("creating shared instance") ||
+                lowerContent.startsWith("http://127.0.0.1:")) {
+                return true;
+            }
+            
+            // 跳过Hibernate相关日志
+            if (lowerContent.contains("hibernate") && 
+                (lowerContent.contains("dialect") || 
+                 lowerContent.contains("sessionfactory") ||
+                 lowerContent.contains("entitymanagerfactory"))) {
+                return true;
+            }
+            
+            return false;
+        }
+
+        @Override
+        public void close() {
+            try {
+                flush();
+                super.close();
+            } catch (Exception e) {
+                // 忽略异常
+            }
+        }
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/UmiOcrConfig.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/UmiOcrConfig.java
new file mode 100644
index 0000000..aa614da
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/UmiOcrConfig.java
@@ -0,0 +1,28 @@
+package com.zhpb.gdyd_zhpb_zgf.config;
+
+import com.fasterxml.jackson.databind.ObjectMapper;
+import com.zhpb.gdyd_zhpb_zgf.service.UmiOcrClient;
+import org.springframework.beans.factory.annotation.Value;
+import org.springframework.context.annotation.Bean;
+import org.springframework.context.annotation.Configuration;
+import org.springframework.web.client.RestTemplate;
+
+/**
+ * Umi-OCR 配置类
+ */
+@Configuration
+public class UmiOcrConfig {
+
+    @Value("${umi-ocr.base-url:http://127.0.0.1:1224}")
+    private String baseUrl;
+
+    @Bean
+    public RestTemplate restTemplate() {
+        return new RestTemplate();
+    }
+
+    @Bean
+    public UmiOcrClient umiOcrClient(RestTemplate restTemplate, ObjectMapper objectMapper) {
+        return new UmiOcrClient(baseUrl, restTemplate, objectMapper);
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/VectorizationHealthIndicator.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/VectorizationHealthIndicator.java
new file mode 100644
index 0000000..accbddb
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/VectorizationHealthIndicator.java
@@ -0,0 +1,127 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.config
+ * @ClassName: VectorizationHealthIndicator
+ * @Description: 向量化服务健康检查指示器，用于Actuator健康检查
+ * @Author: 张志锋
+ * @Date: 2025-11-03
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.config;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.File;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.FileRepository;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.boot.actuate.health.Health;
+import org.springframework.boot.actuate.health.HealthIndicator;
+import org.springframework.stereotype.Component;
+
+import java.time.LocalDateTime;
+import java.util.List;
+
+/**
+ * 向量化服务健康检查指示器
+ */
+@Component
+public class VectorizationHealthIndicator implements HealthIndicator {
+
+    private static final Logger logger = LoggerFactory.getLogger(VectorizationHealthIndicator.class);
+
+    @Autowired
+    private FileRepository fileRepository;
+
+    @Override
+    public Health health() {
+        logger.debug("开始执行向量化服务健康检查");
+        try {
+            // 检查数据库连接和文件表
+            logger.debug("查询所有文件记录");
+            List<File> files = fileRepository.findAll();
+            logger.debug("查询到 {} 个文件记录", files.size());
+
+            // 统计向量化状态
+            int totalFiles = files.size();
+            int completedFiles = 0;
+            int processingFiles = 0;
+            int failedFiles = 0;
+            int notStartedFiles = 0;
+
+            LocalDateTime lastVectorizedTime = null;
+
+            for (File file : files) {
+                String status = file.getVectorizationStatus();
+                if ("completed".equals(status)) {
+                    completedFiles++;
+                    if (lastVectorizedTime == null ||
+                        (file.getVectorizedTime() != null && file.getVectorizedTime().isAfter(lastVectorizedTime))) {
+                        lastVectorizedTime = file.getVectorizedTime();
+                    }
+                } else if ("processing".equals(status)) {
+                    processingFiles++;
+                } else if ("failed".equals(status)) {
+                    failedFiles++;
+                } else {
+                    notStartedFiles++;
+                }
+            }
+
+            // 计算健康状态
+            // 如果有太多失败的文件，认为服务不健康
+            double failureRate = totalFiles > 0 ? (double) failedFiles / totalFiles : 0.0;
+
+            if (failureRate > 0.5) {
+                // 失败率超过50%，认为服务不健康
+                logger.warn("向量化服务健康检查失败：失败率过高 ({})", String.format("%.2f%%", failureRate * 100));
+                return Health.down()
+                        .withDetail("service", "向量化服务")
+                        .withDetail("status", "健康状况异常")
+                        .withDetail("total_files", totalFiles)
+                        .withDetail("completed_files", completedFiles)
+                        .withDetail("processing_files", processingFiles)
+                        .withDetail("failed_files", failedFiles)
+                        .withDetail("not_started_files", notStartedFiles)
+                        .withDetail("failure_rate", String.format("%.2f%%", failureRate * 100))
+                        .withDetail("error", "向量化失败率过高")
+                        .build();
+            } else if (processingFiles > 0) {
+                // 有文件正在处理，认为是正常但有警告
+                logger.info("向量化服务健康检查通过：有文件正在处理");
+                return Health.up()
+                        .withDetail("service", "向量化服务")
+                        .withDetail("status", "正常（有文件正在处理）")
+                        .withDetail("total_files", totalFiles)
+                        .withDetail("completed_files", completedFiles)
+                        .withDetail("processing_files", processingFiles)
+                        .withDetail("failed_files", failedFiles)
+                        .withDetail("not_started_files", notStartedFiles)
+                        .withDetail("completion_rate", String.format("%.2f%%", totalFiles > 0 ? (double) completedFiles / totalFiles * 100 : 0))
+                        .withDetail("last_vectorized", lastVectorizedTime != null ? lastVectorizedTime.toString() : "无")
+                        .build();
+            } else {
+                // 正常状态
+                logger.info("向量化服务健康检查通过：运行正常");
+                return Health.up()
+                        .withDetail("service", "向量化服务")
+                        .withDetail("status", "运行正常")
+                        .withDetail("total_files", totalFiles)
+                        .withDetail("completed_files", completedFiles)
+                        .withDetail("processing_files", processingFiles)
+                        .withDetail("failed_files", failedFiles)
+                        .withDetail("not_started_files", notStartedFiles)
+                        .withDetail("completion_rate", String.format("%.2f%%", totalFiles > 0 ? (double) completedFiles / totalFiles * 100 : 0))
+                        .withDetail("last_vectorized", lastVectorizedTime != null ? lastVectorizedTime.toString() : "无")
+                        .build();
+            }
+
+        } catch (Exception e) {
+            return Health.down()
+                    .withDetail("service", "向量化服务")
+                    .withDetail("status", "检查失败")
+                    .withDetail("error", e.getMessage())
+                    .build();
+        }
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/WebConfig.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/WebConfig.java
new file mode 100644
index 0000000..6c04048
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/config/WebConfig.java
@@ -0,0 +1,52 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.config
+ * @ClassName: WebConfig
+ * @Description: Web配置类，注册API日志拦截器和静态资源处理器
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.config;
+
+import com.zhpb.gdyd_zhpb_zgf.interceptor.ApiLogInterceptor;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.context.annotation.Configuration;
+import org.springframework.web.servlet.config.annotation.InterceptorRegistry;
+import org.springframework.web.servlet.config.annotation.WebMvcConfigurer;
+import org.springframework.web.servlet.config.annotation.ResourceHandlerRegistry;
+
+/**
+ * Web配置类
+ */
+@Configuration
+public class WebConfig implements WebMvcConfigurer {
+
+    @Autowired
+    private ApiLogInterceptor apiLogInterceptor;
+
+    @Override
+    public void addInterceptors(InterceptorRegistry registry) {
+        registry.addInterceptor(apiLogInterceptor)
+                .addPathPatterns("/**")
+                .excludePathPatterns(
+                        "/error",
+                        "/favicon.ico",
+                        "/static/**",
+                        "/webjars/**"
+                );
+    }
+
+    @Override
+    public void addResourceHandlers(ResourceHandlerRegistry registry) {
+        // 确保静态资源正常访问
+        registry.addResourceHandler("/static/**")
+                .addResourceLocations("classpath:/static/");
+
+        // 添加 /view/** 路径映射到 static/view/ 目录
+        registry.addResourceHandler("/view/**")
+                .addResourceLocations("classpath:/static/view/");
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/context/ModelRequestContextHolder.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/context/ModelRequestContextHolder.java
new file mode 100644
index 0000000..05a92e0
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/context/ModelRequestContextHolder.java
@@ -0,0 +1,48 @@
+package com.zhpb.gdyd_zhpb_zgf.context;
+
+/**
+ * 线程级模型请求上下文，用于在调用模型/向量化前标记业务来源。
+ */
+public final class ModelRequestContextHolder {
+
+    private static final ThreadLocal<String> REQUEST_SOURCE = new ThreadLocal<>();
+
+    private ModelRequestContextHolder() {
+    }
+
+    /**
+     * 设置当前线程的模型请求来源。
+     *
+     * @param requestSource 业务来源标识，例如 QA_CHAT、DOC_SUMMARY
+     */
+    public static void setRequestSource(String requestSource) {
+        if (requestSource == null || requestSource.isBlank()) {
+            REQUEST_SOURCE.remove();
+        } else {
+            REQUEST_SOURCE.set(requestSource);
+        }
+    }
+
+    /**
+     * 获取当前线程的模型请求来源，若为空则返回默认值。
+     *
+     * @param defaultValue 默认业务来源
+     * @return 请求来源
+     */
+    public static String getRequestSourceOrDefault(String defaultValue) {
+        String source = REQUEST_SOURCE.get();
+        if (source == null || source.isBlank()) {
+            return defaultValue;
+        }
+        return source;
+    }
+
+    /**
+     * 清理当前线程上下文，防止线程复用导致交叉污染。
+     */
+    public static void clear() {
+        REQUEST_SOURCE.remove();
+    }
+}
+
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/controller/ComparisonController.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/controller/ComparisonController.java
new file mode 100644
index 0000000..512def3
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/controller/ComparisonController.java
@@ -0,0 +1,248 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.controller
+ * @ClassName: ComparisonController
+ * @Description: 供应商对比分析控制器
+ * @Author: 张志锋
+ * @Date: 2025-11-04
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.controller;
+
+import com.zhpb.gdyd_zhpb_zgf.dto.ComparisonRequest;
+import com.zhpb.gdyd_zhpb_zgf.dto.ComparisonResponse;
+import com.zhpb.gdyd_zhpb_zgf.service.ComparisonService;
+import com.zhpb.gdyd_zhpb_zgf.service.UserLogService;
+import io.swagger.v3.oas.annotations.Operation;
+import io.swagger.v3.oas.annotations.Parameter;
+import io.swagger.v3.oas.annotations.responses.ApiResponse;
+import io.swagger.v3.oas.annotations.responses.ApiResponses;
+import io.swagger.v3.oas.annotations.tags.Tag;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.http.ResponseEntity;
+import org.springframework.web.bind.annotation.*;
+
+import jakarta.servlet.http.HttpServletRequest;
+import jakarta.validation.Valid;
+import java.util.HashMap;
+import java.util.Map;
+
+/**
+ * 供应商对比分析控制器
+ */
+@RestController
+@RequestMapping("/api/v1/comparison")
+@Tag(name = "供应商对比分析", description = "供应商横向对比分析相关接口")
+public class ComparisonController {
+
+    private static final Logger logger = LoggerFactory.getLogger(ComparisonController.class);
+
+    @Autowired
+    private ComparisonService comparisonService;
+
+    @Autowired
+    private UserLogService userLogService;
+
+    /**
+     * 获取指定项目的供应商数据
+     */
+    @GetMapping("/vendors")
+    @Operation(
+        summary = "获取项目供应商数据",
+        description = "获取指定项目下的所有供应商及其关联文件，用于在对比时选择。"
+    )
+    @ApiResponses(value = {
+        @ApiResponse(
+            responseCode = "200",
+            description = "获取成功",
+            content = @io.swagger.v3.oas.annotations.media.Content(
+                mediaType = "application/json",
+                schema = @io.swagger.v3.oas.annotations.media.Schema(implementation = ComparisonResponse.VendorsData.class)
+            )
+        ),
+        @ApiResponse(responseCode = "422", description = "参数验证错误")
+    })
+    public ResponseEntity<Map<String, Object>> getProjectVendors(
+        @Parameter(description = "项目ID", required = true, example = "project-001")
+        @RequestParam("project_id") String projectId,
+        @Parameter(description = "用户ID", required = false, example = "user-001")
+        @RequestParam(value = "userId", required = false) String userId,
+        @Parameter(description = "用户ID（兼容字段）", required = false, example = "user-001")
+        @RequestParam(value = "userld", required = false) String legacyUserId,
+        @Parameter(description = "用户名", required = false, example = "张三")
+        @RequestParam(value = "userName", required = false) String userName,
+        HttpServletRequest request) {
+
+        Map<String, Object> response = new HashMap<>();
+        String resolvedUserId = userId != null ? userId : legacyUserId;
+
+        try {
+            ComparisonResponse.VendorsData data = comparisonService.getProjectVendors(projectId);
+
+            response.put("code", 0);
+            response.put("message", "获取供应商数据成功");
+            response.put("data", data);
+
+            userLogService.logUserOperation(resolvedUserId, userName, projectId,
+                "VIEW_VENDORS", "获取项目供应商数据", null, request);
+
+            return ResponseEntity.ok(response);
+
+        } catch (Exception e) {
+            logger.error("获取项目供应商数据失败", e);
+            response.put("code", 500);
+            response.put("message", "获取供应商数据失败: " + e.getMessage());
+            return ResponseEntity.internalServerError().body(response);
+        }
+    }
+
+    /**
+     * 启动供应商对比分析任务
+     */
+    @PostMapping("/start")
+    @Operation(
+        summary = "启动对比分析任务",
+        description = "提交一个异步的横向对比任务。"
+    )
+    @ApiResponses(value = {
+        @ApiResponse(
+            responseCode = "201",
+            description = "任务启动成功",
+            content = @io.swagger.v3.oas.annotations.media.Content(
+                mediaType = "application/json",
+                schema = @io.swagger.v3.oas.annotations.media.Schema(implementation = ComparisonResponse.StartTaskData.class)
+            )
+        ),
+        @ApiResponse(responseCode = "422", description = "参数验证错误")
+    })
+    public ResponseEntity<Map<String, Object>> startComparison(
+        @Parameter(description = "对比分析请求参数", required = true)
+        @Valid @RequestBody ComparisonRequest request) {
+
+        Map<String, Object> response = new HashMap<>();
+
+        try {
+            ComparisonResponse.StartTaskData data = comparisonService.startComparison(request);
+
+            response.put("code", 0);
+            response.put("message", "对比分析任务已启动");
+            response.put("data", data);
+
+            return ResponseEntity.status(201).body(response);
+
+        } catch (Exception e) {
+            logger.error("启动对比分析任务失败", e);
+            response.put("code", 500);
+            response.put("message", "启动对比分析任务失败: " + e.getMessage());
+            return ResponseEntity.internalServerError().body(response);
+        }
+    }
+
+    /**
+     * 获取对比任务历史
+     */
+    @GetMapping("/history/tasks")
+    @Operation(
+        summary = "获取对比任务历史",
+        description = "获取指定用户在某个项目下的所有历史对比任务列表。"
+    )
+    @ApiResponses(value = {
+        @ApiResponse(
+            responseCode = "200",
+            description = "获取成功",
+            content = @io.swagger.v3.oas.annotations.media.Content(
+                mediaType = "application/json",
+                schema = @io.swagger.v3.oas.annotations.media.Schema(implementation = ComparisonResponse.TaskHistoryData.class)
+            )
+        ),
+        @ApiResponse(responseCode = "422", description = "参数验证错误")
+    })
+    public ResponseEntity<Map<String, Object>> getComparisonHistory(
+        @Parameter(description = "项目ID", required = true, example = "project-001")
+        @RequestParam("project_id") String projectId,
+        @Parameter(description = "用户ID", required = true, example = "user-001")
+        @RequestParam("user_id") String userId,
+        @Parameter(description = "用户名", required = false, example = "张三")
+        @RequestParam(value = "userName", required = false) String userName,
+        @Parameter(description = "用户名（兼容字段）", required = false, example = "张三")
+        @RequestParam(value = "user_name", required = false) String legacyUserName,
+        HttpServletRequest request) {
+
+        Map<String, Object> response = new HashMap<>();
+
+        try {
+            ComparisonResponse.TaskHistoryData data = comparisonService.getComparisonHistory(projectId, userId);
+
+            response.put("code", 0);
+            response.put("message", "获取对比任务历史成功");
+            response.put("data", data);
+
+            String resolvedUserName = userName;
+            if (resolvedUserName == null || resolvedUserName.isBlank()) {
+                resolvedUserName = legacyUserName;
+            }
+            if ((resolvedUserName == null || resolvedUserName.isBlank()) && request != null) {
+                resolvedUserName = request.getHeader("User-Name");
+                if (resolvedUserName == null || resolvedUserName.isBlank()) {
+                    resolvedUserName = request.getHeader("userName");
+                }
+            }
+
+            userLogService.logUserOperation(userId, resolvedUserName, projectId,
+                "VIEW_COMPARISON_HISTORY", "获取对比任务历史", null, request);
+
+            return ResponseEntity.ok(response);
+
+        } catch (Exception e) {
+            logger.error("获取对比任务历史失败", e);
+            response.put("code", 500);
+            response.put("message", "获取对比任务历史失败: " + e.getMessage());
+            return ResponseEntity.internalServerError().body(response);
+        }
+    }
+
+    /**
+     * 获取对比任务结果
+     */
+    @GetMapping("/result")
+    @Operation(
+        summary = "获取对比任务结果",
+        description = "根据任务ID，获取对比任务的状态和结果。甲方可通过轮询此接口来获取最终结果。"
+    )
+    @ApiResponses(value = {
+        @ApiResponse(
+            responseCode = "200",
+            description = "获取成功",
+            content = @io.swagger.v3.oas.annotations.media.Content(
+                mediaType = "application/json",
+                schema = @io.swagger.v3.oas.annotations.media.Schema(implementation = ComparisonResponse.ComparisonResultData.class)
+            )
+        ),
+        @ApiResponse(responseCode = "422", description = "参数验证错误")
+    })
+    public ResponseEntity<Map<String, Object>> getComparisonResult(
+        @Parameter(description = "任务ID", required = true, example = "task-001")
+        @RequestParam("task_id") String taskId) {
+
+        Map<String, Object> response = new HashMap<>();
+
+        try {
+            ComparisonResponse.ComparisonResultData data = comparisonService.getComparisonResult(taskId);
+
+            response.put("code", 0);
+            response.put("message", "获取对比任务结果成功");
+            response.put("data", data);
+
+            return ResponseEntity.ok(response);
+
+        } catch (Exception e) {
+            logger.error("获取对比任务结果失败", e);
+            response.put("code", 500);
+            response.put("message", "获取对比任务结果失败: " + e.getMessage());
+            return ResponseEntity.internalServerError().body(response);
+        }
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/controller/FileBrowseController.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/controller/FileBrowseController.java
new file mode 100644
index 0000000..d56fe56
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/controller/FileBrowseController.java
@@ -0,0 +1,110 @@
+package com.zhpb.gdyd_zhpb_zgf.controller;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.File;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.FileRepository;
+import io.swagger.v3.oas.annotations.Operation;
+import io.swagger.v3.oas.annotations.Parameter;
+import io.swagger.v3.oas.annotations.tags.Tag;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.core.io.FileSystemResource;
+import org.springframework.core.io.Resource;
+import org.springframework.http.HttpHeaders;
+import org.springframework.http.HttpStatus;
+import org.springframework.http.MediaType;
+import org.springframework.http.ResponseEntity;
+import org.springframework.util.CollectionUtils;
+import org.springframework.web.bind.annotation.GetMapping;
+import org.springframework.web.bind.annotation.PathVariable;
+import org.springframework.web.bind.annotation.RequestMapping;
+import org.springframework.web.bind.annotation.RequestParam;
+import org.springframework.web.bind.annotation.RestController;
+
+import java.net.URLEncoder;
+import java.nio.charset.StandardCharsets;
+import java.nio.file.Files;
+import java.nio.file.Path;
+import java.nio.file.Paths;
+import java.util.Collections;
+import java.util.List;
+import java.util.Map;
+import java.util.stream.Collectors;
+
+/**
+ * 简单文件浏览与下载接口
+ */
+@RestController
+@RequestMapping("/view/api/v1/files")
+@Tag(name = "文件浏览", description = "列出已存储文件并提供下载")
+public class FileBrowseController {
+
+    private static final Logger logger = LoggerFactory.getLogger(FileBrowseController.class);
+
+    private final FileRepository fileRepository;
+
+    public FileBrowseController(FileRepository fileRepository) {
+        this.fileRepository = fileRepository;
+    }
+
+    /**
+     * 列出文件，可按项目/供应商筛选
+     */
+    @GetMapping("/list")
+    @Operation(summary = "列出已存储文件", description = "可按项目ID、供应商ID过滤")
+    public List<Map<String, Object>> listFiles(
+            @Parameter(description = "项目ID") @RequestParam(required = false) String projectId,
+            @Parameter(description = "供应商ID") @RequestParam(required = false) String supplierId) {
+
+        List<File> files;
+        if (projectId != null && supplierId != null) {
+            files = fileRepository.findByProjectIdAndSupplierId(projectId, supplierId);
+        } else if (projectId != null) {
+            files = fileRepository.findByProjectId(projectId);
+        } else if (supplierId != null) {
+            files = fileRepository.findBySupplierId(supplierId);
+        } else {
+            files = fileRepository.findAll();
+        }
+
+        if (CollectionUtils.isEmpty(files)) {
+            return Collections.emptyList();
+        }
+
+        return files.stream().map(f -> Map.<String, Object>ofEntries(
+                Map.entry("id", f.getId()),
+                Map.entry("fileId", f.getFileId()),
+                Map.entry("projectId", f.getProjectId()),
+                Map.entry("supplierId", f.getSupplierId()),
+                Map.entry("fileName", f.getFileName()),
+                Map.entry("filePath", f.getFilePath()),
+                Map.entry("fileSize", f.getFileSize()),
+                Map.entry("fileExtension", f.getFileExtension()),
+                Map.entry("fileTypeId", f.getFileTypeId()),
+                Map.entry("status", f.getStatus()),
+                Map.entry("downloadUrl", "/api/v1/files/download/" + f.getId())
+        )).collect(Collectors.toList());
+    }
+
+    /**
+     * 按ID下载文件
+     */
+    @GetMapping("/download/{id}")
+    @Operation(summary = "下载文件", description = "根据文件记录ID下载实际文件")
+    public ResponseEntity<Resource> download(@PathVariable Long id) {
+        return fileRepository.findById(id)
+                .map(file -> {
+                    Path path = Paths.get(file.getFilePath());
+                    if (!Files.exists(path)) {
+                        return ResponseEntity.status(HttpStatus.NOT_FOUND).<Resource>build();
+                    }
+                    Resource resource = new FileSystemResource(path);
+                    String encodedName = URLEncoder.encode(file.getFileName(), StandardCharsets.UTF_8);
+                    return ResponseEntity.ok()
+                            .header(HttpHeaders.CONTENT_DISPOSITION, "attachment; filename*=UTF-8''" + encodedName)
+                            .contentType(MediaType.APPLICATION_OCTET_STREAM)
+                            .body(resource);
+                })
+                .orElse(ResponseEntity.status(HttpStatus.NOT_FOUND).<Resource>build());
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/controller/FileController.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/controller/FileController.java
new file mode 100644
index 0000000..8706e6f
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/controller/FileController.java
@@ -0,0 +1,1009 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.controller
+ * @ClassName: FileController
+ * @Description: 文件管理控制器，提供文件上传、ZIP解压等RESTful API接口
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.controller;
+
+import com.zhpb.gdyd_zhpb_zgf.dto.FileUploadRequest;
+import com.zhpb.gdyd_zhpb_zgf.dto.FileUploadResponse;
+import com.zhpb.gdyd_zhpb_zgf.dto.ProjectFilesResponse;
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.File;
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.Project;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.FileDirectoryRepository;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.FileRepository;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.ProjectRepository;
+import com.zhpb.gdyd_zhpb_zgf.service.FileService;
+import com.zhpb.gdyd_zhpb_zgf.service.FilePreviewService;
+import com.zhpb.gdyd_zhpb_zgf.utils.ApiAuthUtils;
+import io.swagger.v3.oas.annotations.Operation;
+import io.swagger.v3.oas.annotations.Parameter;
+import io.swagger.v3.oas.annotations.media.Content;
+import io.swagger.v3.oas.annotations.media.Schema;
+import io.swagger.v3.oas.annotations.responses.ApiResponse;
+import io.swagger.v3.oas.annotations.responses.ApiResponses;
+import io.swagger.v3.oas.annotations.tags.Tag;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.core.io.InputStreamResource;
+import org.springframework.http.HttpHeaders;
+import org.springframework.http.HttpStatus;
+import org.springframework.http.MediaType;
+import org.springframework.http.ResponseEntity;
+import org.springframework.web.bind.annotation.*;
+import org.springframework.web.multipart.MultipartFile;
+import org.springframework.web.servlet.mvc.method.annotation.StreamingResponseBody;
+
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.beans.factory.annotation.Autowired;
+import com.zhpb.gdyd_zhpb_zgf.service.UserLogService;
+
+import jakarta.servlet.http.HttpServletRequest;
+
+import java.io.IOException;
+import java.io.InputStream;
+import java.nio.charset.StandardCharsets;
+import java.util.ArrayList;
+import java.util.HashMap;
+import java.util.List;
+import java.util.Map;
+
+/**
+ * 文件管理控制器
+ */
+@RestController
+@RequestMapping("/api/v1/files")
+@Tag(name = "文件管理", description = "文件上传、下载、管理相关接口")
+public class FileController {
+    @Autowired
+    private FileService fileService;
+
+    @Autowired
+    private FilePreviewService filePreviewService;
+
+    @Autowired
+    private FileRepository fileRepository;
+
+    @Autowired
+    private ProjectRepository projectRepository;
+
+    @Autowired
+    private ApiAuthUtils apiAuthUtils;
+
+    @Autowired
+    private FileDirectoryRepository fileDirectoryRepository;
+
+    @Autowired
+    private UserLogService userLogService;
+
+    private static final Logger logger = LoggerFactory.getLogger(FileController.class);
+
+    /**
+     * 上传ZIP压缩包
+     */
+    @PostMapping(value = "/uploadzip", consumes = MediaType.MULTIPART_FORM_DATA_VALUE)
+    @Operation(
+            summary = "上传ZIP压缩包",
+            description = "上传ZIP压缩包并自动解压，支持的文件格式：PDF、DOC、DOCX、TXT。请求体类型：multipart/form-data"
+    )
+    @io.swagger.v3.oas.annotations.parameters.RequestBody(
+            description = "文件上传表单，使用multipart/form-data格式",
+            required = true,
+            content = @Content(
+                    mediaType = MediaType.MULTIPART_FORM_DATA_VALUE
+            )
+    )
+    @ApiResponses(value = {
+            @ApiResponse(
+                    responseCode = "200",
+                    description = "上传成功",
+                    content = @Content(
+                            mediaType = MediaType.APPLICATION_JSON_VALUE,
+                            schema = @Schema(implementation = FileUploadResponse.class)
+                    )
+            ),
+            @ApiResponse(
+                    responseCode = "422",
+                    description = "验证错误",
+                    content = @Content(mediaType = MediaType.APPLICATION_JSON_VALUE)
+            )
+    })
+    public ResponseEntity<Map<String, Object>> uploadZip(
+            @Parameter(description = "项目ID", required = true, example = "project-001")
+            @RequestParam("project_id") String projectId,
+            @Parameter(description = "ZIP文件", required = true)
+            @RequestParam("file") MultipartFile file,
+            HttpServletRequest request) {
+
+        Map<String, Object> response = new HashMap<>();
+
+        try {
+            // 验证文件
+            if (file.isEmpty()) {
+                response.put("code", 422);
+                response.put("message", "文件不能为空");
+                return ResponseEntity.status(422).body(response);
+            }
+
+            String originalFilename = file.getOriginalFilename();
+            if (originalFilename == null || !originalFilename.toLowerCase().endsWith(".zip")) {
+                response.put("code", 422);
+                response.put("message", "文件必须是ZIP格式");
+                return ResponseEntity.status(422).body(response);
+            }
+
+            // 验证项目ID
+            if (projectId == null || projectId.trim().isEmpty()) {
+                response.put("code", 422);
+                response.put("message", "项目ID不能为空");
+                return ResponseEntity.status(422).body(response);
+            }
+
+            // 上传并解压
+            fileService.uploadAndExtractZip(projectId, file);
+
+            // 获取项目文件信息（与查询API保持一致的格式）
+            ProjectFilesResponse.ProjectFilesData data = fileService.getProjectFiles(projectId);
+
+            response.put("code", 0);
+            response.put("message", "文件上传成功，共解压" + data.getTotalFiles() + "个文件");
+            response.put("data", data);
+
+            // 记录用户操作日志
+            // 从请求头中获取userId和userName
+            String userId = request.getHeader("User-Id");
+            String userName = request.getHeader("User-Name");
+            if (userId != null && !userId.isEmpty()) {
+                userLogService.logUserOperation(userId, userName, projectId, 
+                    "UPLOAD_FILE", "上传ZIP文件: " + file.getOriginalFilename() + ", 解压文件数: " + data.getTotalFiles(), 
+                    null, request);
+            }
+
+            return ResponseEntity.ok(response);
+
+        } catch (IllegalArgumentException e) {
+            response.put("code", 422);
+            response.put("message", e.getMessage());
+            return ResponseEntity.status(422).body(response);
+        } catch (org.springframework.web.multipart.MaxUploadSizeExceededException e) {
+            response.put("code", 413);
+            response.put("message", "文件大小超过限制，最大允许上传500MB");
+            response.put("error", "MaxUploadSizeExceeded");
+            return ResponseEntity.status(413).body(response);
+        } catch (IOException e) {
+            response.put("code", 500);
+            response.put("message", "文件处理失败: " + e.getMessage());
+            return ResponseEntity.status(500).body(response);
+        } catch (Exception e) {
+            response.put("code", 500);
+            response.put("message", "系统错误: " + e.getMessage());
+            return ResponseEntity.status(500).body(response);
+        }
+    }
+
+    /**
+     * PDF文件预览
+     */
+    @GetMapping("/preview/pdf/{file_id}")
+    @Operation(
+            summary = "PDF文件预览",
+            description = "在线预览PDF文件，返回PDF文件流"
+    )
+    @ApiResponses(value = {
+            @ApiResponse(
+                    responseCode = "200",
+                    description = "预览成功",
+                    content = @Content(mediaType = MediaType.APPLICATION_PDF_VALUE)
+            ),
+            @ApiResponse(
+                    responseCode = "404",
+                    description = "文件不存在",
+                    content = @Content(mediaType = MediaType.APPLICATION_JSON_VALUE)
+            ),
+            @ApiResponse(
+                    responseCode = "422",
+                    description = "验证错误",
+                    content = @Content(mediaType = MediaType.APPLICATION_JSON_VALUE)
+            )
+    })
+    public ResponseEntity<StreamingResponseBody> previewPdf(
+            @Parameter(description = "文件ID", required = true)
+            @PathVariable("file_id") String fileId) {
+        // 为兼容旧链接，直接走流式实现，默认chunkSize=8192
+        return buildStreamResponse(fileId, 8192);
+    }
+
+    /**
+     * PDF文件流式传输（支持自定义分片大小）
+     */
+    @GetMapping("/stream/pdf/{file_id}")
+    @Operation(
+            summary = "PDF 文件流式下载",
+            description = "以流的方式返回 PDF 内容，支持指定 chunk_size（默认 8192 字节）"
+    )
+    public ResponseEntity<StreamingResponseBody> streamPdf(
+            @Parameter(description = "文件ID", required = true)
+            @PathVariable("file_id") String fileId,
+            @Parameter(description = "单次写入的字节数，范围 1024~2097152")
+            @RequestParam(value = "chunk_size", required = false, defaultValue = "8192") int chunkSize) {
+        return buildStreamResponse(fileId, chunkSize);
+    }
+
+    /**
+     * Word文档预览（转换为PDF后预览）
+     */
+    @GetMapping("/preview/word/{file_id}")
+    @Operation(
+            summary = "Word文档预览",
+            description = "预览Word文档（转换为PDF后预览），返回PDF文件流"
+    )
+    @ApiResponses(value = {
+            @ApiResponse(
+                    responseCode = "200",
+                    description = "预览成功",
+                    content = @Content(mediaType = MediaType.APPLICATION_PDF_VALUE)
+            ),
+            @ApiResponse(
+                    responseCode = "404",
+                    description = "文件不存在",
+                    content = @Content(mediaType = MediaType.APPLICATION_JSON_VALUE)
+            ),
+            @ApiResponse(
+                    responseCode = "422",
+                    description = "验证错误或不支持的文件类型",
+                    content = @Content(mediaType = MediaType.APPLICATION_JSON_VALUE)
+            )
+    })
+    public ResponseEntity<?> previewWord(
+            @Parameter(description = "文件ID", required = true)
+            @PathVariable("file_id") String fileId) {
+
+        try {
+            InputStream inputStream = java.util.Objects.requireNonNull(
+                    filePreviewService.previewWord(fileId), "预览流为空");
+
+            return ResponseEntity.ok()
+                    .header(HttpHeaders.CONTENT_TYPE, MediaType.APPLICATION_PDF_VALUE)
+                    .header(HttpHeaders.CONTENT_DISPOSITION, "inline; filename=\"preview.pdf\"")
+                    .body(new InputStreamResource(inputStream));
+
+        } catch (IllegalArgumentException e) {
+            Map<String, Object> error = new HashMap<>();
+            error.put("code", 200);
+            error.put("message", e.getMessage());
+            return ResponseEntity.ok().body(error);
+        } catch (UnsupportedOperationException e) {
+            Map<String, Object> error = new HashMap<>();
+            error.put("code", 422);
+            error.put("message", e.getMessage());
+            error.put("total", 0);
+            return ResponseEntity.status(HttpStatus.UNPROCESSABLE_ENTITY).body(error);
+        } catch (Exception e) {
+            Map<String, Object> error = new HashMap<>();
+            error.put("code", 500);
+            error.put("message", "预览失败: " + e.getMessage());
+            error.put("total", 0);
+            return ResponseEntity.status(HttpStatus.INTERNAL_SERVER_ERROR).body(error);
+        }
+    }
+
+    private ResponseEntity<StreamingResponseBody> buildStreamResponse(String fileId, int chunkSize) {
+        try {
+            if (chunkSize < 1024 || chunkSize > 2_097_152) {
+                return jsonStreamResponse(HttpStatus.BAD_REQUEST,
+                        """
+                        {"code":400,"message":"chunk_size 需要在 1024~2097152 之间"}
+                        """);
+            }
+
+            InputStream inputStream = java.util.Objects.requireNonNull(
+                    filePreviewService.previewPdf(fileId), "预览流为空");
+            StreamingResponseBody body = outputStream -> {
+                long written = 0;
+                try (InputStream in = inputStream) {
+                    byte[] buffer = new byte[chunkSize];
+                    int len;
+                    while ((len = in.read(buffer)) != -1) {
+                        outputStream.write(buffer, 0, len);
+                        outputStream.flush();
+                        written += len;
+                    }
+                }
+                logger.info("流式输出完成: fileId={}, bytes={}", fileId, written);
+            };
+
+            return ResponseEntity.ok()
+                    .header(HttpHeaders.CONTENT_TYPE, MediaType.APPLICATION_PDF_VALUE)
+                    .header(HttpHeaders.CONTENT_DISPOSITION, "inline; filename=\"stream.pdf\"")
+                    .body(body);
+
+        } catch (IllegalArgumentException e) {
+            return jsonStreamResponse(HttpStatus.NOT_FOUND,
+                    """
+                    {"code":404,"message":"%s"}
+                    """.formatted(e.getMessage()));
+        } catch (UnsupportedOperationException e) {
+            return jsonStreamResponse(HttpStatus.UNPROCESSABLE_ENTITY,
+                    """
+                    {"code":422,"message":"%s"}
+                    """.formatted(e.getMessage()));
+        } catch (Exception e) {
+            logger.error("流式传输PDF失败: fileId={}, error={}", fileId, e.getMessage(), e);
+            return jsonStreamResponse(HttpStatus.INTERNAL_SERVER_ERROR,
+                    """
+                    {"code":500,"message":"文件传输失败: %s"}
+                    """.formatted(e.getMessage()));
+        }
+    }
+
+    private ResponseEntity<StreamingResponseBody> jsonStreamResponse(HttpStatus status, String json) {
+        StreamingResponseBody body = outputStream ->
+                outputStream.write(json.getBytes(StandardCharsets.UTF_8));
+        return ResponseEntity.status(status.value())
+                .header(HttpHeaders.CONTENT_TYPE, MediaType.APPLICATION_JSON_VALUE)
+                .body(body);
+    }
+
+    /**
+     * 查询所有文件的向量化状态统计
+     */
+    @GetMapping("/vectorization/stats")
+    @Operation(
+            summary = "查询文件向量化状态统计",
+            description = "获取系统中所有文件的向量化状态统计信息，包括完成率、各状态文件数量等"
+    )
+    @ApiResponses(value = {
+            @ApiResponse(
+                    responseCode = "200",
+                    description = "查询成功",
+                    content = @Content(
+                            mediaType = MediaType.APPLICATION_JSON_VALUE,
+                            schema = @Schema(implementation = FileUploadResponse.VectorizationStats.class)
+                    )
+            ),
+            @ApiResponse(
+                    responseCode = "500",
+                    description = "系统错误",
+                    content = @Content(mediaType = MediaType.APPLICATION_JSON_VALUE)
+            )
+    })
+    public ResponseEntity<Map<String, Object>> getVectorizationStats() {
+        Map<String, Object> response = new HashMap<>();
+
+        try {
+            // 查询所有文件
+            List<File> allFiles = fileRepository.findAll();
+
+            // 统计各种状态的文件数量
+            int totalFiles = allFiles.size();
+            int completedFiles = 0;
+            int processingFiles = 0;
+            int failedFiles = 0;
+            int notStartedFiles = 0;
+
+            for (File file : allFiles) {
+                String status = file.getVectorizationStatus();
+                if ("completed".equals(status)) {
+                    completedFiles++;
+                } else if ("processing".equals(status)) {
+                    processingFiles++;
+                } else if ("failed".equals(status)) {
+                    failedFiles++;
+                } else {
+                    notStartedFiles++;
+                }
+            }
+
+            // 计算完成率
+            double completionRate = totalFiles > 0 ? (double) completedFiles / totalFiles : 0.0;
+
+            // 获取最近的文件（按更新时间倒序，取前10个）
+            List<File> recentFiles = allFiles.stream()
+                    .sorted((f1, f2) -> f2.getUpdateTime().compareTo(f1.getUpdateTime()))
+                    .limit(10)
+                    .collect(java.util.stream.Collectors.toList());
+
+            // 构建响应数据
+            FileUploadResponse.VectorizationStats stats = new FileUploadResponse.VectorizationStats();
+            stats.setTotalFiles(totalFiles);
+            stats.setCompletedFiles(completedFiles);
+            stats.setProcessingFiles(processingFiles);
+            stats.setFailedFiles(failedFiles);
+            stats.setNotStartedFiles(notStartedFiles);
+            stats.setCompletionRate(completionRate);
+
+            // 转换最近文件列表
+            List<FileUploadResponse.VectorizationStatus> recentStatusList = recentFiles.stream()
+                    .map(this::convertToVectorizationStatus)
+                    .collect(java.util.stream.Collectors.toList());
+            stats.setRecentFiles(recentStatusList);
+
+            response.put("code", 0);
+            response.put("message", "查询成功");
+            response.put("data", stats);
+
+            return ResponseEntity.ok(response);
+
+        } catch (Exception e) {
+            response.put("code", 500);
+            response.put("message", "查询失败: " + e.getMessage());
+            return ResponseEntity.status(HttpStatus.INTERNAL_SERVER_ERROR).body(response);
+        }
+    }
+
+    /**
+     * 查询指定文件的向量化状态
+     */
+    @GetMapping("/vectorization/status/{file_id}")
+    @Operation(
+            summary = "查询指定文件的向量化状态",
+            description = "根据文件ID查询特定文件的向量化状态信息"
+    )
+    @ApiResponses(value = {
+            @ApiResponse(
+                    responseCode = "200",
+                    description = "查询成功",
+                    content = @Content(
+                            mediaType = MediaType.APPLICATION_JSON_VALUE,
+                            schema = @Schema(implementation = FileUploadResponse.VectorizationStatus.class)
+                    )
+            ),
+            @ApiResponse(
+                    responseCode = "404",
+                    description = "文件不存在",
+                    content = @Content(mediaType = MediaType.APPLICATION_JSON_VALUE)
+            ),
+            @ApiResponse(
+                    responseCode = "500",
+                    description = "系统错误",
+                    content = @Content(mediaType = MediaType.APPLICATION_JSON_VALUE)
+            )
+    })
+    public ResponseEntity<Map<String, Object>> getFileVectorizationStatus(
+            @Parameter(description = "文件ID", required = true)
+            @PathVariable("file_id") String fileId) {
+
+        Map<String, Object> response = new HashMap<>();
+
+        try {
+            java.util.Optional<File> fileOptional = fileRepository.findByFileId(fileId);
+            if (!fileOptional.isPresent()) {
+                response.put("code", 404);
+                response.put("message", "文件不存在");
+                response.put("total", 0);
+                return ResponseEntity.ok().body(response);
+            }
+            File file = fileOptional.get();
+
+            FileUploadResponse.VectorizationStatus status = convertToVectorizationStatus(file);
+
+            response.put("code", 0);
+            response.put("message", "查询成功");
+            response.put("data", status);
+
+            return ResponseEntity.ok(response);
+
+        } catch (Exception e) {
+            response.put("code", 500);
+            response.put("message", "查询失败: " + e.getMessage());
+            return ResponseEntity.status(HttpStatus.INTERNAL_SERVER_ERROR).body(response);
+        }
+    }
+
+    /**
+     * 查询指定项目的文件向量化状态
+     */
+    @GetMapping("/vectorization/project/{project_id}")
+    @Operation(
+            summary = "查询项目文件的向量化状态",
+            description = "根据项目ID查询该项目下所有文件的向量化状态信息"
+    )
+    @ApiResponses(value = {
+            @ApiResponse(
+                    responseCode = "200",
+                    description = "查询成功",
+                    content = @Content(
+                            mediaType = MediaType.APPLICATION_JSON_VALUE,
+                            schema = @Schema(implementation = FileUploadResponse.VectorizationStats.class)
+                    )
+            ),
+            @ApiResponse(
+                    responseCode = "500",
+                    description = "系统错误",
+                    content = @Content(mediaType = MediaType.APPLICATION_JSON_VALUE)
+            )
+    })
+    public ResponseEntity<Map<String, Object>> getProjectVectorizationStatus(
+            @Parameter(description = "项目ID", required = true)
+            @PathVariable("project_id") String projectId) {
+
+        Map<String, Object> response = new HashMap<>();
+
+        try {
+            // 查询项目下的所有文件
+            List<File> projectFiles = fileRepository.findByProjectId(projectId);
+
+            // 统计各种状态的文件数量
+            int totalFiles = projectFiles.size();
+            int completedFiles = 0;
+            int processingFiles = 0;
+            int failedFiles = 0;
+            int notStartedFiles = 0;
+
+            for (File file : projectFiles) {
+                String status = file.getVectorizationStatus();
+                if ("completed".equals(status)) {
+                    completedFiles++;
+                } else if ("processing".equals(status)) {
+                    processingFiles++;
+                } else if ("failed".equals(status)) {
+                    failedFiles++;
+                } else {
+                    notStartedFiles++;
+                }
+            }
+
+            // 计算完成率
+            double completionRate = totalFiles > 0 ? (double) completedFiles / totalFiles : 0.0;
+
+            // 构建响应数据
+            FileUploadResponse.VectorizationStats stats = new FileUploadResponse.VectorizationStats();
+            stats.setTotalFiles(totalFiles);
+            stats.setCompletedFiles(completedFiles);
+            stats.setProcessingFiles(processingFiles);
+            stats.setFailedFiles(failedFiles);
+            stats.setNotStartedFiles(notStartedFiles);
+            stats.setCompletionRate(completionRate);
+
+            // 转换文件列表
+            List<FileUploadResponse.VectorizationStatus> statusList = projectFiles.stream()
+                    .map(this::convertToVectorizationStatus)
+                    .collect(java.util.stream.Collectors.toList());
+            stats.setRecentFiles(statusList);
+
+            response.put("code", 0);
+            response.put("message", "查询成功");
+            response.put("data", stats);
+
+            return ResponseEntity.ok(response);
+
+        } catch (Exception e) {
+            response.put("code", 500);
+            response.put("message", "查询失败: " + e.getMessage());
+            return ResponseEntity.status(HttpStatus.INTERNAL_SERVER_ERROR).body(response);
+        }
+    }
+
+    /**
+     * 测试向量化健康检查
+     */
+    @GetMapping("/test/vectorization-health")
+    @Operation(
+            summary = "测试向量化健康检查",
+            description = "测试向量化服务的健康检查功能"
+    )
+    public ResponseEntity<Map<String, Object>> testVectorizationHealth() {
+        Map<String, Object> response = new HashMap<>();
+
+        try {
+            // 模拟健康检查逻辑
+            List<File> allFiles = fileRepository.findAll();
+
+            int totalFiles = allFiles.size();
+            int completedFiles = 0;
+            int processingFiles = 0;
+            int failedFiles = 0;
+            int notStartedFiles = 0;
+
+            for (File file : allFiles) {
+                String status = file.getVectorizationStatus();
+                if ("completed".equals(status)) {
+                    completedFiles++;
+                } else if ("processing".equals(status)) {
+                    processingFiles++;
+                } else if ("failed".equals(status)) {
+                    failedFiles++;
+                } else {
+                    notStartedFiles++;
+                }
+            }
+
+            double completionRate = totalFiles > 0 ? (double) completedFiles / totalFiles : 0.0;
+
+            Map<String, Object> healthData = new HashMap<>();
+            healthData.put("service", "向量化服务");
+            healthData.put("status", "UP");
+            healthData.put("total_files", totalFiles);
+            healthData.put("completed_files", completedFiles);
+            healthData.put("processing_files", processingFiles);
+            healthData.put("failed_files", failedFiles);
+            healthData.put("not_started_files", notStartedFiles);
+            healthData.put("completion_rate", String.format("%.2f%%", completionRate * 100));
+
+            response.put("code", 0);
+            response.put("message", "健康检查测试成功");
+            response.put("data", healthData);
+
+            return ResponseEntity.ok(response);
+
+        } catch (Exception e) {
+            response.put("code", 500);
+            response.put("message", "健康检查测试失败: " + e.getMessage());
+            return ResponseEntity.status(HttpStatus.INTERNAL_SERVER_ERROR).body(response);
+        }
+    }
+
+    /**
+     * 查询项目文件信息
+     */
+    @GetMapping("/project/{project_id}/suppliers-files")
+    @Operation(
+            summary = "查询项目文件信息",
+            description = "根据项目ID查询该项目的所有文件信息，返回标准格式：包含projectId、totalFiles、files列表和companyFiles分组"
+    )
+    @ApiResponses(value = {
+            @ApiResponse(
+                    responseCode = "200",
+                    description = "查询成功",
+                    content = @Content(
+                            mediaType = MediaType.APPLICATION_JSON_VALUE,
+                            schema = @Schema(example = "{\"code\":0,\"message\":\"查询成功\",\"data\":{\"projectId\":\"7654\",\"totalFiles\":6,\"files\":[...],\"companyFiles\":{\"公司1\":[\"文件1.pdf\"],\"公司2\":[\"文件2.pdf\"]}}}")
+                    )
+            ),
+            @ApiResponse(
+                    responseCode = "404",
+                    description = "项目不存在",
+                    content = @Content(mediaType = MediaType.APPLICATION_JSON_VALUE)
+            ),
+            @ApiResponse(
+                    responseCode = "500",
+                    description = "系统错误",
+                    content = @Content(mediaType = MediaType.APPLICATION_JSON_VALUE)
+            )
+    })
+    public ResponseEntity<Map<String, Object>> getProjectSuppliersAndFiles(
+            @Parameter(description = "项目ID", required = true)
+            @PathVariable("project_id") String projectId) {
+
+        Map<String, Object> response = new HashMap<>();
+
+        try {
+            // 查询项目信息
+            Project project = projectRepository.findByProjectId(projectId);
+            if (project == null) {
+                response.put("code", 404);
+                response.put("message", "项目不存在");
+                response.put("total", 0);
+                return ResponseEntity.ok().body(response);
+            }
+
+            // 使用标准的 getProjectFiles 方法获取数据
+            ProjectFilesResponse.ProjectFilesData data = fileService.getProjectFiles(projectId);
+
+            response.put("code", 0);
+            response.put("message", "查询成功");
+            response.put("data", data);
+
+            return ResponseEntity.ok(response);
+
+        } catch (Exception e) {
+            response.put("code", 500);
+            response.put("message", "查询失败: " + e.getMessage());
+            return ResponseEntity.status(HttpStatus.INTERNAL_SERVER_ERROR).body(response);
+        }
+    }
+
+    /**
+     * 根据项目ID和供应商ID查询文件状态
+     */
+    @GetMapping("/project/{project_id}/supplier/{supplier_id}/status")
+    @Operation(
+            summary = "查询供应商文件状态",
+            description = "根据项目ID和供应商ID查询该供应商的所有文件状态，包括文件ID、向量化状态和目录摘取状态"
+    )
+    @ApiResponses(value = {
+            @ApiResponse(
+                    responseCode = "200",
+                    description = "查询成功",
+                    content = @Content(
+                            mediaType = MediaType.APPLICATION_JSON_VALUE,
+                            schema = @Schema(example = "{\"code\":0,\"message\":\"查询成功\",\"data\":{\"files\":[{\"fileId\":\"xxx\",\"fileName\":\"文件1.pdf\",\"vectorizationStatus\":\"completed\",\"catalogExtractionStatus\":\"completed\"}]}}")
+                    )
+            ),
+            @ApiResponse(
+                    responseCode = "404",
+                    description = "项目或供应商不存在",
+                    content = @Content(mediaType = MediaType.APPLICATION_JSON_VALUE)
+            ),
+            @ApiResponse(
+                    responseCode = "500",
+                    description = "系统错误",
+                    content = @Content(mediaType = MediaType.APPLICATION_JSON_VALUE)
+            )
+    })
+    public ResponseEntity<Map<String, Object>> getSupplierFileStatus(
+            @Parameter(description = "项目ID", required = true)
+            @PathVariable("project_id") String projectId,
+            @Parameter(description = "供应商ID", required = true)
+            @PathVariable("supplier_id") String supplierId) {
+
+        Map<String, Object> response = new HashMap<>();
+
+        try {
+            // 查询项目信息
+            Project project = projectRepository.findByProjectId(projectId);
+            if (project == null) {
+                response.put("code", 404);
+                response.put("message", "项目不存在");
+                response.put("total", 0);
+                return ResponseEntity.ok().body(response);
+            }
+
+            // 查询该项目和供应商下的所有文件
+            List<File> files = fileRepository.findByProjectIdAndSupplierId(projectId, supplierId);
+
+            if (files.isEmpty()) {
+                response.put("code", 404);
+                response.put("message", "未找到该供应商的文件");
+                response.put("total", 0);
+                return ResponseEntity.ok().body(response);
+            }
+
+            // 构建文件状态列表
+            List<Map<String, Object>> fileStatusList = new ArrayList<>();
+            for (File file : files) {
+                Map<String, Object> fileStatus = new HashMap<>();
+                fileStatus.put("fileId", file.getFileId());
+                fileStatus.put("fileName", file.getFileName());
+                fileStatus.put("fileExtension", file.getFileExtension());
+                fileStatus.put("fileSize", file.getFileSize());
+                
+                // 向量化状态
+                String vectorizationStatus = file.getVectorizationStatus();
+                if (vectorizationStatus == null || vectorizationStatus.isEmpty()) {
+                    vectorizationStatus = "not_started";
+                }
+                fileStatus.put("vectorizationStatus", vectorizationStatus);
+                
+                // 目录摘取状态：从文件实体中获取
+                String catalogExtractionStatus = file.getCatalogExtractionStatus();
+                if (catalogExtractionStatus == null || catalogExtractionStatus.isEmpty()) {
+                    catalogExtractionStatus = "not_started";
+                }
+                fileStatus.put("catalogExtractionStatus", catalogExtractionStatus);
+                
+                // 目录摘取重试次数
+                Integer retryCount = file.getCatalogExtractionRetryCount();
+                if (retryCount == null) {
+                    retryCount = 0;
+                }
+                fileStatus.put("catalogExtractionRetryCount", retryCount);
+                
+                // 目录数量
+                long catalogCount = fileDirectoryRepository.countByFileId(file.getFileId());
+                fileStatus.put("catalogCount", catalogCount);
+                
+                // 目录摘取完成时间
+                if (file.getCatalogExtractionTime() != null) {
+                    fileStatus.put("catalogExtractionTime", file.getCatalogExtractionTime());
+                }
+                
+                // 添加其他有用信息
+                fileStatus.put("createTime", file.getCreateTime());
+                fileStatus.put("updateTime", file.getUpdateTime());
+                
+                fileStatusList.add(fileStatus);
+            }
+
+            // 构建响应数据
+            Map<String, Object> data = new HashMap<>();
+            data.put("projectId", projectId);
+            data.put("supplierId", supplierId);
+            data.put("totalFiles", files.size());
+            data.put("files", fileStatusList);
+
+            response.put("code", 0);
+            response.put("message", "查询成功");
+            response.put("data", data);
+
+            return ResponseEntity.ok(response);
+
+        } catch (Exception e) {
+            logger.error("查询供应商文件状态失败: projectId={}, supplierId={}", projectId, supplierId, e);
+            response.put("code", 500);
+            response.put("message", "查询失败: " + e.getMessage());
+            return ResponseEntity.status(HttpStatus.INTERNAL_SERVER_ERROR).body(response);
+        }
+    }
+
+
+    /**
+     * 将File实体转换为VectorizationStatus DTO
+     */
+    private FileUploadResponse.VectorizationStatus convertToVectorizationStatus(File file) {
+        FileUploadResponse.VectorizationStatus status = new FileUploadResponse.VectorizationStatus();
+        status.setFileId(file.getFileId());
+        status.setFileName(file.getFileName());
+        status.setProjectId(file.getProjectId());
+        status.setVectorizationStatus(file.getVectorizationStatus());
+        status.setIsVectorized(file.getIsVectorized());
+        status.setVectorizationCount(file.getVectorizationCount());
+        status.setVectorizedTime(file.getVectorizedTime());
+        status.setFileExtension(file.getFileExtension());
+        status.setFileSize(file.getFileSize());
+        return status;
+    }
+
+    @GetMapping("/project/{project_id}/files")
+    @Operation(summary = "查询项目文件列表", description = "根据项目ID查询该项目下的所有文件信息")
+    @ApiResponses({
+            @ApiResponse(responseCode = "200", description = "查询成功"),
+            @ApiResponse(responseCode = "404", description = "项目不存在")
+    })
+    public ResponseEntity<ProjectFilesResponse> getProjectFiles(
+            @Parameter(description = "项目ID", required = true)
+            @PathVariable("project_id") String projectId) {
+
+        try {
+            // 检查项目是否存在
+            Project project = projectRepository.findByProjectId(projectId);
+            if (project == null) {
+                ProjectFilesResponse response = new ProjectFilesResponse();
+                response.setCode(404);
+                response.setMessage("项目不存在");
+                return ResponseEntity.ok(response);
+            }
+
+            ProjectFilesResponse.ProjectFilesData data = fileService.getProjectFiles(projectId);
+
+            ProjectFilesResponse response = new ProjectFilesResponse();
+            response.setCode(0);
+            response.setMessage("查询成功");
+            response.setData(data);
+
+            return ResponseEntity.ok(response);
+
+        } catch (Exception e) {
+            logger.error("查询项目文件失败: projectId={}", projectId, e);
+            ProjectFilesResponse response = new ProjectFilesResponse();
+            response.setCode(500);
+            response.setMessage("系统错误: " + e.getMessage());
+            return ResponseEntity.internalServerError().body(response);
+        }
+    }
+
+    /**
+     * 发送采购文件json信息接口 (ES平台调用)
+     */
+    @PostMapping("/upload")
+    @Operation(summary = "发送采购文件json信息接口", description = "ES平台发送采购文件JSON信息")
+    @ApiResponses(value = {
+        @ApiResponse(responseCode = "200", description = "成功",
+            content = @Content(schema = @Schema(implementation = FileUploadRequest.ApiResponse.class))),
+        @ApiResponse(responseCode = "400", description = "参数错误"),
+        @ApiResponse(responseCode = "401", description = "认证失败"),
+        @ApiResponse(responseCode = "500", description = "系统错误")
+    })
+    public ResponseEntity<FileUploadRequest.ApiResponse> uploadProcurementFiles(
+            @RequestHeader(value = "X-Server-Param", required = false) String xServerParam,
+            @RequestHeader(value = "X-CurTime", required = false) String xCurTime,
+            @RequestHeader(value = "X-CheckSum", required = false) String xCheckSum,
+            @RequestHeader(value = "traceId", required = false) String traceId,
+            @RequestBody FileUploadRequest.ResponseFileRequest request) {
+
+        logger.info("接收文件上传请求: traceId={}, projectId={}, packageId={}, supplierId={}, fileType={}",
+                   traceId, request.getProjectId(), request.getPackageId(), request.getSupplierId(), request.getFileType());
+
+        try {
+            // 验证认证信息
+            if (!apiAuthUtils.validateAuth(xServerParam, xCurTime, xCheckSum, "your-app-id")) {
+                logger.warn("文件上传认证失败: traceId={}", traceId);
+                return ResponseEntity.status(HttpStatus.UNAUTHORIZED)
+                    .body(new FileUploadRequest.ApiResponse("401", "认证失败"));
+            }
+
+            // 验证请求参数
+            if (request.getProjectId() == null || request.getPackageId() == null ||
+                request.getFileIds() == null || request.getFileIds().isEmpty()) {
+                return ResponseEntity.badRequest()
+                    .body(new FileUploadRequest.ApiResponse("400", "缺少必需参数"));
+            }
+
+            // 判断是采购文件还是供应商应答文件
+            if (request.getSupplierId() != null && !request.getSupplierId().isEmpty()) {
+                // 供应商应答文件逻辑
+                String fileType = request.getFileType();
+                
+                // 按照请求的fileType：1商务、2技术
+                // 只获取技术文件(2)，其他的只接受请求返回200
+                if ("1".equals(fileType)) {
+                    logger.info("接收到商务文件，忽略处理: traceId={}, supplierId={}", traceId, request.getSupplierId());
+                    return ResponseEntity.ok(new FileUploadRequest.ApiResponse("0", "成功"));
+                } else if ("2".equals(fileType)) {
+                    // 技术文件，进行下载和处理
+                    logger.info("接收到技术文件，开始处理: traceId={}, supplierId={}", traceId, request.getSupplierId());
+                    fileService.processResponseFilesAsync(request.getProjectId(), request.getPackageId(),
+                                                        request.getSupplierId(), request.getFileIds(),
+                                                        request.getFileType(), request.getJsonFileId(),
+                                                        request.getUnitId(), traceId);
+                } else {
+                    // 其他类型，暂时也忽略或按需处理，这里默认忽略
+                    logger.info("接收到其他类型文件({}), 忽略处理: traceId={}, supplierId={}", fileType, traceId, request.getSupplierId());
+                    return ResponseEntity.ok(new FileUploadRequest.ApiResponse("0", "成功"));
+                }
+            } else {
+                // 采购文件逻辑 (没有supplierId)
+                fileService.processProcurementFilesAsync(request.getProjectId(), request.getPackageId(),
+                                                       request.getFileIds(), traceId);
+            }
+
+            logger.info("文件上传请求处理完成: traceId={}", traceId);
+            return ResponseEntity.ok(new FileUploadRequest.ApiResponse("0", "成功"));
+
+        } catch (Exception e) {
+            logger.error("文件上传处理失败: traceId={}", traceId, e);
+            return ResponseEntity.internalServerError()
+                .body(new FileUploadRequest.ApiResponse("500", "系统错误: " + e.getMessage()));
+        }
+    }
+
+    /**
+     * 发送应答文件信息接口 (ES平台调用)
+     */
+    @PostMapping("/upload/response")
+    @Operation(summary = "发送应答文件信息接口", description = "ES平台发送供应商应答文件信息")
+    @ApiResponses(value = {
+        @ApiResponse(responseCode = "200", description = "成功",
+            content = @Content(schema = @Schema(implementation = FileUploadRequest.ApiResponse.class))),
+        @ApiResponse(responseCode = "400", description = "参数错误"),
+        @ApiResponse(responseCode = "401", description = "认证失败"),
+        @ApiResponse(responseCode = "500", description = "系统错误")
+    })
+    public ResponseEntity<FileUploadRequest.ApiResponse> uploadResponseFiles(
+            @RequestHeader(value = "X-Server-Param", required = false) String xServerParam,
+            @RequestHeader(value = "X-CurTime", required = false) String xCurTime,
+            @RequestHeader(value = "X-CheckSum", required = false) String xCheckSum,
+            @RequestHeader(value = "traceId", required = false) String traceId,
+            @RequestBody FileUploadRequest.ResponseFileRequest request) {
+
+        logger.info("接收应答文件上传请求: traceId={}, projectId={}, supplierId={}",
+                   traceId, request.getProjectId(), request.getSupplierId());
+
+        try {
+            // 验证认证信息
+            if (!apiAuthUtils.validateAuth(xServerParam, xCurTime, xCheckSum, "your-app-id")) {
+                logger.warn("应答文件上传认证失败: traceId={}", traceId);
+                return ResponseEntity.status(HttpStatus.UNAUTHORIZED)
+                    .body(new FileUploadRequest.ApiResponse("401", "认证失败"));
+            }
+
+            // 验证请求参数
+            if (request.getProjectId() == null || request.getPackageId() == null ||
+                request.getSupplierId() == null || request.getFileIds() == null ||
+                request.getFileIds().isEmpty() || request.getUnitId() == null) {
+                return ResponseEntity.badRequest()
+                    .body(new FileUploadRequest.ApiResponse("400", "缺少必需参数"));
+            }
+
+            // 异步下载和处理文件
+            fileService.processResponseFilesAsync(request.getProjectId(), request.getPackageId(),
+                                                request.getSupplierId(), request.getFileIds(),
+                                                request.getFileType(), request.getJsonFileId(),
+                                                request.getUnitId(), traceId);
+
+            logger.info("应答文件上传请求处理完成: traceId={}", traceId);
+            return ResponseEntity.ok(new FileUploadRequest.ApiResponse("0", "成功"));
+
+        } catch (Exception e) {
+            logger.error("应答文件上传处理失败: traceId={}", traceId, e);
+            return ResponseEntity.internalServerError()
+                .body(new FileUploadRequest.ApiResponse("500", "系统错误: " + e.getMessage()));
+        }
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/controller/IndexController.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/controller/IndexController.java
new file mode 100644
index 0000000..f7a5179
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/controller/IndexController.java
@@ -0,0 +1,61 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.controller
+ * @ClassName: IndexController
+ * @Description: 首页控制器，提供欢迎页面和健康检查页面的路由
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.controller;
+
+import org.springframework.beans.factory.annotation.Value;
+import org.springframework.http.ResponseEntity;
+import org.springframework.stereotype.Controller;
+import org.springframework.web.bind.annotation.GetMapping;
+import org.springframework.web.bind.annotation.RestController;
+
+import java.util.HashMap;
+import java.util.Map;
+
+/**
+ * 首页控制器
+ */
+@Controller
+public class IndexController {
+
+    /**
+     * 访问根路径时显示欢迎页面
+     */
+    @GetMapping("/")
+    public String index() {
+        // 重定向到静态页面
+        return "redirect:/view/index.html";
+    }
+
+    /**
+     * 健康检查页面
+     */
+    @GetMapping("/health")
+    public String health() {
+        // 返回健康检查HTML页面
+        return "redirect:/view/health.html";
+    }
+
+    /**
+     * 版本信息接口
+     */
+    @GetMapping("/version")
+    public ResponseEntity<Map<String, Object>> version() {
+        Map<String, Object> versionInfo = new HashMap<>();
+        versionInfo.put("project", "智慧评标-主观分助手");
+        versionInfo.put("version", System.getProperty("APP_VERSION", "1.0.0"));
+        versionInfo.put("build_time", System.getProperty("BUILD_TIME", "unknown"));
+        versionInfo.put("java_version", System.getProperty("java.version"));
+        versionInfo.put("spring_boot_version", "3.2.0");
+
+        return ResponseEntity.ok(versionInfo);
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/controller/OcrController.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/controller/OcrController.java
new file mode 100644
index 0000000..89aa34c
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/controller/OcrController.java
@@ -0,0 +1,213 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.controller
+ * @ClassName: OcrController
+ * @Description: OCR控制器，提供OCR结果查询和管理接口
+ * @Author: 张志锋
+ * @Date: 2025-12-10
+ * @Version: 1.0
+ * @Copyright: 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.controller;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.OcrResult;
+import com.zhpb.gdyd_zhpb_zgf.service.OcrService;
+import io.swagger.v3.oas.annotations.Operation;
+import io.swagger.v3.oas.annotations.Parameter;
+import io.swagger.v3.oas.annotations.responses.ApiResponse;
+import io.swagger.v3.oas.annotations.responses.ApiResponses;
+import io.swagger.v3.oas.annotations.tags.Tag;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.http.ResponseEntity;
+import org.springframework.web.bind.annotation.*;
+
+import java.util.HashMap;
+import java.util.List;
+import java.util.Map;
+import java.util.Optional;
+
+/**
+ * OCR控制器
+ */
+@RestController
+@RequestMapping("/api/v1/files/ocr")
+@Tag(name = "OCR管理", description = "OCR识别结果查询和管理接口")
+public class OcrController {
+
+    private static final Logger logger = LoggerFactory.getLogger(OcrController.class);
+
+    @Autowired
+    private OcrService ocrService;
+
+    /**
+     * 根据文件ID获取OCR结果
+     */
+    @GetMapping("/file/{fileId}")
+    @Operation(summary = "获取文件OCR结果", description = "根据文件ID获取OCR识别结果")
+    @ApiResponses(value = {
+        @ApiResponse(responseCode = "200", description = "查询成功"),
+        @ApiResponse(responseCode = "204", description = "OCR结果不存在"),
+        @ApiResponse(responseCode = "500", description = "系统错误")
+    })
+    public ResponseEntity<Map<String, Object>> getOcrResultByFileId(
+            @Parameter(description = "文件ID") @PathVariable Long fileId) {
+
+        try {
+            Optional<OcrResult> ocrResult = ocrService.getOcrResultByFileId(fileId);
+
+            if (ocrResult.isPresent()) {
+                Map<String, Object> result = new HashMap<>();
+                result.put("code", 200);
+                result.put("message", "查询成功");
+                result.put("data", ocrResult.get());
+                return ResponseEntity.ok(result);
+            } else {
+                Map<String, Object> result = new HashMap<>();
+                result.put("code", 404);
+                result.put("message", "OCR结果不存在");
+                return ResponseEntity.status(404).body(result);
+            }
+
+        } catch (Exception e) {
+            logger.error("查询OCR结果失败: fileId={}", fileId, e);
+            Map<String, Object> result = new HashMap<>();
+            result.put("code", 500);
+            result.put("message", "系统错误: " + e.getMessage());
+            return ResponseEntity.internalServerError().body(result);
+        }
+    }
+
+    /**
+     * 根据项目ID获取OCR结果列表
+     */
+    @GetMapping("/project/{projectId}")
+    @Operation(summary = "获取项目OCR结果列表", description = "根据项目ID获取所有OCR识别结果")
+    @ApiResponses(value = {
+        @ApiResponse(responseCode = "200", description = "查询成功"),
+        @ApiResponse(responseCode = "500", description = "系统错误")
+    })
+    public ResponseEntity<Map<String, Object>> getOcrResultsByProject(
+            @Parameter(description = "项目ID") @PathVariable String projectId,
+            @Parameter(description = "供应商ID（可选）") @RequestParam(required = false) String supplierId) {
+
+        try {
+            List<OcrResult> ocrResults;
+            if (supplierId != null && !supplierId.trim().isEmpty()) {
+                ocrResults = ocrService.getOcrResultsByProjectAndSupplier(projectId, supplierId);
+            } else {
+                ocrResults = ocrService.getOcrResultsByProject(projectId);
+            }
+
+            Map<String, Object> result = new HashMap<>();
+            result.put("code", 200);
+            result.put("message", "查询成功");
+            result.put("data", ocrResults);
+            result.put("total", ocrResults.size());
+
+            return ResponseEntity.ok(result);
+
+        } catch (Exception e) {
+            logger.error("查询项目OCR结果失败: projectId={}, supplierId={}", projectId, supplierId, e);
+            Map<String, Object> result = new HashMap<>();
+            result.put("code", 500);
+            result.put("message", "系统错误: " + e.getMessage());
+            return ResponseEntity.internalServerError().body(result);
+        }
+    }
+
+    /**
+     * 获取项目OCR统计信息
+     */
+    @GetMapping("/project/{projectId}/stats")
+    @Operation(summary = "获取项目OCR统计", description = "获取项目的OCR处理统计信息")
+    @ApiResponses(value = {
+        @ApiResponse(responseCode = "200", description = "查询成功"),
+        @ApiResponse(responseCode = "500", description = "系统错误")
+    })
+    public ResponseEntity<Map<String, Object>> getProjectOcrStatistics(
+            @Parameter(description = "项目ID") @PathVariable String projectId) {
+
+        try {
+            Map<String, Object> stats = ocrService.getProjectOcrStatistics(projectId);
+
+            Map<String, Object> result = new HashMap<>();
+            result.put("code", 200);
+            result.put("message", "查询成功");
+            result.put("data", stats);
+
+            return ResponseEntity.ok(result);
+
+        } catch (Exception e) {
+            logger.error("查询项目OCR统计失败: projectId={}", projectId, e);
+            Map<String, Object> result = new HashMap<>();
+            result.put("code", 500);
+            result.put("message", "系统错误: " + e.getMessage());
+            return ResponseEntity.internalServerError().body(result);
+        }
+    }
+
+    /**
+     * 重试失败的OCR任务
+     */
+    @PostMapping("/retry/{ocrResultId}")
+    @Operation(summary = "重试OCR任务", description = "重试失败的OCR识别任务")
+    @ApiResponses(value = {
+        @ApiResponse(responseCode = "200", description = "重试任务已启动"),
+        @ApiResponse(responseCode = "404", description = "OCR结果不存在"),
+        @ApiResponse(responseCode = "500", description = "系统错误")
+    })
+    public ResponseEntity<Map<String, Object>> retryOcrTask(
+            @Parameter(description = "OCR结果ID") @PathVariable Long ocrResultId) {
+
+        try {
+            // 这里可以先检查OCR结果是否存在和状态
+            // 但为了简单，直接调用重试方法
+            ocrService.retryFailedOcrTask(ocrResultId);
+
+            Map<String, Object> result = new HashMap<>();
+            result.put("code", 200);
+            result.put("message", "重试任务已启动");
+            return ResponseEntity.ok(result);
+
+        } catch (Exception e) {
+            logger.error("重试OCR任务失败: ocrResultId={}", ocrResultId, e);
+            Map<String, Object> result = new HashMap<>();
+            result.put("code", 500);
+            result.put("message", "系统错误: " + e.getMessage());
+            return ResponseEntity.internalServerError().body(result);
+        }
+    }
+
+    /**
+     * 获取失败的OCR任务列表
+     */
+    @GetMapping("/failed")
+    @Operation(summary = "获取失败的OCR任务", description = "获取所有失败的OCR任务列表")
+    @ApiResponses(value = {
+        @ApiResponse(responseCode = "200", description = "查询成功"),
+        @ApiResponse(responseCode = "500", description = "系统错误")
+    })
+    public ResponseEntity<Map<String, Object>> getFailedOcrTasks() {
+
+        try {
+            List<OcrResult> failedTasks = ocrService.getFailedOcrTasks();
+
+            Map<String, Object> result = new HashMap<>();
+            result.put("code", 200);
+            result.put("message", "查询成功");
+            result.put("data", failedTasks);
+            result.put("total", failedTasks.size());
+
+            return ResponseEntity.ok(result);
+
+        } catch (Exception e) {
+            logger.error("查询失败的OCR任务失败", e);
+            Map<String, Object> result = new HashMap<>();
+            result.put("code", 500);
+            result.put("message", "系统错误: " + e.getMessage());
+            return ResponseEntity.internalServerError().body(result);
+        }
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/controller/PromptTemplateController.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/controller/PromptTemplateController.java
new file mode 100644
index 0000000..680f0d2
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/controller/PromptTemplateController.java
@@ -0,0 +1,171 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.controller
+ * @ClassName: PromptTemplateController
+ * @Description: 提示词模板控制器
+ * @Author: 张志锋
+ * @Date: 2025-11-03
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.controller;
+
+import com.zhpb.gdyd_zhpb_zgf.dto.PromptTemplateDto;
+import com.zhpb.gdyd_zhpb_zgf.service.PromptTemplateService;
+import io.swagger.v3.oas.annotations.Operation;
+import io.swagger.v3.oas.annotations.Parameter;
+import io.swagger.v3.oas.annotations.responses.ApiResponse;
+import io.swagger.v3.oas.annotations.responses.ApiResponses;
+import io.swagger.v3.oas.annotations.tags.Tag;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.http.ResponseEntity;
+import org.springframework.validation.annotation.Validated;
+import org.springframework.web.bind.annotation.*;
+
+import java.util.HashMap;
+import java.util.List;
+import java.util.Map;
+
+/**
+ * 提示词模板控制器
+ */
+@RestController
+@RequestMapping("/api/v1/prompt-templates")
+@Tag(name = "提示词模板", description = "提示词模板管理接口")
+public class PromptTemplateController {
+
+    @Autowired
+    private PromptTemplateService promptTemplateService;
+
+    @GetMapping
+    @Operation(summary = "获取所有提示词模板", description = "获取所有提示词模板，可按类型过滤")
+    @ApiResponses({
+            @ApiResponse(responseCode = "200", description = "获取成功")
+    })
+    public ResponseEntity<List<PromptTemplateDto>> getAllTemplates(
+            @Parameter(description = "模板类型，可为空")
+            @RequestParam(required = false) String templateType) {
+
+        List<PromptTemplateDto> templates = promptTemplateService.getAllTemplates(templateType);
+        return ResponseEntity.ok(templates);
+    }
+
+    @GetMapping("/{templateId}")
+    @Operation(summary = "获取指定模板", description = "根据模板ID获取模板详情")
+    @ApiResponses({
+            @ApiResponse(responseCode = "200", description = "获取成功"),
+            @ApiResponse(responseCode = "404", description = "模板不存在")
+    })
+    public ResponseEntity<Map<String, Object>> getTemplateById(
+            @Parameter(description = "模板ID", required = true)
+            @PathVariable String templateId) {
+
+        PromptTemplateDto template = promptTemplateService.getTemplateById(templateId);
+        if (template == null) {
+            Map<String, Object> response = new HashMap<>();
+            response.put("code", 200);
+            response.put("message", "模板不存在");
+            return ResponseEntity.ok(response);
+        }
+        Map<String, Object> response = new HashMap<>();
+        response.put("code", 0);
+        response.put("message", "Success");
+        response.put("data", template);
+        return ResponseEntity.ok(response);
+    }
+
+    @PostMapping
+    @Operation(summary = "创建提示词模板", description = "创建新的提示词模板")
+    @ApiResponses({
+            @ApiResponse(responseCode = "200", description = "创建成功"),
+            @ApiResponse(responseCode = "400", description = "参数错误")
+    })
+    public ResponseEntity<PromptTemplateDto> createTemplate(
+            @Parameter(description = "模板信息", required = true)
+            @RequestBody @Validated PromptTemplateDto dto,
+            @Parameter(description = "操作用户ID")
+            @RequestParam(defaultValue = "admin") String userId) {
+
+        try {
+            PromptTemplateDto createdTemplate = promptTemplateService.createTemplate(dto, userId);
+            return ResponseEntity.ok(createdTemplate);
+        } catch (IllegalArgumentException e) {
+            return ResponseEntity.badRequest().build();
+        }
+    }
+
+    @PutMapping("/{templateId}")
+    @Operation(summary = "更新提示词模板", description = "更新指定模板的信息")
+    @ApiResponses({
+            @ApiResponse(responseCode = "200", description = "更新成功"),
+            @ApiResponse(responseCode = "404", description = "模板不存在"),
+            @ApiResponse(responseCode = "400", description = "参数错误")
+    })
+    public ResponseEntity<PromptTemplateDto> updateTemplate(
+            @Parameter(description = "模板ID", required = true)
+            @PathVariable String templateId,
+            @Parameter(description = "模板信息", required = true)
+            @RequestBody @Validated PromptTemplateDto dto,
+            @Parameter(description = "操作用户ID")
+            @RequestParam(defaultValue = "admin") String userId) {
+
+        try {
+            PromptTemplateDto updatedTemplate = promptTemplateService.updateTemplate(templateId, dto, userId);
+            return ResponseEntity.ok(updatedTemplate);
+        } catch (IllegalArgumentException e) {
+            return ResponseEntity.badRequest().build();
+        }
+    }
+
+    @DeleteMapping("/{templateId}")
+    @Operation(summary = "删除提示词模板", description = "删除指定的提示词模板")
+    @ApiResponses({
+            @ApiResponse(responseCode = "204", description = "删除成功"),
+            @ApiResponse(responseCode = "404", description = "模板不存在")
+    })
+    public ResponseEntity<Map<String, Object>> deleteTemplate(
+            @Parameter(description = "模板ID", required = true)
+            @PathVariable String templateId) {
+
+        try {
+            promptTemplateService.deleteTemplate(templateId);
+            Map<String, Object> response = new HashMap<>();
+            response.put("code", 0);
+            response.put("message", "删除成功");
+            return ResponseEntity.ok(response);
+        } catch (IllegalArgumentException e) {
+            Map<String, Object> response = new HashMap<>();
+            response.put("code", 200);
+            response.put("message", "模板不存在");
+            return ResponseEntity.ok(response);
+        }
+    }
+
+    @PutMapping("/{templateId}/status")
+    @Operation(summary = "切换模板状态", description = "启用或禁用指定的提示词模板")
+    @ApiResponses({
+            @ApiResponse(responseCode = "200", description = "操作成功"),
+            @ApiResponse(responseCode = "404", description = "模板不存在")
+    })
+    public ResponseEntity<Map<String, Object>> toggleTemplateStatus(
+            @Parameter(description = "模板ID", required = true)
+            @PathVariable String templateId,
+            @Parameter(description = "是否启用", required = true)
+            @RequestParam Boolean isActive,
+            @Parameter(description = "操作用户ID")
+            @RequestParam(defaultValue = "admin") String userId) {
+
+        try {
+            promptTemplateService.toggleTemplateStatus(templateId, isActive, userId);
+            Map<String, Object> response = new HashMap<>();
+            response.put("code", 0);
+            response.put("message", "操作成功");
+            return ResponseEntity.ok(response);
+        } catch (IllegalArgumentException e) {
+            Map<String, Object> response = new HashMap<>();
+            response.put("code", 200);
+            response.put("message", "模板不存在");
+            return ResponseEntity.ok(response);
+        }
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/controller/QaController.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/controller/QaController.java
new file mode 100644
index 0000000..67b7422
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/controller/QaController.java
@@ -0,0 +1,184 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.controller
+ * @ClassName: QaController
+ * @Description: 智能问答控制器
+ * @Author: 张志锋
+ * @Date: 2025-11-03
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.controller;
+
+import com.zhpb.gdyd_zhpb_zgf.dto.QaHistoryResponse;
+import com.zhpb.gdyd_zhpb_zgf.dto.QaRequest;
+import com.zhpb.gdyd_zhpb_zgf.dto.QaResponse;
+import com.zhpb.gdyd_zhpb_zgf.service.QaService;
+import io.swagger.v3.oas.annotations.Operation;
+import io.swagger.v3.oas.annotations.Parameter;
+import io.swagger.v3.oas.annotations.responses.ApiResponse;
+import io.swagger.v3.oas.annotations.responses.ApiResponses;
+import io.swagger.v3.oas.annotations.tags.Tag;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.http.ResponseEntity;
+import org.springframework.util.StringUtils;
+import org.springframework.validation.annotation.Validated;
+import org.springframework.web.bind.annotation.*;
+import org.springframework.web.context.request.RequestContextHolder;
+import org.springframework.web.context.request.ServletRequestAttributes;
+
+import jakarta.servlet.http.HttpServletRequest;
+import com.zhpb.gdyd_zhpb_zgf.service.UserLogService;
+
+/**
+ * 智能问答控制器
+ */
+@RestController
+@RequestMapping("/api/v1/qa")
+@Tag(name = "智能问答", description = "基于文档的智能问答接口")
+public class QaController {
+
+    private static final Logger logger = LoggerFactory.getLogger(QaController.class);
+
+    @Autowired
+    private QaService qaService;
+    
+    @Autowired
+    private UserLogService userLogService;
+
+    @PostMapping("/sendMessage")
+    @Operation(summary = "发送对话消息", description = "用户输入问题，针对一个或多个文件进行提问")
+    @ApiResponses({
+            @ApiResponse(responseCode = "200", description = "Successful Response"),
+            @ApiResponse(responseCode = "422", description = "Validation Error")
+    })
+    public ResponseEntity<QaResponse> sendMessage(
+            @Parameter(description = "智能问答请求")
+            @RequestBody @Validated QaRequest request) {
+
+        logger.info("收到智能问答请求: projectId={}, userId={}, query={}",
+                request.getProjectId(), request.getUserId(), request.getQuery());
+
+        try {
+            QaResponse.QaData data = qaService.sendMessage(request);
+
+            QaResponse response = new QaResponse();
+            response.setCode(0);
+            response.setMessage("查询成功");
+            response.setData(data);
+
+            // 记录用户操作日志
+            ServletRequestAttributes attributes = (ServletRequestAttributes) RequestContextHolder.currentRequestAttributes();
+            HttpServletRequest httpRequest = attributes.getRequest();
+            userLogService.logUserOperation(request.getUserId(), null, request.getProjectId(), 
+                "QA_QUESTION", "发送问答消息: " + request.getQuery(), null, httpRequest);
+
+            return ResponseEntity.ok(response);
+
+        } catch (Exception e) {
+            logger.error("智能问答处理失败", e);
+            QaResponse response = new QaResponse();
+            response.setCode(500);
+            response.setMessage("系统错误: " + e.getMessage());
+            response.setData(new QaResponse.QaData());
+            return ResponseEntity.internalServerError().body(response);
+        }
+    }
+
+    @GetMapping("/history/sessions")
+    @Operation(summary = "查询历史会话", description = "根据项目ID和用户ID查询历史会话列表")
+    @ApiResponses({
+            @ApiResponse(responseCode = "200", description = "查询成功"),
+            @ApiResponse(responseCode = "500", description = "系统错误")
+    })
+    public ResponseEntity<QaHistoryResponse> getSessionHistory(
+            @Parameter(description = "项目ID", required = true)
+            @RequestParam String projectId,
+            @Parameter(description = "用户ID", required = true)
+            @RequestParam String userId,
+            @Parameter(description = "用户名", required = false)
+            @RequestParam(required = false) String userName,
+            @Parameter(description = "用户名（兼容字段）", required = false)
+            @RequestParam(value = "user_name", required = false) String legacyUserName) {
+
+        logger.info("查询历史会话: projectId={}, userId={}", projectId, userId);
+
+        try {
+            QaHistoryResponse.SessionsData data = qaService.getSessionHistory(projectId, userId);
+
+            QaHistoryResponse response = new QaHistoryResponse();
+            response.setCode(0);
+            response.setMessage("Success");
+            response.setData(data);
+
+            // 记录用户操作日志
+            ServletRequestAttributes attributes = (ServletRequestAttributes) RequestContextHolder.currentRequestAttributes();
+            HttpServletRequest httpRequest = attributes.getRequest();
+            String resolvedUserName = StringUtils.hasText(userName) ? userName : legacyUserName;
+            if (!StringUtils.hasText(resolvedUserName) && httpRequest != null) {
+                resolvedUserName = httpRequest.getHeader("User-Name");
+                if (!StringUtils.hasText(resolvedUserName)) {
+                    resolvedUserName = httpRequest.getHeader("userName");
+                }
+            }
+            userLogService.logUserOperation(userId, resolvedUserName, projectId, 
+                "VIEW_QA_SESSIONS", "查询历史会话", null, httpRequest);
+
+            return ResponseEntity.ok(response);
+
+        } catch (Exception e) {
+            logger.error("查询历史会话失败", e);
+            QaHistoryResponse response = new QaHistoryResponse();
+            response.setCode(500);
+            response.setMessage("系统错误: " + e.getMessage());
+            return ResponseEntity.internalServerError().body(response);
+        }
+    }
+
+    @GetMapping("/history/messages")
+    @Operation(summary = "查询会话消息历史", description = "根据会话ID查询消息历史")
+    @ApiResponses({
+            @ApiResponse(responseCode = "200", description = "查询成功"),
+            @ApiResponse(responseCode = "404", description = "会话不存在"),
+            @ApiResponse(responseCode = "500", description = "系统错误")
+    })
+    public ResponseEntity<QaHistoryResponse> getMessageHistory(
+            @Parameter(description = "会话ID", required = true)
+            @RequestParam String sessionId) {
+
+        logger.info("查询会话消息历史: sessionId={}", sessionId);
+
+        try {
+            QaHistoryResponse.MessagesData data = qaService.getMessageHistory(sessionId);
+
+            if (data.getMessages().isEmpty()) {
+                QaHistoryResponse response = new QaHistoryResponse();
+                response.setCode(200);
+                response.setMessage("会话不存在或无消息记录");
+                return ResponseEntity.ok().body(response);
+            }
+
+            QaHistoryResponse response = new QaHistoryResponse();
+            response.setCode(0);
+            response.setMessage("Success");
+            response.setData(data);
+
+            // 记录用户操作日志
+            ServletRequestAttributes attributes = (ServletRequestAttributes) RequestContextHolder.currentRequestAttributes();
+            HttpServletRequest httpRequest = attributes.getRequest();
+            userLogService.logUserOperation(null, null, null, 
+                "VIEW_QA_MESSAGES", "查看问答历史消息", sessionId, httpRequest);
+
+            return ResponseEntity.ok(response);
+
+        } catch (Exception e) {
+            logger.error("查询会话消息历史失败", e);
+            QaHistoryResponse response = new QaHistoryResponse();
+            response.setCode(500);
+            response.setMessage("系统错误: " + e.getMessage());
+            return ResponseEntity.internalServerError().body(response);
+        }
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/controller/SsoController.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/controller/SsoController.java
new file mode 100644
index 0000000..3ce336f
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/controller/SsoController.java
@@ -0,0 +1,217 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.controller
+ * @ClassName: SsoController
+ * @Description: SSO单点登录控制器，通过ES接口验证ticket并生成JWT Token
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.controller;
+
+import com.alibaba.fastjson2.JSONObject;
+import com.zhpb.gdyd_zhpb_zgf.dto.SsoLoginResponse;
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.SsoLog;
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.User;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.SsoLogRepository;
+import com.zhpb.gdyd_zhpb_zgf.service.UserService;
+import com.zhpb.gdyd_zhpb_zgf.utils.HttpClientUtils;
+import com.zhpb.gdyd_zhpb_zgf.utils.JwtUtils;
+import io.swagger.v3.oas.annotations.Operation;
+import io.swagger.v3.oas.annotations.Parameter;
+import io.swagger.v3.oas.annotations.media.Content;
+import io.swagger.v3.oas.annotations.media.ExampleObject;
+import io.swagger.v3.oas.annotations.media.Schema;
+import io.swagger.v3.oas.annotations.responses.ApiResponse;
+import io.swagger.v3.oas.annotations.responses.ApiResponses;
+import io.swagger.v3.oas.annotations.tags.Tag;
+import jakarta.servlet.http.HttpServletRequest;
+import jakarta.servlet.http.HttpServletResponse;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.beans.factory.annotation.Value;
+import org.springframework.http.MediaType;
+import org.springframework.web.bind.annotation.*;
+
+import java.io.IOException;
+import java.net.URLEncoder;
+import java.nio.charset.StandardCharsets;
+import java.time.LocalDateTime;
+
+/**
+ * SSO单点登录控制器
+ */
+@RestController
+@RequestMapping("/api/v1/sso")
+@Tag(name = "单点登录", description = "SSO单点登录相关接口")
+public class SsoController {
+
+    @Autowired
+    private UserService userService;
+    
+    @Autowired
+    private SsoLogRepository ssoLogRepository;
+    
+    @Value("${sso.es-base-url:http://es-integration.es-uat-paas:8890}")
+    private String esBaseUrl;
+    
+    @Value("${sso.getUserInfoPath:/bi/getUserInfoByTicket}")
+    private String getUserInfoPath;
+
+    /**
+     * SSO登录接口
+     * 
+     * @param ticket SSO票据
+     * @param request HTTP请求对象
+     * @param response HTTP响应对象
+     * @throws IOException IO异常
+     */
+    @GetMapping("/login")
+    @Operation(
+            summary = "SSO单点登录",
+            description = "通过ES接口验证ticket，获取用户信息并生成JWT Token返回给前端。用户不存在则自动创建。"
+    )
+    @ApiResponses(value = {
+            @ApiResponse(
+                    responseCode = "200",
+                    description = "登录成功",
+                    content = @Content(
+                            mediaType = MediaType.APPLICATION_JSON_VALUE,
+                            schema = @Schema(implementation = SsoLoginResponse.class),
+                            examples = @ExampleObject(
+                                    name = "登录成功示例",
+                                    value = "{\n" +
+                                            "  \"code\": 200,\n" +
+                                            "  \"message\": \"登录成功\",\n" +
+                                            "  \"data\": {\n" +
+                                            "    \"token\": \"eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9...\",\n" +
+                                            "    \"userId\": \"n001\",\n" +
+                                            "    \"userName\": \"张三\",\n" +
+                                            "    \"projectId\": \"0001\"\n" +
+                                            "  }\n" +
+                                            "}"
+                            )
+                    )
+            ),
+            @ApiResponse(
+                    responseCode = "302",
+                    description = "Ticket验证失败，重定向到登录页",
+                    content = @Content
+            )
+    })
+    public void ssoLogin(
+            @Parameter(
+                    description = "SSO票据，从ES系统获取",
+                    required = true,
+                    example = "ticket_abc123xyz456"
+            )
+            @RequestParam String ticket,
+            HttpServletRequest request,
+            HttpServletResponse response) throws IOException {
+        
+        // 创建日志对象
+        SsoLog ssoLog = new SsoLog();
+        ssoLog.setTicket(ticket);
+        ssoLog.setLoginIp(getClientIp(request));
+        ssoLog.setLoginTime(LocalDateTime.now());
+        ssoLog.setCreateTime(LocalDateTime.now());
+        
+        try {
+            // 1. 调用ES接口验证ticket
+            String time = String.valueOf(System.currentTimeMillis());
+            String esUrl = esBaseUrl + getUserInfoPath + "?ticket=" + ticket + "&time=" + time;
+            ssoLog.setRequestUrl(esUrl);
+            
+            String result = HttpClientUtils.doGet(esUrl);
+            ssoLog.setResponseData(result);
+            
+            JSONObject res = JSONObject.parseObject(result);
+            
+            if (res.getInteger("ret") != 200) {
+                String errorMsg = res.getString("msg");
+                ssoLog.setLoginStatus("FAILED");
+                ssoLog.setErrorMsg("Ticket验证失败: " + errorMsg);
+                ssoLogRepository.save(ssoLog);
+                
+                // URL编码中文参数
+                String encodedError = URLEncoder.encode("ticket无效", StandardCharsets.UTF_8);
+                response.sendRedirect("/login?error=" + encodedError);
+                return;
+            }
+
+            // 2. 解析用户信息
+            JSONObject data = res.getJSONObject("data");
+            String userId = data.getString("id");
+            String name = data.getString("name");
+            String projectId = data.getString("projectid");
+            
+            ssoLog.setUserId(userId);
+            ssoLog.setUserName(name);
+            ssoLog.setProjectId(projectId);
+
+            // 3. 检查用户是否存在，不存在则创建
+            User user = userService.findByUserId(userId);
+            if (user == null) {
+                user = userService.createUser(userId, name, projectId);
+            } else {
+                // 更新项目ID（如果需要）
+                if (projectId != null && !projectId.equals(user.getProjectId())) {
+                    user.setProjectId(projectId);
+                    user.setUpdateTime(LocalDateTime.now());
+                    userService.updateUser(user);
+                }
+            }
+
+            // 4. 生成JWT Token
+            String token = JwtUtils.generateToken(userId, name);
+            ssoLog.setToken(token);
+            ssoLog.setLoginStatus("SUCCESS");
+
+            // 5. 保存登录日志
+            ssoLogRepository.save(ssoLog);
+
+            // 6. 将Token和用户信息返回给前端
+            JSONObject responseData = new JSONObject();
+            responseData.put("code", 200);
+            responseData.put("message", "登录成功");
+            JSONObject userData = new JSONObject();
+            userData.put("token", token);
+            userData.put("userId", userId);
+            userData.put("userName", name);
+            userData.put("projectId", projectId);
+            responseData.put("data", userData);
+            
+            // 设置响应头
+            response.setContentType("application/json;charset=UTF-8");
+            response.getWriter().write(responseData.toJSONString());
+            
+        } catch (Exception e) {
+            // 记录错误日志
+            ssoLog.setLoginStatus("ERROR");
+            ssoLog.setErrorMsg("登录异常: " + e.getMessage());
+            ssoLogRepository.save(ssoLog);
+            
+            // URL编码中文参数
+            String encodedError = URLEncoder.encode("登录失败", StandardCharsets.UTF_8);
+            response.sendRedirect("/login?error=" + encodedError);
+        }
+    }
+    
+    /**
+     * 获取客户端IP地址
+     */
+    private String getClientIp(HttpServletRequest request) {
+        String ip = request.getHeader("X-Forwarded-For");
+        if (ip == null || ip.isEmpty() || "unknown".equalsIgnoreCase(ip)) {
+            ip = request.getHeader("Proxy-Client-IP");
+        }
+        if (ip == null || ip.isEmpty() || "unknown".equalsIgnoreCase(ip)) {
+            ip = request.getHeader("WL-Proxy-Client-IP");
+        }
+        if (ip == null || ip.isEmpty() || "unknown".equalsIgnoreCase(ip)) {
+            ip = request.getRemoteAddr();
+        }
+        return ip;
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/controller/StatisticsController.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/controller/StatisticsController.java
new file mode 100644
index 0000000..12b0edb
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/controller/StatisticsController.java
@@ -0,0 +1,69 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.controller
+ * @ClassName: StatisticsController
+ * @Description: API访问统计控制器，提供API调用次数统计查询接口
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.controller;
+
+import com.zhpb.gdyd_zhpb_zgf.service.ApiStatisticsService;
+import io.swagger.v3.oas.annotations.Operation;
+import io.swagger.v3.oas.annotations.tags.Tag;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.web.bind.annotation.GetMapping;
+import org.springframework.web.bind.annotation.RequestMapping;
+import org.springframework.web.bind.annotation.RestController;
+
+import java.util.HashMap;
+import java.util.Map;
+
+/**
+ * 统计数据控制器
+ */
+@RestController
+@RequestMapping("/api/v1/statistics")
+@Tag(name = "统计信息", description = "API访问统计相关接口")
+public class StatisticsController {
+
+    @Autowired
+    private ApiStatisticsService apiStatisticsService;
+
+    /**
+     * 获取API访问总次数
+     *
+     * @return 总访问次数
+     */
+    @GetMapping("/total")
+    @Operation(summary = "获取API总访问次数", description = "返回系统中所有API接口的总访问次数")
+    public Map<String, Object> getTotalCount() {
+        Map<String, Object> result = new HashMap<>();
+        result.put("code", 200);
+        result.put("message", "获取成功");
+        result.put("data", apiStatisticsService.getTotalApiCount());
+        return result;
+    }
+
+    /**
+     * 获取特定接口的访问次数
+     *
+     * @param path 接口路径
+     * @return 访问次数
+     */
+    @GetMapping("/count")
+    @Operation(summary = "获取特定接口访问次数", description = "根据接口路径返回该接口的访问次数")
+    public Map<String, Object> getApiCount(String path) {
+        Map<String, Object> result = new HashMap<>();
+        result.put("code", 200);
+        result.put("message", "获取成功");
+        Map<String, Object> data = new HashMap<>();
+        data.put("path", path);
+        data.put("count", apiStatisticsService.getApiCount(path));
+        result.put("data", data);
+        return result;
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/controller/SummaryController.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/controller/SummaryController.java
new file mode 100644
index 0000000..1c5105b
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/controller/SummaryController.java
@@ -0,0 +1,673 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.controller
+ * @ClassName: SummaryController
+ * @Description: 总结相关控制器
+ * @Author: 张志锋
+ * @Date: 2025-11-03
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.controller;
+
+import com.zhpb.gdyd_zhpb_zgf.context.ModelRequestContextHolder;
+import com.zhpb.gdyd_zhpb_zgf.dto.SummaryResponse;
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.FileDirectory;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.FileDirectoryRepository;
+import com.zhpb.gdyd_zhpb_zgf.service.AIModelService;
+import com.zhpb.gdyd_zhpb_zgf.service.ChapterContentResolver;
+import com.zhpb.gdyd_zhpb_zgf.service.FilePathResolver;
+import com.zhpb.gdyd_zhpb_zgf.service.PromptTemplateService;
+import com.zhpb.gdyd_zhpb_zgf.utils.HttpClientUtils;
+import com.alibaba.fastjson2.JSON;
+import org.springframework.http.ResponseEntity;
+
+import java.util.*;
+import java.util.stream.Collectors;
+import io.swagger.v3.oas.annotations.Operation;
+import io.swagger.v3.oas.annotations.Parameter;
+import io.swagger.v3.oas.annotations.responses.ApiResponse;
+import io.swagger.v3.oas.annotations.responses.ApiResponses;
+import io.swagger.v3.oas.annotations.tags.Tag;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.web.bind.annotation.*;
+
+import jakarta.servlet.http.HttpServletRequest;
+import com.zhpb.gdyd_zhpb_zgf.service.UserLogService;
+
+import org.springframework.web.context.request.ServletRequestAttributes;
+
+import java.util.ArrayList;
+import java.util.List;
+
+/**
+ * 总结相关控制器
+ */
+@RestController
+@RequestMapping("/api/v1/summary")
+@Tag(name = "总结管理", description = "文件目录和章节总结相关接口")
+public class SummaryController {
+
+    private static final Logger logger = LoggerFactory.getLogger(SummaryController.class);
+
+    @Autowired
+    private FileDirectoryRepository fileDirectoryRepository;
+
+    @Autowired
+    private com.zhpb.gdyd_zhpb_zgf.repository.mysql.FileRepository fileRepository;
+
+    @Autowired
+    private com.zhpb.gdyd_zhpb_zgf.service.FileService fileService;
+
+    @Autowired
+    private PromptTemplateService promptTemplateService;
+
+    @Autowired
+    private AIModelService aiModelService;
+
+    @Autowired
+    private ChapterContentResolver chapterContentResolver;
+
+    @Autowired
+    private org.springframework.core.env.Environment environment;
+
+    @Autowired
+    private FilePathResolver filePathResolver;
+    
+    @Autowired
+    private UserLogService userLogService;
+
+    @GetMapping("/directory/count")
+    @Operation(summary = "查询t_file_directory表记录数")
+    public ResponseEntity<?> getDirectoryCount() {
+        logger.info("查询t_file_directory表记录数");
+        try {
+            long count = fileDirectoryRepository.count();
+            logger.info("t_file_directory表总记录数: {}", count);
+
+            Map<String, Object> response = new HashMap<>();
+            response.put("table", "t_file_directory");
+            response.put("totalRecords", count);
+            response.put("timestamp", java.time.LocalDateTime.now());
+
+            return ResponseEntity.ok(response);
+        } catch (Exception e) {
+            logger.error("查询表记录数失败", e);
+            Map<String, Object> errorResponse = new HashMap<>();
+            errorResponse.put("status", "error");
+            errorResponse.put("message", e.getMessage());
+            return ResponseEntity.status(500).body(errorResponse);
+        }
+    }
+
+    @GetMapping("/directory/status")
+    @Operation(summary = "检查文件目录状态", description = "检查文件是否存在以及是否有目录数据")
+    @ApiResponses({
+            @ApiResponse(responseCode = "200", description = "查询成功")
+    })
+    public ResponseEntity<Map<String, Object>> getDirectoryStatus(
+            @Parameter(description = "文件ID", required = true)
+            @RequestParam String file_id) {
+
+        logger.info("检查文件目录状态: fileId={}", file_id);
+
+        Map<String, Object> result = new HashMap<>();
+
+        try {
+            // 检查文件是否在t_file表中存在
+            java.util.Optional<com.zhpb.gdyd_zhpb_zgf.entity.mysql.File> fileOptional = fileRepository.findByFileId(file_id);
+            boolean fileExists = fileOptional.isPresent();
+
+            // 检查目录数据是否存在
+            long directoryCount = fileDirectoryRepository.countByFileId(file_id);
+
+            result.put("fileId", file_id);
+            result.put("fileExists", fileExists);
+            result.put("directoryExists", directoryCount > 0);
+            result.put("directoryCount", directoryCount);
+
+            if (fileExists) {
+                com.zhpb.gdyd_zhpb_zgf.entity.mysql.File file = fileOptional.get();
+                result.put("fileName", file.getFileName());
+                result.put("fileSize", file.getFileSize());
+                result.put("uploadTime", file.getCreateTime());
+                result.put("vectorizationStatus", file.getVectorizationStatus());
+            }
+
+            result.put("code", 0);
+            result.put("message", "查询成功");
+
+        } catch (Exception e) {
+            logger.error("检查文件目录状态失败", e);
+            result.put("code", 500);
+            result.put("message", "系统错误: " + e.getMessage());
+        }
+
+        return ResponseEntity.ok(result);
+    }
+
+    @GetMapping("/directory")
+    @Operation(summary = "获取文件目录结构", description = "根据文件ID获取文件的目录结构")
+    @ApiResponses({
+            @ApiResponse(responseCode = "200", description = "查询成功"),
+            @ApiResponse(responseCode = "404", description = "文件不存在"),
+            @ApiResponse(responseCode = "500", description = "系统错误")
+    })
+    public ResponseEntity<SummaryResponse.DirectoryResponse> getDirectory(
+            @Parameter(description = "文件ID", required = true)
+            @RequestParam String file_id) {
+
+        logger.info("查询文件目录: fileId={}", file_id);
+
+        try {
+            List<FileDirectory> directories = fileDirectoryRepository.findByFileIdOrderBySortOrderAsc(file_id);
+
+            if (directories.isEmpty()) {
+                SummaryResponse.DirectoryResponse response = new SummaryResponse.DirectoryResponse();
+                response.setCode(200);
+                response.setMessage("文件目录不存在，可能文件还没有进行目录抽取处理");
+                return ResponseEntity.ok().body(response);
+            }
+
+            // 构建树形结构
+            List<SummaryResponse.DirectoryItem> directoryItems = buildDirectoryTree(directories);
+
+            SummaryResponse.DirectoryData data = new SummaryResponse.DirectoryData();
+            data.setFileId(file_id);
+            data.setDirectory(directoryItems);
+
+            SummaryResponse.DirectoryResponse response = new SummaryResponse.DirectoryResponse();
+            response.setCode(0);
+            response.setMessage("Success");
+            response.setData(data);
+
+            // 记录用户操作日志
+            HttpServletRequest httpRequest = ((org.springframework.web.context.request.ServletRequestAttributes) org.springframework.web.context.request.RequestContextHolder.currentRequestAttributes())
+                .getRequest();
+            userLogService.logUserOperation(null, null, null, 
+                "VIEW_CATALOG", "查看文件目录结构", file_id, httpRequest);
+
+            return ResponseEntity.ok(response);
+
+        } catch (Exception e) {
+            logger.error("查询文件目录失败", e);
+            SummaryResponse.DirectoryResponse response = new SummaryResponse.DirectoryResponse();
+            response.setCode(500);
+            response.setMessage("系统错误: " + e.getMessage());
+            return ResponseEntity.internalServerError().body(response);
+        }
+    }
+
+    @GetMapping("/chapter")
+    @Operation(summary = "获取章节总结", description = "根据文件ID和章节ID获取章节总结")
+    @ApiResponses({
+            @ApiResponse(responseCode = "200", description = "查询成功"),
+            @ApiResponse(responseCode = "404", description = "章节不存在"),
+            @ApiResponse(responseCode = "500", description = "系统错误")
+    })
+    public ResponseEntity<SummaryResponse.ChapterSummaryResponse> getChapterSummary(
+            @Parameter(description = "文件ID", required = true)
+            @RequestParam String file_id,
+            @Parameter(description = "章节ID", required = true)
+            @RequestParam String chapter_id) {
+
+        logger.info("查询章节总结: fileId={}, chapterId={}", file_id, chapter_id);
+
+        try {
+            FileDirectory directory = fileDirectoryRepository.findByFileIdAndChapterId(file_id, chapter_id);
+
+            if (directory == null) {
+                SummaryResponse.ChapterSummaryResponse response = new SummaryResponse.ChapterSummaryResponse();
+                response.setCode(200);
+                response.setMessage("章节不存在");
+                return ResponseEntity.ok().body(response);
+            }
+
+            // 如果数据库中有总结内容，直接返回
+            if (directory.getSummary() != null && !directory.getSummary().isEmpty()) {
+                SummaryResponse.ChapterSummaryData data = new SummaryResponse.ChapterSummaryData();
+                data.setFileId(file_id);
+                data.setChapterId(chapter_id);
+                data.setSummary(directory.getSummary());
+
+                // 构建sources信息（如果需要的话）
+                data.setSources(new ArrayList<>());
+
+                SummaryResponse.ChapterSummaryResponse response = new SummaryResponse.ChapterSummaryResponse();
+                response.setCode(0);
+                response.setMessage("Success");
+                response.setData(data);
+
+                // 记录用户操作日志
+                HttpServletRequest httpRequest = ((org.springframework.web.context.request.ServletRequestAttributes) org.springframework.web.context.request.RequestContextHolder.currentRequestAttributes())
+                    .getRequest();
+                userLogService.logUserOperation(null, null, null, 
+                    "CATALOG_SUMMARY", "查看章节总结", chapter_id, httpRequest);
+
+                return ResponseEntity.ok(response);
+            }
+
+            // 如果数据库中没有总结内容，实时调用模型生成（3分钟超时）
+            try {
+                logger.info("数据库中无总结内容，实时调用模型生成: fileId={}, chapterId={}", file_id, chapter_id);
+
+                // 使用CompletableFuture实现3分钟超时控制
+                final FileDirectory finalDirectory = directory;
+                java.util.concurrent.CompletableFuture<String> summaryFuture = java.util.concurrent.CompletableFuture.supplyAsync(() -> {
+                    return generateChapterSummaryWithDeepSeek(finalDirectory);
+                });
+
+                // 等待最多3分钟
+                String summaryContent = null;
+                try {
+                    summaryContent = summaryFuture.get(3, java.util.concurrent.TimeUnit.MINUTES);
+                } catch (java.util.concurrent.TimeoutException te) {
+                    logger.warn("实时生成摘要超时(3分钟): fileId={}, chapterId={}", file_id, chapter_id);
+                    summaryFuture.cancel(true);
+                }
+
+                if (summaryContent != null && !summaryContent.isEmpty()) {
+                    // 更新数据库
+                    directory.setSummary(summaryContent);
+                    fileDirectoryRepository.save(directory);
+                    logger.info("实时生成摘要成功并保存: fileId={}, chapterId={}, 长度={}", 
+                        file_id, chapter_id, summaryContent.length());
+
+                    SummaryResponse.ChapterSummaryData data = new SummaryResponse.ChapterSummaryData();
+                    data.setFileId(file_id);
+                    data.setChapterId(chapter_id);
+                    data.setSummary(summaryContent);
+                    data.setSources(new ArrayList<>());
+
+                    SummaryResponse.ChapterSummaryResponse response = new SummaryResponse.ChapterSummaryResponse();
+                    response.setCode(0);
+                    response.setMessage("Success");
+                    response.setData(data);
+
+                    return ResponseEntity.ok(response);
+                }
+            } catch (Exception e) {
+                logger.warn("实时生成章节总结失败: fileId={}, chapterId={}, error={}", file_id, chapter_id, e.getMessage());
+            }
+
+            // 如果都获取不到，返回默认消息
+            SummaryResponse.ChapterSummaryData data = new SummaryResponse.ChapterSummaryData();
+            data.setFileId(file_id);
+            data.setChapterId(chapter_id);
+            data.setSummary("暂无总结内容");
+            data.setSources(new ArrayList<>());
+
+            SummaryResponse.ChapterSummaryResponse response = new SummaryResponse.ChapterSummaryResponse();
+            response.setCode(0);
+            response.setMessage("Success");
+            response.setData(data);
+
+            return ResponseEntity.ok(response);
+
+        } catch (Exception e) {
+            logger.error("查询章节总结失败", e);
+            SummaryResponse.ChapterSummaryResponse response = new SummaryResponse.ChapterSummaryResponse();
+            response.setCode(500);
+            response.setMessage("系统错误: " + e.getMessage());
+            return ResponseEntity.internalServerError().body(response);
+        }
+    }
+
+    /**
+     * 从外部API获取章节总结（这里使用DeepSeek API）
+     */
+    private String fetchChapterSummaryFromExternalAPI(String fileId, String chapterId) {
+        try {
+            // 获取章节信息
+            com.zhpb.gdyd_zhpb_zgf.entity.mysql.FileDirectory directory = fileDirectoryRepository.findByFileIdAndChapterId(fileId, chapterId);
+            if (directory == null) {
+                logger.warn("未找到章节信息: fileId={}, chapterId={}", fileId, chapterId);
+                return null;
+            }
+
+            // 使用DeepSeek API生成章节总结
+            return generateChapterSummaryWithDeepSeek(directory);
+
+        } catch (Exception e) {
+            logger.error("获取章节总结失败", e);
+            return null;
+        }
+    }
+
+    /**
+     * 使用AI模型生成章节总结
+     */
+    private String generateChapterSummaryWithDeepSeek(com.zhpb.gdyd_zhpb_zgf.entity.mysql.FileDirectory directory) {
+        try {
+            // 获取文档摘要提示词模板
+            com.zhpb.gdyd_zhpb_zgf.entity.mysql.PromptTemplate summaryTemplate = promptTemplateService.getActiveTemplate("summary");
+            if (summaryTemplate == null) {
+                throw new RuntimeException("未找到激活的文档摘要提示词模板，请在提示词配置页面启用或创建summary类型的提示词模板");
+            }
+
+            String promptTemplate = summaryTemplate.getPromptContent();
+            logger.info("使用数据库提示词模板: {} (ID: {}), 内容长度: {}", summaryTemplate.getTemplateName(), summaryTemplate.getTemplateId(), promptTemplate.length());
+            logger.debug("提示词内容: {}", promptTemplate);
+
+            // 使用模板替换变量
+            String prompt = promptTemplate.replace("{title}", directory.getTitle());
+            logger.debug("最终使用的提示词: {}", prompt);
+
+            logger.info("调用AI模型生成章节总结: chapterId={}", directory.getChapterId());
+
+            String userMessage = chapterContentResolver.buildUserMessage(directory);
+
+            ModelRequestContextHolder.setRequestSource("DOC_SUMMARY_MANUAL");
+            String summary;
+            try {
+                summary = aiModelService.generateResponse(prompt, userMessage);
+            } finally {
+                ModelRequestContextHolder.clear();
+            }
+
+            if (summary != null && !summary.trim().isEmpty()) {
+                logger.info("AI模型返回章节总结，长度: {}", summary.length());
+                return summary.trim();
+            }
+
+            return null;
+
+        } catch (Exception e) {
+            logger.error("调用AI模型生成章节总结失败: chapterId={}", directory.getChapterId(), e);
+            return null;
+        }
+    }
+
+    /**
+     * 获取配置文件的值
+     */
+    private String getConfigValue(String key, String defaultValue) {
+        try {
+            String value = environment.getProperty(key);
+            return value != null ? value : defaultValue;
+        } catch (Exception e) {
+            logger.warn("获取配置值失败: key={}, 使用默认值: {}", key, defaultValue);
+            return defaultValue;
+        }
+    }
+
+    @PostMapping("/directory/process")
+    @Operation(summary = "手动触发目录抽取", description = "为指定文件手动触发目录抽取和总结处理")
+    @ApiResponses({
+            @ApiResponse(responseCode = "200", description = "处理开始"),
+            @ApiResponse(responseCode = "404", description = "文件不存在"),
+            @ApiResponse(responseCode = "500", description = "系统错误")
+    })
+    public ResponseEntity<Map<String, Object>> processDirectory(
+            @Parameter(description = "文件ID", required = true)
+            @RequestParam String file_id) {
+
+        logger.info("手动触发目录抽取: fileId={}", file_id);
+
+        Map<String, Object> result = new HashMap<>();
+
+        try {
+            // 检查文件是否存在
+            java.util.Optional<com.zhpb.gdyd_zhpb_zgf.entity.mysql.File> fileOptional = fileRepository.findByFileId(file_id);
+            if (!fileOptional.isPresent()) {
+                result.put("code", 200);
+                result.put("message", "文件不存在");
+                return ResponseEntity.ok().body(result);
+            }
+
+            com.zhpb.gdyd_zhpb_zgf.entity.mysql.File file = fileOptional.get();
+
+            // 获取文件的完整路径
+            String filePath = constructFilePath(file);
+            if (filePath == null) {
+                result.put("code", 200);
+                result.put("message", "文件路径不存在，无法找到文件：" + file.getFileName() +
+                    "（fileId: " + file.getFileId() + "）");
+                result.put("fileId", file_id);
+                result.put("fileName", file.getFileName());
+                result.put("storedPath", file.getFilePath());
+                return ResponseEntity.ok().body(result);
+            }
+
+            // 异步处理文件目录
+            try {
+                java.lang.reflect.Method method = fileService.getClass().getDeclaredMethod("processFileDirectoriesAsync",
+                    java.util.List.class, String.class);
+                method.setAccessible(true);
+
+                java.util.List<String> filePaths = java.util.Arrays.asList(filePath);
+                method.invoke(fileService, filePaths, file.getProjectId());
+
+                result.put("code", 0);
+                result.put("message", "目录抽取任务已启动，请稍后查询");
+                result.put("fileId", file_id);
+                result.put("fileName", file.getFileName());
+
+            } catch (Exception e) {
+                logger.error("启动目录抽取任务失败", e);
+                result.put("code", 500);
+                result.put("message", "启动任务失败: " + e.getMessage());
+                return ResponseEntity.internalServerError().body(result);
+            }
+
+        } catch (Exception e) {
+            logger.error("手动触发目录抽取失败", e);
+            result.put("code", 500);
+            result.put("message", "系统错误: " + e.getMessage());
+            return ResponseEntity.internalServerError().body(result);
+        }
+
+        return ResponseEntity.ok(result);
+    }
+
+    /**
+     * 根据文件信息构造文件路径
+     */
+    private String constructFilePath(com.zhpb.gdyd_zhpb_zgf.entity.mysql.File file) {
+        try {
+            java.util.Optional<java.nio.file.Path> resolvedOpt = filePathResolver.resolveExistingPath(
+                file.getFilePath(), file.getProjectId(), file.getFileName());
+            if (resolvedOpt.isPresent()) {
+                java.nio.file.Path resolvedPath = resolvedOpt.get();
+                String normalized = filePathResolver.normalizeForStorage(resolvedPath);
+                if (normalized != null && !normalized.equals(file.getFilePath())) {
+                    file.setFilePath(normalized);
+                    fileRepository.save(file);
+                    logger.info("已自动纠正文件路径: {} -> {}", file.getFileId(), normalized);
+                }
+                return resolvedPath.toString();
+            }
+            logger.warn("无法解析文件路径: fileId={}, storedPath={}", file.getFileId(), file.getFilePath());
+            return null;
+
+        } catch (Exception e) {
+            logger.error("构造文件路径失败: fileId={}", file.getFileId(), e);
+        }
+        return null;
+    }
+
+    /**
+     * 获取文件扩展名
+     */
+    private String getFileExtension(String fileName) {
+        if (fileName == null) return "";
+        int lastDotIndex = fileName.lastIndexOf('.');
+        return lastDotIndex > 0 ? fileName.substring(lastDotIndex) : "";
+    }
+
+    /**
+     * 构建目录树形结构
+     */
+    private List<SummaryResponse.DirectoryItem> buildDirectoryTree(List<FileDirectory> directories) {
+        List<SummaryResponse.DirectoryItem> result = new ArrayList<>();
+
+        // 获取顶级目录（没有父章节的）
+        List<FileDirectory> rootDirectories = directories.stream()
+            .filter(dir -> dir.getParentChapterId() == null || dir.getParentChapterId().isEmpty())
+            .toList();
+
+        for (FileDirectory rootDir : rootDirectories) {
+            SummaryResponse.DirectoryItem item = convertToDirectoryItem(rootDir);
+            item.setChildren(buildChildren(directories, rootDir.getChapterId()));
+            result.add(item);
+        }
+
+        return result;
+    }
+
+    /**
+     * 递归构建子目录
+     */
+    private List<SummaryResponse.DirectoryItem> buildChildren(List<FileDirectory> allDirectories, String parentChapterId) {
+        List<SummaryResponse.DirectoryItem> children = new ArrayList<>();
+
+        List<FileDirectory> childDirectories = allDirectories.stream()
+            .filter(dir -> parentChapterId.equals(dir.getParentChapterId()))
+            .toList();
+
+        for (FileDirectory childDir : childDirectories) {
+            SummaryResponse.DirectoryItem item = convertToDirectoryItem(childDir);
+            item.setChildren(buildChildren(allDirectories, childDir.getChapterId()));
+            children.add(item);
+        }
+
+        return children;
+    }
+
+    /**
+     * 转换FileDirectory为DirectoryItem
+     */
+    private SummaryResponse.DirectoryItem convertToDirectoryItem(FileDirectory directory) {
+        SummaryResponse.DirectoryItem item = new SummaryResponse.DirectoryItem();
+        item.setChapterId(directory.getChapterId());
+        item.setTitle(directory.getTitle());
+        item.setPage(directory.getPage());
+        item.setStartPage(directory.getStartPage());
+        item.setEndPage(directory.getEndPage());
+        item.setChildren(new ArrayList<>());
+        return item;
+    }
+
+    @Autowired
+    private com.zhpb.gdyd_zhpb_zgf.service.CatalogExtractionRetryService catalogExtractionRetryService;
+
+    @PostMapping("/retry/failed")
+    @Operation(summary = "重试所有失败任务", description = "重新处理所有目录抽取失败和向量化失败的文件")
+    @ApiResponses({
+            @ApiResponse(responseCode = "200", description = "重试任务已启动"),
+            @ApiResponse(responseCode = "500", description = "系统错误")
+    })
+    public ResponseEntity<Map<String, Object>> retryAllFailedTasks(
+            @Parameter(description = "项目ID（可选，不传则处理所有项目）", required = false)
+            @RequestParam(value = "project_id", required = false) String projectId) {
+
+        logger.info("重试所有失败任务: projectId={}", projectId);
+
+        Map<String, Object> result = new HashMap<>();
+
+        try {
+            int catalogRetryCount = 0;
+            int vectorRetryCount = 0;
+
+            // 1. 重试目录抽取失败的文件
+            java.util.List<com.zhpb.gdyd_zhpb_zgf.entity.mysql.File> catalogFailedFiles;
+            if (projectId != null && !projectId.isEmpty()) {
+                catalogFailedFiles = fileRepository.findByProjectIdAndCatalogExtractionStatus(projectId, "failed");
+            } else {
+                catalogFailedFiles = fileRepository.findByCatalogExtractionStatus("failed");
+            }
+
+            for (com.zhpb.gdyd_zhpb_zgf.entity.mysql.File file : catalogFailedFiles) {
+                file.setCatalogExtractionStatus("not_started");
+                file.setCatalogExtractionRetryCount(0);
+                fileRepository.save(file);
+                catalogRetryCount++;
+            }
+
+            // 2. 重试向量化失败的文件
+            java.util.List<com.zhpb.gdyd_zhpb_zgf.entity.mysql.File> vectorFailedFiles;
+            if (projectId != null && !projectId.isEmpty()) {
+                vectorFailedFiles = fileRepository.findByProjectIdAndVectorizationStatus(projectId, "failed");
+            } else {
+                vectorFailedFiles = fileRepository.findByVectorizationStatus("failed");
+            }
+
+            for (com.zhpb.gdyd_zhpb_zgf.entity.mysql.File file : vectorFailedFiles) {
+                file.setVectorizationStatus("not_started");
+                file.setIsVectorized(false);
+                fileRepository.save(file);
+                vectorRetryCount++;
+            }
+
+            // 3. 处理卡住的任务（processing 状态超过30分钟的）
+            catalogExtractionRetryService.scanAndRetryStuckCatalogExtraction();
+
+            result.put("code", 0);
+            result.put("message", "重试任务已启动");
+            result.put("catalogRetryCount", catalogRetryCount);
+            result.put("vectorRetryCount", vectorRetryCount);
+            result.put("projectId", projectId);
+
+            logger.info("重试失败任务完成: 目录抽取重置{}个, 向量化重置{}个", catalogRetryCount, vectorRetryCount);
+
+        } catch (Exception e) {
+            logger.error("重试失败任务出错", e);
+            result.put("code", 500);
+            result.put("message", "系统错误: " + e.getMessage());
+            return ResponseEntity.internalServerError().body(result);
+        }
+
+        return ResponseEntity.ok(result);
+    }
+
+    @GetMapping("/status/failed")
+    @Operation(summary = "查询失败任务数量", description = "查询目录抽取失败和向量化失败的文件数量")
+    @ApiResponses({
+            @ApiResponse(responseCode = "200", description = "查询成功")
+    })
+    public ResponseEntity<Map<String, Object>> getFailedTasksStatus(
+            @Parameter(description = "项目ID（可选，不传则查询所有项目）", required = false)
+            @RequestParam(value = "project_id", required = false) String projectId) {
+
+        logger.info("查询失败任务状态: projectId={}", projectId);
+
+        Map<String, Object> result = new HashMap<>();
+
+        try {
+            long catalogFailedCount;
+            long catalogProcessingCount;
+            long vectorFailedCount;
+            long vectorProcessingCount;
+
+            if (projectId != null && !projectId.isEmpty()) {
+                catalogFailedCount = fileRepository.countByProjectIdAndCatalogExtractionStatus(projectId, "failed");
+                catalogProcessingCount = fileRepository.countByProjectIdAndCatalogExtractionStatus(projectId, "processing");
+                vectorFailedCount = fileRepository.countByProjectIdAndVectorizationStatus(projectId, "failed");
+                vectorProcessingCount = fileRepository.countByProjectIdAndVectorizationStatus(projectId, "processing");
+            } else {
+                catalogFailedCount = fileRepository.countByCatalogExtractionStatus("failed");
+                catalogProcessingCount = fileRepository.countByCatalogExtractionStatus("processing");
+                vectorFailedCount = fileRepository.countByVectorizationStatus("failed");
+                vectorProcessingCount = fileRepository.countByVectorizationStatus("processing");
+            }
+
+            result.put("code", 0);
+            result.put("message", "查询成功");
+            result.put("projectId", projectId);
+            result.put("catalogExtractionFailed", catalogFailedCount);
+            result.put("catalogExtractionProcessing", catalogProcessingCount);
+            result.put("vectorizationFailed", vectorFailedCount);
+            result.put("vectorizationProcessing", vectorProcessingCount);
+
+        } catch (Exception e) {
+            logger.error("查询失败任务状态出错", e);
+            result.put("code", 500);
+            result.put("message", "系统错误: " + e.getMessage());
+            return ResponseEntity.internalServerError().body(result);
+        }
+
+        return ResponseEntity.ok(result);
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/controller/SysLogController.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/controller/SysLogController.java
new file mode 100644
index 0000000..3150f7a
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/controller/SysLogController.java
@@ -0,0 +1,153 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.controller
+ * @ClassName: SysLogController
+ * @Description: 系统日志查询接口
+ * @Author: 张志锋
+ * @Date: 2025-11-24
+ * @Version: 1.0
+ */
+package com.zhpb.gdyd_zhpb_zgf.controller;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.SysLog;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.SysLogRepository;
+import io.swagger.v3.oas.annotations.Operation;
+import io.swagger.v3.oas.annotations.tags.Tag;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.data.domain.Page;
+import org.springframework.data.domain.PageRequest;
+import org.springframework.data.domain.Pageable;
+import org.springframework.data.domain.Sort;
+import org.springframework.data.jpa.domain.Specification;
+import org.springframework.util.StringUtils;
+import org.springframework.web.bind.annotation.GetMapping;
+import org.springframework.web.bind.annotation.RequestMapping;
+import org.springframework.web.bind.annotation.RequestParam;
+import org.springframework.web.bind.annotation.RestController;
+
+import jakarta.persistence.criteria.Predicate;
+import java.time.LocalDateTime;
+import java.time.format.DateTimeFormatter;
+import java.time.format.DateTimeParseException;
+import java.util.*;
+
+/**
+ * 系统日志查询接口
+ */
+@RestController
+@RequestMapping("/api/logs")
+@Tag(name = "系统日志", description = "系统日志查询接口")
+public class SysLogController {
+
+    private static final DateTimeFormatter ISO_FORMATTER = DateTimeFormatter.ISO_DATE_TIME;
+    private static final DateTimeFormatter NORMAL_FORMATTER = DateTimeFormatter.ofPattern("yyyy-MM-dd HH:mm:ss");
+    private static final List<String> KNOWN_LEVELS = Arrays.asList("ERROR", "WARN", "INFO", "DEBUG", "TRACE", "OUT");
+
+    @Autowired
+    private SysLogRepository sysLogRepository;
+
+    @GetMapping
+    @Operation(summary = "分页查询日志", description = "支持按级别、关键字、时间范围筛选日志")
+    public Map<String, Object> queryLogs(
+            @RequestParam(defaultValue = "0") int page,
+            @RequestParam(defaultValue = "20") int size,
+            @RequestParam(required = false) String level,
+            @RequestParam(required = false) String keyword,
+            @RequestParam(required = false) String startTime,
+            @RequestParam(required = false) String endTime) {
+
+        int safePage = Math.max(page, 0);
+        int safeSize = Math.min(Math.max(size, 1), 200);
+
+        Pageable pageable = PageRequest.of(safePage, safeSize, Sort.by(Sort.Direction.DESC, "createTime"));
+        Specification<SysLog> specification = buildSpecification(level, keyword, startTime, endTime);
+
+        Page<SysLog> pageData = sysLogRepository.findAll(specification, pageable);
+
+        Map<String, Object> response = new HashMap<>();
+        response.put("content", pageData.getContent());
+        response.put("totalElements", pageData.getTotalElements());
+        response.put("totalPages", pageData.getTotalPages());
+        response.put("page", pageData.getNumber());
+        response.put("size", pageData.getSize());
+        return response;
+    }
+
+    @GetMapping("/stats")
+    @Operation(summary = "日志统计信息", description = "返回日志数量、各级别分布等统计数据")
+    public Map<String, Object> logStats() {
+        Map<String, Object> result = new HashMap<>();
+        long total = sysLogRepository.count();
+        result.put("total", total);
+
+        LocalDateTime last24h = LocalDateTime.now().minusHours(24);
+        long last24hCount = sysLogRepository.count((root, query, cb) ->
+                cb.greaterThanOrEqualTo(root.get("createTime"), last24h));
+        result.put("last24Hours", last24hCount);
+
+        Map<String, Long> levelStats = new LinkedHashMap<>();
+        for (String lvl : KNOWN_LEVELS) {
+            long count = sysLogRepository.count((root, query, cb) ->
+                    cb.equal(cb.upper(root.get("logLevel")), lvl));
+            levelStats.put(lvl, count);
+        }
+        result.put("levelStats", levelStats);
+
+        SysLog latest = sysLogRepository.findTopByOrderByCreateTimeDesc();
+        result.put("latest", latest);
+        return result;
+    }
+
+    private Specification<SysLog> buildSpecification(String level, String keyword,
+                                                     String startTime, String endTime) {
+        return (root, query, cb) -> {
+            List<Predicate> predicates = new ArrayList<>();
+
+            if (StringUtils.hasText(level) && !"ALL".equalsIgnoreCase(level)) {
+                predicates.add(cb.equal(cb.upper(root.get("logLevel")), level.trim().toUpperCase()));
+            }
+
+            if (StringUtils.hasText(keyword)) {
+                String likeValue = "%" + keyword.trim().toLowerCase() + "%";
+                predicates.add(cb.or(
+                        cb.like(cb.lower(root.get("logContent")), likeValue),
+                        cb.like(cb.lower(root.get("fullMessage")), likeValue),
+                        cb.like(cb.lower(root.get("loggerName")), likeValue),
+                        cb.like(cb.lower(root.get("threadName")), likeValue)
+                ));
+            }
+
+            LocalDateTime start = parseDateTime(startTime);
+            LocalDateTime end = parseDateTime(endTime);
+            if (start != null) {
+                predicates.add(cb.greaterThanOrEqualTo(root.get("createTime"), start));
+            }
+            if (end != null) {
+                predicates.add(cb.lessThanOrEqualTo(root.get("createTime"), end));
+            }
+
+            return cb.and(predicates.toArray(new Predicate[0]));
+        };
+    }
+
+    private LocalDateTime parseDateTime(String raw) {
+        if (!StringUtils.hasText(raw)) {
+            return null;
+        }
+        String value = raw.trim();
+        // datetime-local 会带 T
+        if (value.contains("T") && !value.contains("Z")) {
+            value = value.replace("T", " ");
+        }
+        List<DateTimeFormatter> formatters = Arrays.asList(ISO_FORMATTER, NORMAL_FORMATTER,
+                DateTimeFormatter.ofPattern("yyyy-MM-dd HH:mm"), DateTimeFormatter.ISO_LOCAL_DATE_TIME);
+        for (DateTimeFormatter formatter : formatters) {
+            try {
+                return LocalDateTime.parse(value, formatter);
+            } catch (DateTimeParseException ignored) {
+            }
+        }
+        return null;
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/dto/ComparisonRequest.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/dto/ComparisonRequest.java
new file mode 100644
index 0000000..4ec27ab
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/dto/ComparisonRequest.java
@@ -0,0 +1,93 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.dto
+ * @ClassName: ComparisonRequest
+ * @Description: 对比分析请求DTO
+ * @Author: 张志锋
+ * @Date: 2025-11-04
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.dto;
+
+import io.swagger.v3.oas.annotations.media.Schema;
+
+import java.util.List;
+
+/**
+ * 对比分析请求DTO
+ */
+@Schema(description = "对比分析请求")
+public class ComparisonRequest {
+
+    @Schema(description = "项目ID", example = "project-001", required = true)
+    private String projectId;
+
+    @Schema(description = "用户ID", example = "user-001", required = true)
+    private String userId;
+
+    @Schema(description = "供应商ID列表", required = true)
+    private List<String> vendorIds;
+
+    @Schema(description = "对比维度", required = true)
+    private List<ComparisonDimension> dimensions;
+
+    public String getProjectId() {
+        return projectId;
+    }
+
+    public void setProjectId(String projectId) {
+        this.projectId = projectId;
+    }
+
+    public String getUserId() {
+        return userId;
+    }
+
+    public void setUserId(String userId) {
+        this.userId = userId;
+    }
+
+    public List<String> getVendorIds() {
+        return vendorIds;
+    }
+
+    public void setVendorIds(List<String> vendorIds) {
+        this.vendorIds = vendorIds;
+    }
+
+    public List<ComparisonDimension> getDimensions() {
+        return dimensions;
+    }
+
+    public void setDimensions(List<ComparisonDimension> dimensions) {
+        this.dimensions = dimensions;
+    }
+
+    /**
+     * 对比维度
+     */
+    public static class ComparisonDimension {
+        @Schema(description = "维度类型", example = "price", allowableValues = {"price", "quality", "delivery", "service", "technology"})
+        private String type;
+
+        @Schema(description = "维度值", example = "价格")
+        private String value;
+
+        public String getType() {
+            return type;
+        }
+
+        public void setType(String type) {
+            this.type = type;
+        }
+
+        public String getValue() {
+            return value;
+        }
+
+        public void setValue(String value) {
+            this.value = value;
+        }
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/dto/ComparisonResponse.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/dto/ComparisonResponse.java
new file mode 100644
index 0000000..1e231b7
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/dto/ComparisonResponse.java
@@ -0,0 +1,471 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.dto
+ * @ClassName: ComparisonResponse
+ * @Description: 对比分析响应DTO
+ * @Author: 张志锋
+ * @Date: 2025-11-04
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.dto;
+
+import io.swagger.v3.oas.annotations.media.Schema;
+
+import java.util.List;
+
+/**
+ * 对比分析响应DTO
+ */
+public class ComparisonResponse {
+
+    /**
+     * 获取供应商数据响应
+     */
+    @Schema(description = "获取供应商数据响应")
+    public static class VendorsData {
+        @Schema(description = "供应商列表")
+        private List<VendorInfo> vendors;
+
+        public List<VendorInfo> getVendors() {
+            return vendors;
+        }
+
+        public void setVendors(List<VendorInfo> vendors) {
+            this.vendors = vendors;
+        }
+    }
+
+    /**
+     * 供应商信息
+     */
+    @Schema(description = "供应商信息")
+    public static class VendorInfo {
+        @Schema(description = "供应商ID", example = "supplier-001")
+        private String vendorId;
+
+        @Schema(description = "供应商名称", example = "彩讯科技股份有限公司")
+        private String vendorName;
+
+        @Schema(description = "供应商类型", example = "技术服务商")
+        private String vendorType;
+
+        @Schema(description = "文件列表")
+        private List<FileInfo> files;
+
+        public String getVendorId() {
+            return vendorId;
+        }
+
+        public void setVendorId(String vendorId) {
+            this.vendorId = vendorId;
+        }
+
+        public String getVendorName() {
+            return vendorName;
+        }
+
+        public void setVendorName(String vendorName) {
+            this.vendorName = vendorName;
+        }
+
+        public String getVendorType() {
+            return vendorType;
+        }
+
+        public void setVendorType(String vendorType) {
+            this.vendorType = vendorType;
+        }
+
+        public List<FileInfo> getFiles() {
+            return files;
+        }
+
+        public void setFiles(List<FileInfo> files) {
+            this.files = files;
+        }
+    }
+
+    /**
+     * 文件信息
+     */
+    @Schema(description = "文件信息")
+    public static class FileInfo {
+        @Schema(description = "文件ID", example = "file-001")
+        private String fileId;
+
+        @Schema(description = "文件名称", example = "技术方案.pdf")
+        private String fileName;
+
+        @Schema(description = "文件类型", example = "技术文件")
+        private String fileType;
+
+        @Schema(description = "文件大小(字节)", example = "2048576")
+        private Long fileSize;
+
+        public String getFileId() {
+            return fileId;
+        }
+
+        public void setFileId(String fileId) {
+            this.fileId = fileId;
+        }
+
+        public String getFileName() {
+            return fileName;
+        }
+
+        public void setFileName(String fileName) {
+            this.fileName = fileName;
+        }
+
+        public String getFileType() {
+            return fileType;
+        }
+
+        public void setFileType(String fileType) {
+            this.fileType = fileType;
+        }
+
+        public Long getFileSize() {
+            return fileSize;
+        }
+
+        public void setFileSize(Long fileSize) {
+            this.fileSize = fileSize;
+        }
+    }
+
+    /**
+     * 启动对比任务响应
+     */
+    @Schema(description = "启动对比任务响应")
+    public static class StartTaskData {
+        @Schema(description = "任务ID", example = "task-001")
+        private String taskId;
+
+        @Schema(description = "任务状态", example = "pending:待处理")
+        private String status;
+
+        @Schema(description = "创建时间", example = "2025-11-04T12:00:00Z")
+        private String createdTime;
+
+        public String getTaskId() {
+            return taskId;
+        }
+
+        public void setTaskId(String taskId) {
+            this.taskId = taskId;
+        }
+
+        public String getStatus() {
+            return status;
+        }
+
+        public void setStatus(String status) {
+            this.status = status;
+        }
+
+        public String getCreatedTime() {
+            return createdTime;
+        }
+
+        public void setCreatedTime(String createdTime) {
+            this.createdTime = createdTime;
+        }
+    }
+
+    /**
+     * 对比任务历史响应
+     */
+    @Schema(description = "对比任务历史响应")
+    public static class TaskHistoryData {
+        @Schema(description = "任务列表")
+        private List<TaskInfo> tasks;
+
+        public List<TaskInfo> getTasks() {
+            return tasks;
+        }
+
+        public void setTasks(List<TaskInfo> tasks) {
+            this.tasks = tasks;
+        }
+    }
+
+    /**
+     * 任务信息
+     */
+    @Schema(description = "任务信息")
+    public static class TaskInfo {
+        @Schema(description = "任务ID", example = "task-001")
+        private String taskId;
+
+        @Schema(description = "任务名称", example = "对比任务1")
+        private String taskName;
+
+        @Schema(description = "项目ID", example = "project-001")
+        private String projectId;
+
+        @Schema(description = "用户ID", example = "user-001")
+        private String userId;
+
+        @Schema(description = "任务状态", example = "completed:已完成", allowableValues = {"pending:待处理", "processing:处理中", "completed:已完成", "failed:失败"})
+        private String status;
+
+        @Schema(description = "供应商数量", example = "3")
+        private Integer vendorCount;
+
+        @Schema(description = "对比维度数量", example = "5")
+        private Integer dimensionCount;
+
+        @Schema(description = "创建时间", example = "2025-11-04T12:00:00Z")
+        private String createdTime;
+
+        @Schema(description = "完成时间", example = "2025-11-04T12:05:00Z")
+        private String completedTime;
+
+        @Schema(description = "对比维度列表", example = "[\"技术能力\", \"商务条款\", \"报价分析\"]")
+        private List<String> dimensions;
+
+        @Schema(description = "供应商列表", example = "[\"供应商A\", \"供应商B\", \"供应商C\"]")
+        private List<String> vendors;
+
+        @Schema(description = "任务耗时（秒）", example = "120")
+        private Long time;
+
+        public String getTaskId() {
+            return taskId;
+        }
+
+        public void setTaskId(String taskId) {
+            this.taskId = taskId;
+        }
+
+        public String getProjectId() {
+            return projectId;
+        }
+
+        public void setProjectId(String projectId) {
+            this.projectId = projectId;
+        }
+
+        public String getUserId() {
+            return userId;
+        }
+
+        public void setUserId(String userId) {
+            this.userId = userId;
+        }
+
+        public String getStatus() {
+            return status;
+        }
+
+        public void setStatus(String status) {
+            this.status = status;
+        }
+
+        public Integer getVendorCount() {
+            return vendorCount;
+        }
+
+        public void setVendorCount(Integer vendorCount) {
+            this.vendorCount = vendorCount;
+        }
+
+        public Integer getDimensionCount() {
+            return dimensionCount;
+        }
+
+        public void setDimensionCount(Integer dimensionCount) {
+            this.dimensionCount = dimensionCount;
+        }
+
+        public String getCreatedTime() {
+            return createdTime;
+        }
+
+        public void setCreatedTime(String createdTime) {
+            this.createdTime = createdTime;
+        }
+
+        public String getCompletedTime() {
+            return completedTime;
+        }
+
+        public void setCompletedTime(String completedTime) {
+            this.completedTime = completedTime;
+        }
+
+        public List<String> getDimensions() {
+            return dimensions;
+        }
+
+        public void setDimensions(List<String> dimensions) {
+            this.dimensions = dimensions;
+        }
+
+        public List<String> getVendors() {
+            return vendors;
+        }
+
+        public void setVendors(List<String> vendors) {
+            this.vendors = vendors;
+        }
+
+        public Long getTime() {
+            return time;
+        }
+
+        public void setTime(Long time) {
+            this.time = time;
+        }
+
+        public String getTaskName() {
+            return taskName;
+        }
+
+        public void setTaskName(String taskName) {
+            this.taskName = taskName;
+        }
+    }
+
+    /**
+     * 对比结果响应
+     */
+    @Schema(description = "对比结果响应")
+    public static class ComparisonResultData {
+        @Schema(description = "任务ID", example = "task-001")
+        private String taskId;
+
+        @Schema(description = "任务状态", example = "completed:已完成")
+        private String status;
+
+        @Schema(description = "对比结果矩阵")
+        private List<ResultMatrixItem> resultMatrix;
+
+        public String getTaskId() {
+            return taskId;
+        }
+
+        public void setTaskId(String taskId) {
+            this.taskId = taskId;
+        }
+
+        public String getStatus() {
+            return status;
+        }
+
+        public void setStatus(String status) {
+            this.status = status;
+        }
+
+        public List<ResultMatrixItem> getResultMatrix() {
+            return resultMatrix;
+        }
+
+        public void setResultMatrix(List<ResultMatrixItem> resultMatrix) {
+            this.resultMatrix = resultMatrix;
+        }
+    }
+
+    /**
+     * 结果矩阵项
+     */
+    @Schema(description = "结果矩阵项")
+    public static class ResultMatrixItem {
+        @Schema(description = "对比维度", example = "价格")
+        private String dimension;
+
+        @Schema(description = "维度类型", example = "price")
+        private String dimensionType;
+
+        @Schema(description = "供应商对比结果")
+        private List<VendorComparison> comparisons;
+
+        public String getDimension() {
+            return dimension;
+        }
+
+        public void setDimension(String dimension) {
+            this.dimension = dimension;
+        }
+
+        public String getDimensionType() {
+            return dimensionType;
+        }
+
+        public void setDimensionType(String dimensionType) {
+            this.dimensionType = dimensionType;
+        }
+
+        public List<VendorComparison> getComparisons() {
+            return comparisons;
+        }
+
+        public void setComparisons(List<VendorComparison> comparisons) {
+            this.comparisons = comparisons;
+        }
+    }
+
+    /**
+     * 供应商对比结果
+     */
+    @Schema(description = "供应商对比结果")
+    public static class VendorComparison {
+        @Schema(description = "供应商ID", example = "supplier-001")
+        private String vendorId;
+
+        @Schema(description = "供应商名称", example = "彩讯科技股份有限公司")
+        private String vendorName;
+
+        @Schema(description = "对比总结", example = "该供应商在价格方面具有竞争力...")
+        private String summary;
+
+        @Schema(description = "相似度分数列表")
+        private List<Double> similarityScores;
+
+        @Schema(description = "相关文件信息")
+        private List<FileInfo> fileInfo;
+
+        public String getVendorId() {
+            return vendorId;
+        }
+
+        public void setVendorId(String vendorId) {
+            this.vendorId = vendorId;
+        }
+
+        public String getVendorName() {
+            return vendorName;
+        }
+
+        public void setVendorName(String vendorName) {
+            this.vendorName = vendorName;
+        }
+
+        public String getSummary() {
+            return summary;
+        }
+
+        public void setSummary(String summary) {
+            this.summary = summary;
+        }
+
+        public List<Double> getSimilarityScores() {
+            return similarityScores;
+        }
+
+        public void setSimilarityScores(List<Double> similarityScores) {
+            this.similarityScores = similarityScores;
+        }
+
+        public List<FileInfo> getFileInfo() {
+            return fileInfo;
+        }
+
+        public void setFileInfo(List<FileInfo> fileInfo) {
+            this.fileInfo = fileInfo;
+        }
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/dto/FileUploadRequest.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/dto/FileUploadRequest.java
new file mode 100644
index 0000000..fee03a7
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/dto/FileUploadRequest.java
@@ -0,0 +1,178 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.dto
+ * @ClassName: FileUploadRequest
+ * @Description: 文件上传请求DTO
+ * @Author: 张志锋
+ * @Date: 2025-11-05
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.dto;
+
+import io.swagger.v3.oas.annotations.media.Schema;
+
+import java.util.List;
+
+/**
+ * ES平台文件上传请求DTO
+ */
+public class FileUploadRequest {
+
+    /**
+     * 采购文件上传请求
+     */
+    @Schema(description = "采购文件上传请求")
+    public static class ProcurementFileRequest {
+        @Schema(description = "项目ID", example = "550e8400-e29b-41d4-a716-446655440000")
+        private String projectId;
+
+        @Schema(description = "包ID", example = "550e8400-e29b-41d4-a716-446655440001")
+        private String packageId;
+
+        @Schema(description = "采购文件JSON的URL列表")
+        private List<String> fileIds;
+
+        public String getProjectId() {
+            return projectId;
+        }
+
+        public void setProjectId(String projectId) {
+            this.projectId = projectId;
+        }
+
+        public String getPackageId() {
+            return packageId;
+        }
+
+        public void setPackageId(String packageId) {
+            this.packageId = packageId;
+        }
+
+        public List<String> getFileIds() {
+            return fileIds;
+        }
+
+        public void setFileIds(List<String> fileIds) {
+            this.fileIds = fileIds;
+        }
+    }
+
+    /**
+     * 应答文件上传请求
+     */
+    @Schema(description = "应答文件上传请求")
+    public static class ResponseFileRequest {
+        @Schema(description = "项目ID", example = "550e8400-e29b-41d4-a716-446655440000")
+        private String projectId;
+
+        @Schema(description = "包ID", example = "550e8400-e29b-41d4-a716-446655440001")
+        private String packageId;
+
+        @Schema(description = "供应商ID", example = "supplier-001")
+        private String supplierId;
+
+        @Schema(description = "供应商应答文件URL列表")
+        private List<String> fileIds;
+
+        @Schema(description = "文件类型：1商务、2技术", example = "1")
+        private String fileType;
+
+        @Schema(description = "采购文件JSON的URL地址")
+        private String jsonFileId;
+
+        @Schema(description = "单位ID", example = "00030000000000000000")
+        private String unitId;
+
+        public String getProjectId() {
+            return projectId;
+        }
+
+        public void setProjectId(String projectId) {
+            this.projectId = projectId;
+        }
+
+        public String getPackageId() {
+            return packageId;
+        }
+
+        public void setPackageId(String packageId) {
+            this.packageId = packageId;
+        }
+
+        public String getSupplierId() {
+            return supplierId;
+        }
+
+        public void setSupplierId(String supplierId) {
+            this.supplierId = supplierId;
+        }
+
+        public List<String> getFileIds() {
+            return fileIds;
+        }
+
+        public void setFileIds(List<String> fileIds) {
+            this.fileIds = fileIds;
+        }
+
+        public String getFileType() {
+            return fileType;
+        }
+
+        public void setFileType(String fileType) {
+            this.fileType = fileType;
+        }
+
+        public String getJsonFileId() {
+            return jsonFileId;
+        }
+
+        public void setJsonFileId(String jsonFileId) {
+            this.jsonFileId = jsonFileId;
+        }
+
+        public String getUnitId() {
+            return unitId;
+        }
+
+        public void setUnitId(String unitId) {
+            this.unitId = unitId;
+        }
+    }
+
+    /**
+     * 通用响应
+     */
+    @Schema(description = "API响应")
+    public static class ApiResponse {
+        @Schema(description = "响应编码：0成功，其他失败", example = "0")
+        private String code;
+
+        @Schema(description = "响应消息", example = "成功")
+        private String msg;
+
+        public ApiResponse() {}
+
+        public ApiResponse(String code, String msg) {
+            this.code = code;
+            this.msg = msg;
+        }
+
+        public String getCode() {
+            return code;
+        }
+
+        public void setCode(String code) {
+            this.code = code;
+        }
+
+        public String getMsg() {
+            return msg;
+        }
+
+        public void setMsg(String msg) {
+            this.msg = msg;
+        }
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/dto/FileUploadResponse.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/dto/FileUploadResponse.java
new file mode 100644
index 0000000..b2a7c9a
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/dto/FileUploadResponse.java
@@ -0,0 +1,189 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.dto
+ * @ClassName: FileUploadResponse
+ * @Description: 文件上传响应DTO，包含上传结果、文件路径列表和按公司、文件类型整理后的文件列表
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.dto;
+
+import io.swagger.v3.oas.annotations.media.Schema;
+import lombok.Data;
+
+import java.time.LocalDateTime;
+import java.util.List;
+import java.util.Map;
+
+/**
+ * 文件上传响应DTO
+ */
+@Data
+@Schema(description = "文件上传响应")
+public class FileUploadResponse {
+
+    @Schema(description = "响应状态码，0表示成功", example = "0")
+    private Integer code;
+
+    @Schema(description = "响应消息", example = "上传成功")
+    private String message;
+
+    @Schema(description = "响应数据")
+    private FileUploadData data;
+
+    @Data
+    @Schema(description = "文件上传响应数据")
+    public static class FileUploadData {
+        @Schema(description = "项目ID", example = "project-001")
+        private String projectId;
+
+        @Schema(description = "原始ZIP文件路径", example = "./uploads/project-001/file.zip")
+        private String zipFilePath;
+
+        @Schema(description = "解压路径", example = "./extracts/project-001/")
+        private String extractPath;
+
+        @Schema(description = "解压后的文件路径列表", example = "[\"./extracts/project-001/file1.pdf\", \"./extracts/project-001/file2.doc\"]")
+        private List<String> extractedFilePaths;
+
+        @Schema(description = "文件数量", example = "5")
+        private Integer fileCount;
+
+        @Schema(description = "按公司和文件类型整理后的文件列表", example = "{\"公司1\": {\"技术文件\": [\"...\"], \"商务文件\": [...]}}")
+        private Map<String, Map<String, List<String>>> organizedFiles;
+    }
+
+    /**
+     * 文件向量化状态
+     */
+    @Data
+    @Schema(description = "文件向量化状态信息")
+    public static class VectorizationStatus {
+        @Schema(description = "文件ID", example = "550e8400-e29b-41d4-a716-446655440000")
+        private String fileId;
+
+        @Schema(description = "文件名", example = "技术方案.pdf")
+        private String fileName;
+
+        @Schema(description = "项目ID", example = "project-001")
+        private String projectId;
+
+        @Schema(description = "向量化状态", example = "completed", allowableValues = {"not_started", "processing", "completed", "failed"})
+        private String vectorizationStatus;
+
+        @Schema(description = "是否已向量化", example = "true")
+        private Boolean isVectorized;
+
+        @Schema(description = "向量化数据条数", example = "150")
+        private Integer vectorizationCount;
+
+        @Schema(description = "向量化完成时间", example = "2025-11-03T15:30:00")
+        private LocalDateTime vectorizedTime;
+
+        @Schema(description = "文件扩展名", example = "pdf")
+        private String fileExtension;
+
+        @Schema(description = "文件大小（字节）", example = "2048576")
+        private Long fileSize;
+    }
+
+    /**
+     * 向量化状态统计
+     */
+    @Data
+    @Schema(description = "向量化状态统计信息")
+    public static class VectorizationStats {
+        @Schema(description = "总文件数", example = "25")
+        private int totalFiles;
+
+        @Schema(description = "已完成向量化文件数", example = "20")
+        private int completedFiles;
+
+        @Schema(description = "正在处理的文件数", example = "2")
+        private int processingFiles;
+
+        @Schema(description = "向量化失败的文件数", example = "1")
+        private int failedFiles;
+
+        @Schema(description = "未开始向化的文件数", example = "2")
+        private int notStartedFiles;
+
+        @Schema(description = "向量化完成率", example = "0.8")
+        private double completionRate;
+
+        @Schema(description = "最近的文件列表")
+        private List<VectorizationStatus> recentFiles;
+    }
+
+    /**
+     * 供应商信息
+     */
+    @Data
+    @Schema(description = "供应商基本信息")
+    public static class SupplierInfo {
+        @Schema(description = "供应商ID", example = "supplier-001")
+        private String supplierId;
+
+        @Schema(description = "供应商名称", example = "彩讯科技股份有限公司")
+        private String supplierName;
+
+        @Schema(description = "供应商代码", example = "CXKJ")
+        private String supplierCode;
+
+        @Schema(description = "供应商类型", example = "企业")
+        private String supplierType;
+
+        @Schema(description = "联系人", example = "张三")
+        private String contactPerson;
+
+        @Schema(description = "联系电话", example = "13800138000")
+        private String contactPhone;
+
+        @Schema(description = "联系邮箱", example = "contact@company.com")
+        private String contactEmail;
+
+        @Schema(description = "供应商地址", example = "北京市朝阳区")
+        private String address;
+
+        @Schema(description = "供应商状态", example = "active")
+        private String status;
+    }
+
+    /**
+     * 项目供应商和文件信息（简化格式）
+     */
+    @Data
+    @Schema(description = "项目供应商和文件信息（供应商为键，文件列表为值）")
+    public static class ProjectSuppliersAndFilesSimple {
+        @Schema(description = "供应商列表")
+        private List<SupplierInfo> 供应商;
+
+        // 动态字段：供应商名称 -> 文件列表
+        // 使用Map来存储供应商名称和对应的文件列表
+    }
+
+    /**
+     * 项目供应商和文件信息（完整格式）
+     */
+    @Data
+    @Schema(description = "项目供应商和文件信息")
+    public static class ProjectSuppliersAndFiles {
+        @Schema(description = "项目ID", example = "project-001")
+        private String projectId;
+
+        @Schema(description = "项目名称", example = "智慧评标系统项目")
+        private String projectName;
+
+        @Schema(description = "供应商列表")
+        private List<SupplierInfo> suppliers;
+
+        @Schema(description = "文件列表")
+        private List<VectorizationStatus> files;
+
+        @Schema(description = "供应商文件分组")
+        private Map<String, List<VectorizationStatus>> supplierFiles;
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/dto/ProjectFilesResponse.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/dto/ProjectFilesResponse.java
new file mode 100644
index 0000000..c332f4d
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/dto/ProjectFilesResponse.java
@@ -0,0 +1,81 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.dto
+ * @ClassName: ProjectFilesResponse
+ * @Description: 项目文件列表查询响应DTO
+ * @Author: 张志锋
+ * @Date: 2025-11-03
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.dto;
+
+import io.swagger.v3.oas.annotations.media.Schema;
+import lombok.Data;
+
+import java.time.LocalDateTime;
+import java.util.List;
+import java.util.Map;
+
+/**
+ * 项目文件列表查询响应DTO
+ */
+@Data
+@Schema(description = "项目文件列表查询响应")
+public class ProjectFilesResponse {
+
+    @Schema(description = "响应状态码，0表示成功", example = "0")
+    private Integer code;
+
+    @Schema(description = "响应消息", example = "查询成功")
+    private String message;
+
+    @Schema(description = "响应数据")
+    private ProjectFilesData data;
+
+    @Data
+    @Schema(description = "项目文件数据")
+    public static class ProjectFilesData {
+
+        @Schema(description = "项目ID", example = "7654")
+        private String projectId;
+
+        @Schema(description = "文件总数", example = "6")
+        private Integer totalFiles;
+
+        @Schema(description = "文件列表")
+        private List<FileInfo> files;
+
+        @Schema(description = "按公司分组的文件列表")
+        private Map<String, List<String>> companyFiles;
+    }
+
+    @Data
+    @Schema(description = "文件信息")
+    public static class FileInfo {
+
+        @Schema(description = "文件ID", example = "1a0a5bdd-b073-4249-be61-e925be42e87f")
+        private String fileId;
+
+        @Schema(description = "公司名称", example = "xx科技有限公司")
+        private String companyName;
+
+        @Schema(description = "文件名", example = "1 技术规范文件应答.docx")
+        private String fileName;
+
+        @Schema(description = "文件大小（字节）", example = "58680")
+        private Long fileSize;
+
+        @Schema(description = "文件类型", example = "docx")
+        private String fileType;
+
+        @Schema(description = "相对路径", example = "test_files\\xx科技有限公司\\技术文件-2502071344\\1 技术规范文件应答.docx")
+        private String relativePath;
+
+        @Schema(description = "上传时间", example = "2025-11-03T15:42:35.372798")
+        private LocalDateTime uploadTime;
+
+        @Schema(description = "向量化状态", example = "processing", allowableValues = {"not_started", "processing", "completed", "failed"})
+        private String vectorizationStatus;
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/dto/PromptTemplateDto.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/dto/PromptTemplateDto.java
new file mode 100644
index 0000000..ba349c1
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/dto/PromptTemplateDto.java
@@ -0,0 +1,57 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.dto
+ * @ClassName: PromptTemplateDto
+ * @Description: 提示词模板DTO
+ * @Author: 张志锋
+ * @Date: 2025-11-03
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.dto;
+
+import io.swagger.v3.oas.annotations.media.Schema;
+import lombok.Data;
+
+import java.time.LocalDateTime;
+
+/**
+ * 提示词模板DTO
+ */
+@Data
+@Schema(description = "提示词模板")
+public class PromptTemplateDto {
+
+    @Schema(description = "模板ID", example = "template-001")
+    private String templateId;
+
+    @Schema(description = "模板名称", example = "智能问答提示词")
+    private String templateName;
+
+    @Schema(description = "模板类型", example = "qa", allowableValues = {"qa", "summary"})
+    private String templateType;
+
+    @Schema(description = "提示词内容", example = "请基于以下文档内容回答用户的问题...")
+    private String promptContent;
+
+    @Schema(description = "模板描述", example = "用于智能问答的提示词模板")
+    private String description;
+
+    @Schema(description = "是否启用", example = "true")
+    private Boolean isActive;
+
+    @Schema(description = "排序号", example = "1")
+    private Integer sortOrder;
+
+    @Schema(description = "创建人", example = "admin")
+    private String createdBy;
+
+    @Schema(description = "更新人", example = "admin")
+    private String updatedBy;
+
+    @Schema(description = "创建时间", example = "2025-11-03T10:00:00")
+    private LocalDateTime createTime;
+
+    @Schema(description = "更新时间", example = "2025-11-03T10:00:00")
+    private LocalDateTime updateTime;
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/dto/QaHistoryResponse.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/dto/QaHistoryResponse.java
new file mode 100644
index 0000000..98dd393
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/dto/QaHistoryResponse.java
@@ -0,0 +1,101 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.dto
+ * @ClassName: QaHistoryResponse
+ * @Description: QA历史记录响应DTO
+ * @Author: 张志锋
+ * @Date: 2025-11-03
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.dto;
+
+import io.swagger.v3.oas.annotations.media.Schema;
+import lombok.Data;
+
+import java.util.List;
+
+/**
+ * QA历史记录响应DTO
+ */
+@Data
+@Schema(description = "QA历史记录响应")
+public class QaHistoryResponse {
+
+    @Schema(description = "响应状态码，0表示成功", example = "0")
+    private Integer code;
+
+    @Schema(description = "响应消息", example = "Success")
+    private String message;
+
+    @Schema(description = "响应数据")
+    private Object data;
+
+    @Data
+    @Schema(description = "会话历史数据")
+    public static class SessionsData {
+
+        @Schema(description = "会话列表")
+        private List<SessionInfo> sessions;
+    }
+
+    @Data
+    @Schema(description = "消息历史数据")
+    public static class MessagesData {
+
+        @Schema(description = "会话ID", example = "session-5e205312")
+        private String sessionId;
+
+        @Schema(description = "消息列表")
+        private List<MessageInfo> messages;
+    }
+
+    @Data
+    @Schema(description = "会话信息")
+    public static class SessionInfo {
+
+        @Schema(description = "会话ID", example = "session-5e205312")
+        private String sessionId;
+
+        @Schema(description = "开始时间", example = "2025-11-03T16:50:00")
+        private String startTime;
+
+        @Schema(description = "首次问题内容", example = "这个项目的预算有多少？")
+        private String firstQuery;
+
+        @Schema(description = "当前会话涉及的文件ID列表")
+        private List<String> fileIds;
+
+        @Schema(description = "当前会话涉及的文件名列表")
+        private List<String> fileNames;
+
+        @Schema(description = "当前会话涉及的供应商名称列表")
+        private List<String> supplierNames;
+
+        @Schema(description = "当前会话涉及的供应商ID列表")
+        private List<String> supplierIds;
+    }
+
+    @Data
+    @Schema(description = "消息信息")
+    public static class MessageInfo {
+
+        @Schema(description = "消息ID", example = "msg-4e529b65")
+        private String messageId;
+
+        @Schema(description = "发送者", example = "user", allowableValues = {"user", "ai"})
+        private String sender;
+
+        @Schema(description = "消息内容", example = "这个项目的预算有多少？")
+        private String content;
+
+        @Schema(description = "来源信息")
+        private Object sources;
+
+        @Schema(description = "时间戳", example = "2025-11-03T16:50:00")
+        private String timestamp;
+
+        @Schema(description = "引用文件ID列表")
+        private List<String> fileIds;
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/dto/QaRequest.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/dto/QaRequest.java
new file mode 100644
index 0000000..a59cd4e
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/dto/QaRequest.java
@@ -0,0 +1,39 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.dto
+ * @ClassName: QaRequest
+ * @Description: 智能问答请求DTO
+ * @Author: 张志锋
+ * @Date: 2025-11-03
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.dto;
+
+import io.swagger.v3.oas.annotations.media.Schema;
+import lombok.Data;
+
+import java.util.List;
+
+/**
+ * 智能问答请求DTO
+ */
+@Data
+@Schema(description = "智能问答请求")
+public class QaRequest {
+
+    @Schema(description = "项目ID", example = "project-001", required = true)
+    private String projectId;
+
+    @Schema(description = "用户ID", example = "user-001", required = true)
+    private String userId;
+
+    @Schema(description = "会话ID", example = "session-001")
+    private String sessionId;
+
+    @Schema(description = "用户查询问题", example = "这个项目的预算有多少？", required = true)
+    private String query;
+
+    @Schema(description = "相关文件ID列表", example = "[\"file-001\", \"file-002\"]", required = true)
+    private List<String> fileIds;
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/dto/QaResponse.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/dto/QaResponse.java
new file mode 100644
index 0000000..7097258
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/dto/QaResponse.java
@@ -0,0 +1,76 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.dto
+ * @ClassName: QaResponse
+ * @Description: 智能问答响应DTO
+ * @Author: 张志锋
+ * @Date: 2025-11-03
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.dto;
+
+import io.swagger.v3.oas.annotations.media.Schema;
+import lombok.Data;
+
+import java.util.List;
+
+/**
+ * 智能问答响应DTO
+ */
+@Data
+@Schema(description = "智能问答响应")
+public class QaResponse {
+
+    @Schema(description = "响应状态码，0表示成功", example = "0")
+    private Integer code;
+
+    @Schema(description = "响应消息", example = "查询成功")
+    private String message;
+
+    @Schema(description = "响应数据")
+    private QaData data;
+
+    @Data
+    @Schema(description = "智能问答响应数据")
+    public static class QaData {
+
+        @Schema(description = "会话ID", example = "session-001")
+        private String sessionId;
+
+        @Schema(description = "消息ID", example = "msg-001")
+        private String messageId;
+
+        @Schema(description = "AI回答内容", example = "根据项目文档显示，这个项目的总预算为500万元人民币。")
+        private String answer;
+
+        @Schema(description = "答案来源信息")
+        private List<SourceInfo> sources;
+    }
+
+    @Data
+    @Schema(description = "答案来源信息")
+    public static class SourceInfo {
+
+        @Schema(description = "文件ID", example = "file-001")
+        private String fileId;
+
+        @Schema(description = "文件名", example = "技术方案.pdf")
+        private String fileName;
+
+        @Schema(description = "供应商名称", example = "彩讯科技股份有限公司")
+        private String vendorName;
+
+        @Schema(description = "页码", example = "5")
+        private Integer page;
+
+        @Schema(description = "章节", example = "第3章 项目预算")
+        private String chapter;
+
+        @Schema(description = "内容片段", example = "项目总预算为500万元，其中包含...")
+        private String snippet;
+
+        @Schema(description = "相似度", example = "0.85")
+        private Double similarity;
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/dto/SsoLoginResponse.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/dto/SsoLoginResponse.java
new file mode 100644
index 0000000..2aa574f
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/dto/SsoLoginResponse.java
@@ -0,0 +1,55 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.dto
+ * @ClassName: SsoLoginResponse
+ * @Description: SSO登录响应DTO，用于Swagger API文档说明
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.dto;
+
+import io.swagger.v3.oas.annotations.media.Schema;
+import lombok.AllArgsConstructor;
+import lombok.Data;
+import lombok.NoArgsConstructor;
+
+/**
+ * SSO登录响应DTO
+ */
+@Data
+@NoArgsConstructor
+@AllArgsConstructor
+@Schema(description = "SSO登录响应")
+public class SsoLoginResponse {
+    
+    @Schema(description = "响应码", example = "200")
+    private Integer code;
+    
+    @Schema(description = "响应消息", example = "登录成功")
+    private String message;
+    
+    @Schema(description = "用户数据")
+    private UserData data;
+    
+    @Data
+    @NoArgsConstructor
+    @AllArgsConstructor
+    @Schema(description = "用户数据")
+    public static class UserData {
+        
+        @Schema(description = "JWT Token", example = "eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9...")
+        private String token;
+        
+        @Schema(description = "用户ID", example = "n001")
+        private String userId;
+        
+        @Schema(description = "用户名称", example = "张三")
+        private String userName;
+        
+        @Schema(description = "项目ID", example = "0001")
+        private String projectId;
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/dto/SummaryResponse.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/dto/SummaryResponse.java
new file mode 100644
index 0000000..170cafc
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/dto/SummaryResponse.java
@@ -0,0 +1,124 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.dto
+ * @ClassName: SummaryResponse
+ * @Description: 总结相关响应DTO
+ * @Author: 张志锋
+ * @Date: 2025-11-03
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.dto;
+
+import io.swagger.v3.oas.annotations.media.Schema;
+import lombok.Data;
+
+import java.util.List;
+
+/**
+ * 总结相关响应DTO
+ */
+public class SummaryResponse {
+
+    @Data
+    @Schema(description = "目录响应")
+    public static class DirectoryResponse {
+
+        @Schema(description = "响应状态码，0表示成功", example = "0")
+        private Integer code;
+
+        @Schema(description = "响应消息", example = "Success")
+        private String message;
+
+        @Schema(description = "目录数据")
+        private DirectoryData data;
+    }
+
+    @Data
+    @Schema(description = "目录数据")
+    public static class DirectoryData {
+
+        @Schema(description = "文件ID", example = "9e607a37-ab87-4e02-8b42-125270d6bf58")
+        private String fileId;
+
+        @Schema(description = "目录结构")
+        private List<DirectoryItem> directory;
+    }
+
+    @Data
+    @Schema(description = "目录项")
+    public static class DirectoryItem {
+
+        @Schema(description = "章节ID", example = "chap-0")
+        private String chapterId;
+
+        @Schema(description = "章节标题", example = "1、技术规范文件应答")
+        private String title;
+
+        @Schema(description = "页码信息", example = "1")
+        private String page;
+
+        @Schema(description = "起始页码", example = "1")
+        private Integer startPage;
+
+        @Schema(description = "结束页码", example = "1")
+        private Integer endPage;
+
+        @Schema(description = "子章节列表")
+        private List<DirectoryItem> children;
+    }
+
+    @Data
+    @Schema(description = "章节总结响应")
+    public static class ChapterSummaryResponse {
+
+        @Schema(description = "响应状态码，0表示成功", example = "0")
+        private Integer code;
+
+        @Schema(description = "响应消息", example = "Success")
+        private String message;
+
+        @Schema(description = "章节总结数据")
+        private ChapterSummaryData data;
+    }
+
+    @Data
+    @Schema(description = "章节总结数据")
+    public static class ChapterSummaryData {
+
+        @Schema(description = "文件ID", example = "9e607a37-ab87-4e02-8b42-125270d6bf58")
+        private String fileId;
+
+        @Schema(description = "章节ID", example = "chap-0")
+        private String chapterId;
+
+        @Schema(description = "章节总结", example = "本章节主要介绍了...")
+        private String summary;
+
+        @Schema(description = "来源信息")
+        private List<SourceInfo> sources;
+    }
+
+    @Data
+    @Schema(description = "来源信息")
+    public static class SourceInfo {
+
+        @Schema(description = "页码", example = "1")
+        private String page;
+
+        @Schema(description = "文件ID", example = "9e607a37-ab87-4e02-8b42-125270d6bf58")
+        private String fileId;
+
+        @Schema(description = "内容片段", example = "具体内容...")
+        private String snippet;
+
+        @Schema(description = "章节ID", example = "chap-0")
+        private String chapterId;
+
+        @Schema(description = "起始页码", example = "1")
+        private Integer startPage;
+
+        @Schema(description = "结束页码", example = "1")
+        private Integer endPage;
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/ApiLog.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/ApiLog.java
new file mode 100644
index 0000000..dbcfcf6
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/ApiLog.java
@@ -0,0 +1,107 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.entity.mysql
+ * @ClassName: ApiLog
+ * @Description: API调用日志实体类，映射t_api_log表，记录所有API调用的详细信息
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.entity.mysql;
+
+import jakarta.persistence.*;
+import lombok.Data;
+import java.time.LocalDateTime;
+
+/**
+ * API调用日志实体类
+ */
+@Entity
+@Table(name = "t_api_log")
+@Data
+public class ApiLog {
+
+    @Id
+    @GeneratedValue(strategy = GenerationType.IDENTITY)
+    private Long id;
+
+    /**
+     * 调用的接口路径
+     */
+    @Column(name = "api_path", nullable = false, length = 512)
+    private String apiPath;
+
+    /**
+     * 接口中文名称
+     */
+    @Column(name = "api_name", length = 128)
+    private String apiName;
+
+    /**
+     * HTTP方法（GET, POST, PUT, DELETE等）
+     */
+    @Column(name = "http_method", length = 16)
+    private String httpMethod;
+
+    /**
+     * 请求参数（JSON格式）
+     */
+    @Column(name = "request_params", columnDefinition = "TEXT")
+    private String requestParams;
+
+    /**
+     * 请求体内容（JSON格式）
+     */
+    @Column(name = "request_body", columnDefinition = "TEXT")
+    private String requestBody;
+
+    /**
+     * 返回参数（JSON格式）
+     */
+    @Column(name = "response_body", columnDefinition = "TEXT")
+    private String responseBody;
+
+    /**
+     * 响应状态码（200, 404, 500等）
+     */
+    @Column(name = "status_code")
+    private Integer statusCode;
+
+    /**
+     * 耗费时长（毫秒）
+     */
+    @Column(name = "duration")
+    private Long duration;
+
+    /**
+     * 客户端IP地址
+     */
+    @Column(name = "client_ip", length = 64)
+    private String clientIp;
+
+    /**
+     * 用户ID（如果有）
+     */
+    @Column(name = "user_id", length = 64)
+    private String userId;
+
+    /**
+     * 用户名（如果有）
+     */
+    @Column(name = "user_name", length = 128)
+    private String userName;
+
+    /**
+     * 错误信息（如果有）
+     */
+    @Column(name = "error_msg", columnDefinition = "TEXT")
+    private String errorMsg;
+
+    /**
+     * 创建时间
+     */
+    @Column(name = "create_time", nullable = false)
+    private LocalDateTime createTime;
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/ComparisonTask.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/ComparisonTask.java
new file mode 100644
index 0000000..c63ab6b
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/ComparisonTask.java
@@ -0,0 +1,171 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.entity.mysql
+ * @ClassName: ComparisonTask
+ * @Description: 对比分析任务实体
+ * @Author: 张志锋
+ * @Date: 2025-11-04
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.entity.mysql;
+
+import jakarta.persistence.*;
+import java.time.LocalDateTime;
+
+/**
+ * 对比分析任务实体
+ */
+@Entity
+@Table(name = "t_comparison_task")
+public class ComparisonTask {
+
+    @Id
+    @Column(name = "task_id", length = 64)
+    private String taskId;
+
+    @Column(name = "project_id", length = 64, nullable = false)
+    private String projectId;
+
+    @Column(name = "user_id", length = 64, nullable = false)
+    private String userId;
+
+    @Column(name = "vendor_ids", columnDefinition = "TEXT")
+    private String vendorIds; // JSON格式存储供应商ID列表
+
+    @Column(name = "dimensions", columnDefinition = "TEXT")
+    private String dimensions; // JSON格式存储对比维度
+
+    @Column(name = "status", length = 32, nullable = false)
+    private String status; // pending, processing, completed, failed
+
+    @Lob
+    @Column(name = "result", columnDefinition = "TEXT")
+    private String result; // JSON格式存储对比结果
+
+    @Column(name = "error_message", columnDefinition = "TEXT")
+    private String errorMessage;
+
+    @Column(name = "progress", nullable = false)
+    private Integer progress = 0; // 进度百分比 0-100
+
+    @Column(name = "create_time", nullable = false)
+    private LocalDateTime createTime;
+
+    @Column(name = "update_time")
+    private LocalDateTime updateTime;
+
+    @Column(name = "completed_time")
+    private LocalDateTime completedTime;
+
+    // Getters and Setters
+    public String getTaskId() {
+        return taskId;
+    }
+
+    public void setTaskId(String taskId) {
+        this.taskId = taskId;
+    }
+
+    public String getProjectId() {
+        return projectId;
+    }
+
+    public void setProjectId(String projectId) {
+        this.projectId = projectId;
+    }
+
+    public String getUserId() {
+        return userId;
+    }
+
+    public void setUserId(String userId) {
+        this.userId = userId;
+    }
+
+    public String getVendorIds() {
+        return vendorIds;
+    }
+
+    public void setVendorIds(String vendorIds) {
+        this.vendorIds = vendorIds;
+    }
+
+    public String getDimensions() {
+        return dimensions;
+    }
+
+    public void setDimensions(String dimensions) {
+        this.dimensions = dimensions;
+    }
+
+    public String getStatus() {
+        return status;
+    }
+
+    public void setStatus(String status) {
+        this.status = status;
+    }
+
+    public String getResult() {
+        return result;
+    }
+
+    public void setResult(String result) {
+        this.result = result;
+    }
+
+    public String getErrorMessage() {
+        return errorMessage;
+    }
+
+    public void setErrorMessage(String errorMessage) {
+        this.errorMessage = errorMessage;
+    }
+
+    public Integer getProgress() {
+        return progress;
+    }
+
+    public void setProgress(Integer progress) {
+        this.progress = progress;
+    }
+
+    public LocalDateTime getCreateTime() {
+        return createTime;
+    }
+
+    public void setCreateTime(LocalDateTime createTime) {
+        this.createTime = createTime;
+    }
+
+    public LocalDateTime getUpdateTime() {
+        return updateTime;
+    }
+
+    public void setUpdateTime(LocalDateTime updateTime) {
+        this.updateTime = updateTime;
+    }
+
+    public LocalDateTime getCompletedTime() {
+        return completedTime;
+    }
+
+    public void setCompletedTime(LocalDateTime completedTime) {
+        this.completedTime = completedTime;
+    }
+
+    @PrePersist
+    protected void onCreate() {
+        createTime = LocalDateTime.now();
+        updateTime = LocalDateTime.now();
+        if (progress == null) {
+            progress = 0;
+        }
+    }
+
+    @PreUpdate
+    protected void onUpdate() {
+        updateTime = LocalDateTime.now();
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/File.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/File.java
new file mode 100644
index 0000000..62350ff
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/File.java
@@ -0,0 +1,171 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.entity.mysql
+ * @ClassName: File
+ * @Description: 文件实体类，映射t_file表，存储文件信息
+ * @Author: 张志锋
+ * @Date: 2025-11-01
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.entity.mysql;
+
+import jakarta.persistence.*;
+import lombok.Data;
+import java.time.LocalDateTime;
+
+/**
+ * 文件实体类
+ */
+@Entity
+@Table(name = "t_file")
+@Data
+public class File {
+
+    @Id
+    @GeneratedValue(strategy = GenerationType.IDENTITY)
+    private Long id;
+
+    /**
+     * 文件ID（业务ID，唯一）
+     */
+    @Column(name = "file_id", unique = true, nullable = false, length = 64)
+    private String fileId;
+
+    /**
+     * 项目ID（关联t_project表）
+     */
+    @Column(name = "project_id", nullable = false, length = 64)
+    private String projectId;
+
+    /**
+     * 供应商ID（关联t_supplier表）
+     */
+    @Column(name = "supplier_id", length = 64)
+    private String supplierId;
+
+    /**
+     * 文件名称
+     */
+    @Column(name = "file_name", nullable = false, length = 512)
+    private String fileName;
+
+    /**
+     * 文件路径（完整路径）
+     */
+    @Column(name = "file_path", nullable = false, columnDefinition = "TEXT")
+    private String filePath;
+
+    /**
+     * 文件扩展名（pdf, doc, xls等）
+     */
+    @Column(name = "file_extension", length = 32)
+    private String fileExtension;
+
+    /**
+     * 文件类型ID（关联t_file_type表）
+     */
+    @Column(name = "file_type_id")
+    private Long fileTypeId;
+
+    /**
+     * 文件大小（字节）
+     */
+    @Column(name = "file_size")
+    private Long fileSize;
+
+    /**
+     * 文件MD5值
+     */
+    @Column(name = "file_md5", length = 64)
+    private String fileMd5;
+
+    /**
+     * 文件状态（uploaded, processed, vectorized, failed等）
+     */
+    @Column(name = "status", length = 32)
+    private String status;
+
+    /**
+     * 向量化状态（not_started-未开始, processing-处理中, completed-已完成, failed-失败）
+     */
+    @Column(name = "vectorization_status", length = 32)
+    private String vectorizationStatus = "not_started";
+
+    /**
+     * 是否已向量化（兼容字段，根据vectorization_status自动设置）
+     */
+    @Column(name = "is_vectorized", nullable = false)
+    private Boolean isVectorized = false;
+
+    /**
+     * 向量化时间
+     */
+    @Column(name = "vectorized_time")
+    private LocalDateTime vectorizedTime;
+
+    /**
+     * 向量化数据条数（该文件对应的向量化记录数量）
+     */
+    @Column(name = "vectorization_count", nullable = false)
+    private Integer vectorizationCount = 0;
+
+    /**
+     * 目录摘取状态（not_started-未开始, processing-处理中, completed-已完成, failed-失败）
+     * 注意：失败且重试次数<3时会重置为 not_started，重试次数>=3时为 failed
+     */
+    @Column(name = "catalog_extraction_status", length = 32, columnDefinition = "varchar(32) default 'not_started'")
+    private String catalogExtractionStatus = "not_started";
+
+    /**
+     * 目录摘取重试次数（最多重试3次）
+     */
+    @Column(name = "catalog_extraction_retry_count", nullable = false, columnDefinition = "integer default 0")
+    private Integer catalogExtractionRetryCount = 0;
+
+    /**
+     * 目录摘取完成时间
+     */
+    @Column(name = "catalog_extraction_time")
+    private LocalDateTime catalogExtractionTime;
+
+    /**
+     * 文件描述
+     */
+    @Column(name = "description", columnDefinition = "TEXT")
+    private String description;
+
+    /**
+     * 预览接口URL（用于在线预览，如：/api/v1/files/preview/pdf/{file_id}）
+     */
+    @Column(name = "preview_path", columnDefinition = "TEXT")
+    private String previewPath;
+
+    /**
+     * 创建时间
+     */
+    @Column(name = "create_time", nullable = false)
+    private LocalDateTime createTime;
+
+    /**
+     * 更新时间
+     */
+    @Column(name = "update_time")
+    private LocalDateTime updateTime;
+
+    @PrePersist
+    public void prePersist() {
+        this.createTime = LocalDateTime.now();
+        this.updateTime = LocalDateTime.now();
+        if (this.fileId == null) {
+            // 如果没有设置fileId，自动生成UUID
+            this.fileId = java.util.UUID.randomUUID().toString().replace("-", "");
+        }
+    }
+
+    @PreUpdate
+    public void preUpdate() {
+        this.updateTime = LocalDateTime.now();
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/FileDirectory.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/FileDirectory.java
new file mode 100644
index 0000000..a23f60c
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/FileDirectory.java
@@ -0,0 +1,80 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.entity.mysql
+ * @ClassName: FileDirectory
+ * @Description: 文件目录实体
+ * @Author: 张志锋
+ * @Date: 2025-11-03
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.entity.mysql;
+
+import jakarta.persistence.*;
+import lombok.Data;
+
+import java.time.LocalDateTime;
+import java.util.List;
+
+/**
+ * 文件目录实体
+ */
+@Data
+@Entity
+@Table(name = "t_file_directory")
+public class FileDirectory {
+
+    @Id
+    @GeneratedValue(strategy = GenerationType.IDENTITY)
+    private Long id;
+
+    @Column(name = "file_id", nullable = false, length = 100)
+    private String fileId;
+
+    @Column(name = "chapter_id", nullable = false, length = 100)
+    private String chapterId;
+
+    @Column(name = "title", nullable = false, columnDefinition = "TEXT")
+    private String title;
+
+    @Column(name = "page", length = 50)
+    private String page;
+
+    @Column(name = "start_page")
+    private Integer startPage;
+
+    @Column(name = "end_page")
+    private Integer endPage;
+
+    @Column(name = "parent_chapter_id", length = 100)
+    private String parentChapterId;
+
+    @Column(name = "level", columnDefinition = "int default 1")
+    private Integer level = 1;
+
+    @Column(name = "sort_order", columnDefinition = "int default 0")
+    private Integer sortOrder = 0;
+
+    @Column(name = "summary", columnDefinition = "TEXT")
+    private String summary;
+
+    @Column(name = "create_time", nullable = false, updatable = false)
+    private LocalDateTime createTime;
+
+    @Column(name = "update_time", nullable = false)
+    private LocalDateTime updateTime;
+
+    @Transient
+    private List<FileDirectory> children;
+
+    @PrePersist
+    protected void onCreate() {
+        createTime = LocalDateTime.now();
+        updateTime = LocalDateTime.now();
+    }
+
+    @PreUpdate
+    protected void onUpdate() {
+        updateTime = LocalDateTime.now();
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/FileType.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/FileType.java
new file mode 100644
index 0000000..c35d09c
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/FileType.java
@@ -0,0 +1,112 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.entity.mysql
+ * @ClassName: FileType
+ * @Description: 文件类型实体类，映射t_file_type表，存储文件类型信息（如：投标文件、采购文件、结果文件等）
+ * @Author: 张志锋
+ * @Date: 2025-11-01
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.entity.mysql;
+
+import jakarta.persistence.*;
+import lombok.Data;
+import java.time.LocalDateTime;
+
+/**
+ * 文件类型实体类
+ */
+@Entity
+@Table(name = "t_file_type")
+@Data
+public class FileType {
+
+    @Id
+    @GeneratedValue(strategy = GenerationType.IDENTITY)
+    private Long id;
+
+    /**
+     * 文件类型代码（唯一，如：tender, purchase, result等）
+     */
+    @Column(name = "type_code", unique = true, nullable = false, length = 64)
+    private String typeCode;
+
+    /**
+     * 文件类型名称（如：投标文件、采购文件、结果文件等）
+     */
+    @Column(name = "type_name", nullable = false, length = 128)
+    private String typeName;
+
+    /**
+     * 文件类型描述
+     */
+    @Column(name = "description", columnDefinition = "TEXT")
+    private String description;
+
+    /**
+     * 识别关键词（JSON格式，用于自动识别文件类型）
+     * 例如：["投标", "标书", "tender"] 或 ["采购", "purchase", "采购文件"]
+     */
+    @Column(name = "keywords", columnDefinition = "TEXT")
+    private String keywords;
+
+    /**
+     * 文件扩展名（JSON格式，可选，如：["pdf", "doc", "docx"]）
+     */
+    @Column(name = "file_extensions", columnDefinition = "TEXT")
+    private String fileExtensions;
+
+    /**
+     * 优先级（数字越大优先级越高，用于匹配时优先使用）
+     */
+    @Column(name = "priority")
+    private Integer priority = 0;
+
+    /**
+     * 是否启用
+     */
+    @Column(name = "enabled", nullable = false)
+    private Boolean enabled = true;
+
+    /**
+     * 创建时间
+     */
+    @Column(name = "create_time", nullable = false)
+    private LocalDateTime createTime;
+
+    /**
+     * 更新时间
+     */
+    @Column(name = "update_time")
+    private LocalDateTime updateTime;
+
+    @PrePersist
+    public void prePersist() {
+        this.createTime = LocalDateTime.now();
+        this.updateTime = LocalDateTime.now();
+        if (this.typeCode == null || this.typeCode.isEmpty()) {
+            // 如果没有提供typeCode，从typeName生成
+            this.typeCode = generateTypeCode(this.typeName);
+        }
+    }
+
+    @PreUpdate
+    public void preUpdate() {
+        this.updateTime = LocalDateTime.now();
+    }
+
+    /**
+     * 从类型名称生成类型代码
+     */
+    private String generateTypeCode(String typeName) {
+        if (typeName == null || typeName.isEmpty()) {
+            return "unknown";
+        }
+        // 转换为小写，替换空格和下划线为连字符
+        return typeName.toLowerCase()
+                .replaceAll("\\s+", "_")
+                .replaceAll("[^a-z0-9_]", "");
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/ModelRequestLog.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/ModelRequestLog.java
new file mode 100644
index 0000000..ef80836
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/ModelRequestLog.java
@@ -0,0 +1,143 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.entity.mysql
+ * @ClassName: ModelRequestLog
+ * @Description: 模型请求日志实体，记录所有模型/向量化调用信息
+ * @Author: 张志锋
+ * @Date: 2025-11-25
+ * @Version: 1.0
+ * @Copyright: © 2025
+ */
+package com.zhpb.gdyd_zhpb_zgf.entity.mysql;
+
+import jakarta.persistence.*;
+import lombok.Data;
+
+import java.time.LocalDateTime;
+
+@Entity
+@Table(name = "t_model_request_log")
+@Data
+public class ModelRequestLog {
+
+    @Id
+    @GeneratedValue(strategy = GenerationType.IDENTITY)
+    private Long id;
+
+    /**
+     * 请求发起时间
+     */
+    @Column(name = "request_time", nullable = false)
+    private LocalDateTime requestTime;
+
+    /**
+     * 请求来源（QA、SUMMARY、VECTORIZE等）
+     */
+    @Column(name = "request_source", length = 64)
+    private String requestSource;
+
+    /**
+     * 请求地址
+     */
+    @Column(name = "request_url", length = 512)
+    private String requestUrl;
+
+    /**
+     * HTTP方法
+     */
+    @Column(name = "request_method", length = 16)
+    private String requestMethod;
+
+    /**
+     * 模型名称
+     */
+    @Column(name = "model_name", length = 128)
+    private String modelName;
+
+    /**
+     * 请求体
+     */
+    @Column(name = "request_payload", columnDefinition = "TEXT")
+    private String requestPayload;
+
+    /**
+     * 响应体
+     */
+    @Column(name = "response_payload", columnDefinition = "TEXT")
+    private String responsePayload;
+
+    /**
+     * 请求状态（SUCCESS/FAILED）
+     */
+    @Column(name = "status", length = 32)
+    private String status;
+
+    /**
+     * 耗时（毫秒）
+     */
+    @Column(name = "duration_ms")
+    private Long durationMs;
+
+    /**
+     * API密钥
+     */
+    @Column(name = "api_key", length = 256)
+    private String apiKey;
+
+    /**
+     * AppKey（私有/办公模型用）
+     */
+    @Column(name = "app_key", length = 256)
+    private String appKey;
+
+    /**
+     * AppSecret/AppSecret
+     */
+    @Column(name = "app_secret", length = 256)
+    private String appSecret;
+
+    /**
+     * Prompt Token数量
+     */
+    @Column(name = "prompt_tokens")
+    private Integer promptTokens;
+
+    /**
+     * Completion Token数量
+     */
+    @Column(name = "completion_tokens")
+    private Integer completionTokens;
+
+    /**
+     * 总Token数量
+     */
+    @Column(name = "total_tokens")
+    private Integer totalTokens;
+
+    /**
+     * TraceId
+     */
+    @Column(name = "trace_id", length = 128)
+    private String traceId;
+
+    /**
+     * 错误信息
+     */
+    @Column(name = "error_message", columnDefinition = "TEXT")
+    private String errorMessage;
+
+    /**
+     * 额外信息（用于存放header、扩展字段等）
+     */
+    @Column(name = "extra_info", columnDefinition = "TEXT")
+    private String extraInfo;
+
+    @PrePersist
+    public void prePersist() {
+        if (this.requestTime == null) {
+            this.requestTime = LocalDateTime.now();
+        }
+    }
+}
+
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/OcrResult.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/OcrResult.java
new file mode 100644
index 0000000..030b468
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/OcrResult.java
@@ -0,0 +1,118 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.entity.mysql
+ * @ClassName: OcrResult
+ * @Description: OCR识别结果实体类，映射t_ocr_result表
+ * @Author: 张志锋
+ * @Date: 2025-12-09
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.entity.mysql;
+
+import jakarta.persistence.*;
+import lombok.Data;
+import java.time.LocalDateTime;
+
+/**
+ * OCR识别结果实体类
+ */
+@Entity
+@Table(name = "t_ocr_result")
+@Data
+public class OcrResult {
+
+    @Id
+    @GeneratedValue(strategy = GenerationType.IDENTITY)
+    private Long id;
+
+    /**
+     * 文件ID（关联t_file表）
+     */
+    @Column(name = "file_id", nullable = false)
+    private Long fileId;
+
+    /**
+     * 项目ID
+     */
+    @Column(name = "project_id", nullable = false, length = 64)
+    private String projectId;
+
+    /**
+     * 供应商ID
+     */
+    @Column(name = "supplier_id", length = 64)
+    private String supplierId;
+
+    /**
+     * OCR任务ID
+     */
+    @Column(name = "ocr_task_id", length = 100)
+    private String ocrTaskId;
+
+    /**
+     * 识别状态（processing-处理中, success-成功, failed-失败）
+     */
+    @Column(name = "status", length = 32, columnDefinition = "varchar(32) default 'processing'")
+    private String status = "processing";
+
+    /**
+     * 识别文本内容
+     */
+    @Column(name = "recognized_text", columnDefinition = "TEXT")
+    private String recognizedText;
+
+    /**
+     * 下载链接（OCR结果文件下载链接）
+     */
+    @Column(name = "download_url", columnDefinition = "TEXT")
+    private String downloadUrl;
+
+    /**
+     * 下载文件名
+     */
+    @Column(name = "download_file_name", length = 512)
+    private String downloadFileName;
+
+    /**
+     * 处理页数
+     */
+    @Column(name = "processed_pages")
+    private Integer processedPages;
+
+    /**
+     * 总页数
+     */
+    @Column(name = "total_pages")
+    private Integer totalPages;
+
+    /**
+     * 错误信息
+     */
+    @Column(name = "error_message", columnDefinition = "TEXT")
+    private String errorMessage;
+
+    /**
+     * 开始时间
+     */
+    @Column(name = "start_time")
+    private LocalDateTime startTime;
+
+    /**
+     * 完成时间
+     */
+    @Column(name = "end_time")
+    private LocalDateTime endTime;
+
+    /**
+     * 创建时间
+     */
+    @Column(name = "create_time", nullable = false)
+    private LocalDateTime createTime = LocalDateTime.now();
+
+    /**
+     * 更新时间
+     */
+    @Column(name = "update_time", nullable = false)
+    private LocalDateTime updateTime = LocalDateTime.now();
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/Project.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/Project.java
new file mode 100644
index 0000000..58d603a
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/Project.java
@@ -0,0 +1,71 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.entity.mysql
+ * @ClassName: Project
+ * @Description: 项目实体类，映射t_project表，存储项目信息和文件路径列表
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.entity.mysql;
+
+import jakarta.persistence.*;
+import lombok.Data;
+import java.time.LocalDateTime;
+
+/**
+ * 项目实体类
+ */
+@Entity
+@Table(name = "t_project")
+@Data
+public class Project {
+
+    @Id
+    @GeneratedValue(strategy = GenerationType.IDENTITY)
+    private Long id;
+
+    /**
+     * 项目ID（UUID或业务ID）
+     */
+    @Column(name = "project_id", unique = true, nullable = false, length = 64)
+    private String projectId;
+
+    /**
+     * 项目名称
+     */
+    @Column(name = "project_name", length = 256)
+    private String projectName;
+
+    /**
+     * 项目描述
+     */
+    @Column(name = "description", columnDefinition = "TEXT")
+    private String description;
+
+    /**
+     * 项目状态
+     */
+    @Column(name = "status", length = 32)
+    private String status;
+
+    /**
+     * 文件路径（上传或解压的文件路径，JSON格式存储多个路径）
+     */
+    @Column(name = "file_paths", columnDefinition = "TEXT")
+    private String filePaths;
+
+    /**
+     * 创建时间
+     */
+    @Column(name = "create_time", nullable = false)
+    private LocalDateTime createTime;
+
+    /**
+     * 更新时间
+     */
+    @Column(name = "update_time")
+    private LocalDateTime updateTime;
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/PromptTemplate.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/PromptTemplate.java
new file mode 100644
index 0000000..786d213
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/PromptTemplate.java
@@ -0,0 +1,108 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.entity.mysql
+ * @ClassName: PromptTemplate
+ * @Description: 提示词模板实体类，映射t_prompt_template表，存储智能问答的提示词配置
+ * @Author: 张志锋
+ * @Date: 2025-11-03
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.entity.mysql;
+
+import jakarta.persistence.*;
+import lombok.Data;
+import java.time.LocalDateTime;
+
+/**
+ * 提示词模板实体类
+ */
+@Entity
+@Table(name = "t_prompt_template")
+@Data
+public class PromptTemplate {
+
+    @Id
+    @GeneratedValue(strategy = GenerationType.IDENTITY)
+    private Long id;
+
+    /**
+     * 模板ID（业务ID，唯一）
+     */
+    @Column(name = "template_id", unique = true, nullable = false, length = 64)
+    private String templateId;
+
+    /**
+     * 模板名称
+     */
+    @Column(name = "template_name", nullable = false, length = 256)
+    private String templateName;
+
+    /**
+     * 模板类型（qa-智能问答, summary-摘要, etc）
+     */
+    @Column(name = "template_type", nullable = false, length = 32)
+    private String templateType = "qa";
+
+    /**
+     * 提示词内容
+     */
+    @Column(name = "prompt_content", nullable = false, columnDefinition = "TEXT")
+    private String promptContent;
+
+    /**
+     * 模板描述
+     */
+    @Column(name = "description", columnDefinition = "TEXT")
+    private String description;
+
+    /**
+     * 是否启用
+     */
+    @Column(name = "is_active", nullable = false)
+    private Boolean isActive = true;
+
+    /**
+     * 排序号
+     */
+    @Column(name = "sort_order", nullable = false)
+    private Integer sortOrder = 0;
+
+    /**
+     * 创建人
+     */
+    @Column(name = "created_by", length = 64)
+    private String createdBy;
+
+    /**
+     * 更新人
+     */
+    @Column(name = "updated_by", length = 64)
+    private String updatedBy;
+
+    /**
+     * 创建时间
+     */
+    @Column(name = "create_time", nullable = false)
+    private LocalDateTime createTime;
+
+    /**
+     * 更新时间
+     */
+    @Column(name = "update_time")
+    private LocalDateTime updateTime;
+
+    @PrePersist
+    public void prePersist() {
+        this.createTime = LocalDateTime.now();
+        this.updateTime = LocalDateTime.now();
+        if (this.templateId == null) {
+            this.templateId = java.util.UUID.randomUUID().toString().replace("-", "");
+        }
+    }
+
+    @PreUpdate
+    public void preUpdate() {
+        this.updateTime = LocalDateTime.now();
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/QaMessage.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/QaMessage.java
new file mode 100644
index 0000000..a8292fa
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/QaMessage.java
@@ -0,0 +1,58 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.entity.mysql
+ * @ClassName: QaMessage
+ * @Description: QA消息实体
+ * @Author: 张志锋
+ * @Date: 2025-11-03
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.entity.mysql;
+
+import jakarta.persistence.*;
+import lombok.Data;
+
+import java.time.LocalDateTime;
+
+/**
+ * QA消息实体
+ */
+@Data
+@Entity
+@Table(name = "t_qa_message")
+public class QaMessage {
+
+    @Id
+    @GeneratedValue(strategy = GenerationType.IDENTITY)
+    private Long id;
+
+    @Column(name = "message_id", nullable = false, length = 100)
+    private String messageId;
+
+    @Column(name = "session_id", nullable = false, length = 100)
+    private String sessionId;
+
+    @Column(name = "sender", nullable = false, length = 20)
+    private String sender; // "user" or "ai"
+
+    @Column(name = "content", nullable = false, columnDefinition = "TEXT")
+    private String content;
+
+    @Column(name = "sources", columnDefinition = "TEXT")
+    private String sources; // JSON string
+
+    @Column(name = "file_ids", columnDefinition = "TEXT")
+    private String fileIds; // JSON string
+
+    @Column(name = "timestamp", nullable = false)
+    private LocalDateTime timestamp;
+
+    @Column(name = "create_time", nullable = false, updatable = false)
+    private LocalDateTime createTime;
+
+    @PrePersist
+    protected void onCreate() {
+        createTime = LocalDateTime.now();
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/QaSession.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/QaSession.java
new file mode 100644
index 0000000..b2c508d
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/QaSession.java
@@ -0,0 +1,70 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.entity.mysql
+ * @ClassName: QaSession
+ * @Description: QA会话实体
+ * @Author: 张志锋
+ * @Date: 2025-11-03
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.entity.mysql;
+
+import jakarta.persistence.*;
+import lombok.Data;
+
+import java.time.LocalDateTime;
+
+/**
+ * QA会话实体
+ */
+@Data
+@Entity
+@Table(name = "t_qa_session")
+public class QaSession {
+
+    @Id
+    @GeneratedValue(strategy = GenerationType.IDENTITY)
+    private Long id;
+
+    @Column(name = "session_id", nullable = false, length = 100)
+    private String sessionId;
+
+    @Column(name = "project_id", nullable = false, length = 50)
+    private String projectId;
+
+    @Column(name = "user_id", nullable = false, length = 50)
+    private String userId;
+
+    @Column(name = "first_query", columnDefinition = "TEXT")
+    private String firstQuery;
+
+    @Column(name = "start_time", nullable = false)
+    private LocalDateTime startTime;
+
+    @Column(name = "last_message_time")
+    private LocalDateTime lastMessageTime;
+
+    @Column(name = "message_count", columnDefinition = "int default 0")
+    private Integer messageCount = 0;
+
+    @Column(name = "status", length = 20, columnDefinition = "varchar(20) default 'active'")
+    private String status = "active";
+
+    @Column(name = "create_time", nullable = false, updatable = false)
+    private LocalDateTime createTime;
+
+    @Column(name = "update_time", nullable = false)
+    private LocalDateTime updateTime;
+
+    @PrePersist
+    protected void onCreate() {
+        createTime = LocalDateTime.now();
+        updateTime = LocalDateTime.now();
+    }
+
+    @PreUpdate
+    protected void onUpdate() {
+        updateTime = LocalDateTime.now();
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/SsoLog.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/SsoLog.java
new file mode 100644
index 0000000..81059f2
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/SsoLog.java
@@ -0,0 +1,65 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.entity.mysql
+ * @ClassName: SsoLog
+ * @Description: SSO登录日志实体类，映射t_sso_log表，记录SSO登录的详细信息
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.entity.mysql;
+
+import jakarta.persistence.*;
+import lombok.Data;
+import java.time.LocalDateTime;
+
+/**
+ * SSO登录日志实体类
+ */
+@Entity
+@Table(name = "t_sso_log")
+@Data
+public class SsoLog {
+    
+    @Id
+    @GeneratedValue(strategy = GenerationType.IDENTITY)
+    private Long id;
+    
+    @Column(name = "ticket", length = 256)
+    private String ticket;
+    
+    @Column(name = "user_id", length = 64)
+    private String userId;
+    
+    @Column(name = "user_name", length = 128)
+    private String userName;
+    
+    @Column(name = "project_id", length = 64)
+    private String projectId;
+    
+    @Column(name = "token", length = 512)
+    private String token;
+    
+    @Column(name = "login_ip", length = 64)
+    private String loginIp;
+    
+    @Column(name = "login_status", length = 32)
+    private String loginStatus;
+    
+    @Column(name = "error_msg", length = 512)
+    private String errorMsg;
+    
+    @Column(name = "request_url", length = 512)
+    private String requestUrl;
+    
+    @Column(name = "response_data", columnDefinition = "TEXT")
+    private String responseData;
+    
+    @Column(name = "login_time")
+    private LocalDateTime loginTime;
+    
+    @Column(name = "create_time")
+    private LocalDateTime createTime;
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/Supplier.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/Supplier.java
new file mode 100644
index 0000000..9a4a7a3
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/Supplier.java
@@ -0,0 +1,112 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.entity.mysql
+ * @ClassName: Supplier
+ * @Description: 供应商实体类，映射t_supplier表，存储供应商基本信息
+ * @Author: 张志锋
+ * @Date: 2025-11-01
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.entity.mysql;
+
+import jakarta.persistence.*;
+import lombok.Data;
+import java.time.LocalDateTime;
+
+/**
+ * 供应商实体类
+ */
+@Entity
+@Table(name = "t_supplier")
+@Data
+public class Supplier {
+
+    @Id
+    @GeneratedValue(strategy = GenerationType.IDENTITY)
+    private Long id;
+
+    /**
+     * 供应商ID（业务ID，唯一）
+     */
+    @Column(name = "supplier_id", unique = true, nullable = false, length = 64)
+    private String supplierId;
+
+    /**
+     * 供应商名称
+     */
+    @Column(name = "supplier_name", nullable = false, length = 256)
+    private String supplierName;
+
+    /**
+     * 供应商代码
+     */
+    @Column(name = "supplier_code", length = 128)
+    private String supplierCode;
+
+    /**
+     * 供应商类型
+     */
+    @Column(name = "supplier_type", length = 64)
+    private String supplierType;
+
+    /**
+     * 联系人
+     */
+    @Column(name = "contact_person", length = 128)
+    private String contactPerson;
+
+    /**
+     * 联系电话
+     */
+    @Column(name = "contact_phone", length = 64)
+    private String contactPhone;
+
+    /**
+     * 联系邮箱
+     */
+    @Column(name = "contact_email", length = 128)
+    private String contactEmail;
+
+    /**
+     * 供应商地址
+     */
+    @Column(name = "address", columnDefinition = "TEXT")
+    private String address;
+
+    /**
+     * 供应商状态（active, inactive等）
+     */
+    @Column(name = "status", length = 32)
+    private String status;
+
+    /**
+     * 备注
+     */
+    @Column(name = "remarks", columnDefinition = "TEXT")
+    private String remarks;
+
+    /**
+     * 创建时间
+     */
+    @Column(name = "create_time", nullable = false)
+    private LocalDateTime createTime;
+
+    /**
+     * 更新时间
+     */
+    @Column(name = "update_time")
+    private LocalDateTime updateTime;
+
+    @PrePersist
+    public void prePersist() {
+        this.createTime = LocalDateTime.now();
+        this.updateTime = LocalDateTime.now();
+    }
+
+    @PreUpdate
+    public void preUpdate() {
+        this.updateTime = LocalDateTime.now();
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/SysLog.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/SysLog.java
new file mode 100644
index 0000000..872d948
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/SysLog.java
@@ -0,0 +1,119 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.entity.mysql
+ * @ClassName: SysLog
+ * @Description: 系统日志实体类，映射t_sys_log表，记录所有控制台输出和系统日志
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.entity.mysql;
+
+import jakarta.persistence.*;
+import lombok.Data;
+
+import java.time.LocalDateTime;
+
+/**
+ * 系统日志实体类
+ */
+@Entity
+@Table(name = "t_sys_log", indexes = {
+        @Index(name = "idx_log_level", columnList = "log_level"),
+        @Index(name = "idx_create_time", columnList = "create_time"),
+        @Index(name = "idx_thread_name", columnList = "thread_name")
+})
+@Data
+public class SysLog {
+
+    @Id
+    @GeneratedValue(strategy = GenerationType.IDENTITY)
+    private Long id;
+
+    /**
+     * 日志级别（INFO, DEBUG, WARN, ERROR, OUT等）
+     */
+    @Column(name = "log_level", length = 20, nullable = false)
+    private String logLevel;
+
+    /**
+     * 日志来源（logger名称或SYSTEM_OUT）
+     */
+    @Column(name = "logger_name", length = 255)
+    private String loggerName;
+
+    /**
+     * 线程名称
+     */
+    @Column(name = "thread_name", length = 255)
+    private String threadName;
+
+    /**
+     * 日志内容
+     */
+    @Column(name = "log_content", columnDefinition = "TEXT", nullable = false)
+    private String logContent;
+
+    /**
+     * 完整的日志消息（包含格式化信息）
+     */
+    @Column(name = "full_message", columnDefinition = "TEXT")
+    private String fullMessage;
+
+    /**
+     * 异常堆栈信息（如果有）
+     */
+    @Column(name = "exception_stack", columnDefinition = "TEXT")
+    private String exceptionStack;
+
+    /**
+     * 类名
+     */
+    @Column(name = "class_name", length = 500)
+    private String className;
+
+    /**
+     * 方法名
+     */
+    @Column(name = "method_name", length = 255)
+    private String methodName;
+
+    /**
+     * 行号
+     */
+    @Column(name = "line_number")
+    private Integer lineNumber;
+
+    /**
+     * 创建时间
+     */
+    @Column(name = "create_time", nullable = false)
+    private LocalDateTime createTime;
+
+    /**
+     * 应用名称
+     */
+    @Column(name = "application_name", length = 100)
+    private String applicationName;
+
+    /**
+     * 用户ID（如果有）
+     */
+    @Column(name = "user_id", length = 64)
+    private String userId;
+
+    /**
+     * 请求ID或跟踪ID（如果有）
+     */
+    @Column(name = "trace_id", length = 64)
+    private String traceId;
+
+    @PrePersist
+    public void prePersist() {
+        if (this.createTime == null) {
+            this.createTime = LocalDateTime.now();
+        }
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/User.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/User.java
new file mode 100644
index 0000000..1119ffb
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/User.java
@@ -0,0 +1,47 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.entity.mysql
+ * @ClassName: User
+ * @Description: MySQL用户实体类，映射t_user表，存储用户基本信息
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.entity.mysql;
+
+import jakarta.persistence.*;
+import lombok.Data;
+import java.time.LocalDateTime;
+
+/**
+ * MySQL 用户实体类
+ */
+@Entity
+@Table(name = "t_user")
+@Data
+public class User {
+    
+    @Id
+    @GeneratedValue(strategy = GenerationType.IDENTITY)
+    private Long id;
+    
+    @Column(name = "user_id", unique = true, nullable = false, length = 64)
+    private String userId;
+    
+    @Column(name = "name", nullable = false, length = 128)
+    private String name;
+    
+    @Column(name = "project_id", length = 64)
+    private String projectId;
+    
+    @Column(name = "status", length = 16)
+    private String status;
+    
+    @Column(name = "create_time")
+    private LocalDateTime createTime;
+    
+    @Column(name = "update_time")
+    private LocalDateTime updateTime;
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/UserLog.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/UserLog.java
new file mode 100644
index 0000000..78afd14
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/mysql/UserLog.java
@@ -0,0 +1,88 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.entity.mysql
+ * @ClassName: UserLog
+ * @Description: 用户操作日志实体类，映射t_user_log表，记录用户的关键操作信息
+ * @Author: 
+ * @Date: 2025-11-28
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.entity.mysql;
+
+import jakarta.persistence.*;
+import lombok.Data;
+import java.time.LocalDateTime;
+
+/**
+ * 用户操作日志实体类
+ */
+@Entity
+@Table(name = "t_user_log")
+@Data
+public class UserLog {
+
+    @Id
+    @GeneratedValue(strategy = GenerationType.IDENTITY)
+    private Long id;
+
+    /**
+     * 用户ID
+     */
+    @Column(name = "user_id", nullable = false, length = 64)
+    private String userId;
+
+    /**
+     * 用户名
+     */
+    @Column(name = "user_name", length = 128)
+    private String userName;
+
+    /**
+     * 项目ID
+     */
+    @Column(name = "project_id", length = 64)
+    private String projectId;
+
+    /**
+     * 操作类型
+     * UPLOAD_FILE-上传文件
+     * QA_QUESTION-智能问答
+     * QA_MESSAGE-问答消息
+     * VIEW_CATALOG-查看目录
+     * CATALOG_SUMMARY-目录总结
+     * MODEL_REPLY-大模型回复
+     */
+    @Column(name = "operation_type", nullable = false, length = 64)
+    private String operationType;
+
+    /**
+     * 操作详情
+     */
+    @Column(name = "operation_detail", columnDefinition = "TEXT")
+    private String operationDetail;
+
+    /**
+     * 关联ID（如文件ID、会话ID等）
+     */
+    @Column(name = "related_id", length = 64)
+    private String relatedId;
+
+    /**
+     * 操作IP地址
+     */
+    @Column(name = "ip_address", length = 64)
+    private String ipAddress;
+
+    /**
+     * 用户代理
+     */
+    @Column(name = "user_agent", length = 512)
+    private String userAgent;
+
+    /**
+     * 创建时间
+     */
+    @Column(name = "create_time", nullable = false)
+    private LocalDateTime createTime = LocalDateTime.now();
+}
\ No newline at end of file
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/pgsql/PdfVector.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/pgsql/PdfVector.java
new file mode 100644
index 0000000..29fc21d
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/pgsql/PdfVector.java
@@ -0,0 +1,175 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.entity.pgsql
+ * @ClassName: PdfVector
+ * @Description: PDF向量化实体类，映射t_pdf_vector表，存储PDF文件的向量化结果（使用PostgreSQL的pgvector扩展）
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.entity.pgsql;
+
+import jakarta.persistence.*;
+import lombok.Data;
+
+import java.time.LocalDateTime;
+
+/**
+ * PDF向量化实体类
+ */
+@Entity
+@Table(name = "t_pdf_vector")
+@Data
+public class PdfVector {
+
+    @Id
+    @GeneratedValue(strategy = GenerationType.IDENTITY)
+    private Long id;
+
+    /**
+     * PDF文件路径
+     */
+    @Column(name = "file_path", nullable = false, length = 1000)
+    private String filePath;
+
+    /**
+     * PDF文件名
+     */
+    @Column(name = "file_name", nullable = false, length = 500)
+    private String fileName;
+
+    /**
+     * PDF原始文本内容（可存储部分内容或摘要）
+     */
+    @Column(name = "content", columnDefinition = "TEXT")
+    private String content;
+
+    /**
+     * 向量化后的向量（使用pgvector类型存储）
+     * 注意：Hibernate无法自动创建vector类型，表通过原生SQL手动创建
+     * 此字段不参与Hibernate的DDL，完全通过原生SQL操作
+     */
+    @Transient
+    private String vectorText;
+    
+    /**
+     * 向量数组（临时字段，不直接存储到数据库，通过vectorText间接存储）
+     */
+    @Transient
+    private float[] vector;
+    
+    /**
+     * 设置向量数组（自动转换为字符串）
+     */
+    public void setVector(float[] vector) {
+        this.vector = vector;
+        if (vector != null && vector.length > 0) {
+            StringBuilder sb = new StringBuilder("[");
+            for (int i = 0; i < vector.length; i++) {
+                if (i > 0) {
+                    sb.append(",");
+                }
+                sb.append(vector[i]);
+            }
+            sb.append("]");
+            this.vectorText = sb.toString();
+        }
+    }
+    
+    /**
+     * 获取向量数组（从字符串解析）
+     */
+    public float[] getVector() {
+        if (vector == null && vectorText != null && !vectorText.isEmpty()) {
+            try {
+                String content = vectorText.trim();
+                if (content.startsWith("[") && content.endsWith("]")) {
+                    content = content.substring(1, content.length() - 1);
+                    String[] parts = content.split(",");
+                    vector = new float[parts.length];
+                    for (int i = 0; i < parts.length; i++) {
+                        vector[i] = Float.parseFloat(parts[i].trim());
+                    }
+                }
+            } catch (Exception e) {
+                // 解析失败
+            }
+        }
+        return vector;
+    }
+    
+    /**
+     * 向量字符串格式（用于原生SQL查询）
+     * 格式：[0.1,0.2,0.3,...]
+     */
+    @Transient
+    public String getVectorString() {
+        if (vector == null || vector.length == 0) {
+            return "[]";
+        }
+        StringBuilder sb = new StringBuilder("[");
+        for (int i = 0; i < vector.length; i++) {
+            if (i > 0) {
+                sb.append(",");
+            }
+            sb.append(vector[i]);
+        }
+        sb.append("]");
+        return sb.toString();
+    }
+
+    /**
+     * 向量维度
+     */
+    @Column(name = "dimension", nullable = false)
+    private Integer dimension;
+
+    /**
+     * 文件ID（关联t_file表的file_id）
+     */
+    @Column(name = "file_id", length = 64)
+    private String fileId;
+
+    /**
+     * 项目ID（关联t_project表的project_id）
+     */
+    @Column(name = "project_id", length = 100)
+    private String projectId;
+
+    /**
+     * 供应商ID（关联t_supplier表的supplier_id）
+     */
+    @Column(name = "supplier_id", length = 64)
+    private String supplierId;
+
+    /**
+     * 文件类型ID（关联t_file_type表的id）
+     */
+    @Column(name = "file_type_id")
+    private Long fileTypeId;
+
+    /**
+     * 创建时间
+     */
+    @Column(name = "create_time", nullable = false)
+    private LocalDateTime createTime;
+
+    /**
+     * 更新时间
+     */
+    @Column(name = "update_time")
+    private LocalDateTime updateTime;
+
+    @PrePersist
+    public void prePersist() {
+        this.createTime = LocalDateTime.now();
+        this.updateTime = LocalDateTime.now();
+    }
+
+    @PreUpdate
+    public void preUpdate() {
+        this.updateTime = LocalDateTime.now();
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/pgsql/Product.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/pgsql/Product.java
new file mode 100644
index 0000000..997fe5a
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/entity/pgsql/Product.java
@@ -0,0 +1,34 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.entity.pgsql
+ * @ClassName: Product
+ * @Description: PostgreSQL产品实体类示例，映射t_product表
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.entity.pgsql;
+
+import jakarta.persistence.*;
+import lombok.Data;
+
+/**
+ * PostgreSQL 产品实体类示例
+ */
+@Entity
+@Table(name = "t_product")
+@Data
+public class Product {
+    
+    @Id
+    @GeneratedValue(strategy = GenerationType.IDENTITY)
+    private Long id;
+    
+    @Column(name = "name")
+    private String name;
+    
+    @Column(name = "price")
+    private Double price;
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/exception/GlobalExceptionHandler.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/exception/GlobalExceptionHandler.java
new file mode 100644
index 0000000..c341b66
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/exception/GlobalExceptionHandler.java
@@ -0,0 +1,46 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.exception
+ * @ClassName: GlobalExceptionHandler
+ * @Description: 全局异常处理器（暂时禁用，避免SpringDoc兼容性问题）
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.exception;
+
+/**
+ * 全局异常处理器
+ * 注意：为了避免SpringDoc兼容性问题，暂时禁用此处理器
+ * 文件上传大小限制异常已在FileController中直接处理
+ * 
+ * 如果未来SpringDoc版本修复了兼容性问题，可以重新启用：
+ * - 取消注释 @RestControllerAdvice
+ * - 取消注释 @ExceptionHandler 方法
+ */
+// 暂时禁用，避免SpringDoc 2.6.0与Spring Boot 3.5.7的兼容性问题
+/*
+@RestControllerAdvice
+public class GlobalExceptionHandler {
+    
+    private static final Logger logger = LoggerFactory.getLogger(GlobalExceptionHandler.class);
+    
+    @ExceptionHandler(MaxUploadSizeExceededException.class)
+    public ResponseEntity<Map<String, Object>> handleMaxUploadSizeExceededException(
+            MaxUploadSizeExceededException e) {
+        logger.warn("文件上传大小超限: {}", e.getMessage());
+        
+        Map<String, Object> response = new HashMap<>();
+        response.put("code", 413);
+        response.put("message", "文件大小超过限制，最大允许上传500MB");
+        response.put("error", "MaxUploadSizeExceeded");
+        
+        return ResponseEntity.status(HttpStatus.PAYLOAD_TOO_LARGE).body(response);
+    }
+}
+*/
+public class GlobalExceptionHandler {
+    // 暂时为空类，保留以备将来使用
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/filter/ApiStatisticsFilter.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/filter/ApiStatisticsFilter.java
new file mode 100644
index 0000000..b1417eb
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/filter/ApiStatisticsFilter.java
@@ -0,0 +1,40 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.filter
+ * @ClassName: ApiStatisticsFilter
+ * @Description: API统计过滤器（已废弃，现在由ApiLogInterceptor处理日志记录）
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.filter;
+
+import jakarta.servlet.FilterChain;
+import jakarta.servlet.ServletException;
+import jakarta.servlet.http.HttpServletRequest;
+import jakarta.servlet.http.HttpServletResponse;
+import org.springframework.stereotype.Component;
+import org.springframework.web.filter.OncePerRequestFilter;
+
+import java.io.IOException;
+
+/**
+ * API访问统计过滤器（已废弃）
+ * 统计功能已由 ApiLogInterceptor 接管，日志保存在数据库表中
+ * 保留此类是为了避免删除后的兼容性问题
+ * 
+ * @deprecated 使用 {@link com.zhpb.gdyd_zhpb_zgf.interceptor.ApiLogInterceptor} 代替
+ */
+@Component
+@Deprecated
+public class ApiStatisticsFilter extends OncePerRequestFilter {
+
+    @Override
+    protected void doFilterInternal(HttpServletRequest request, HttpServletResponse response, FilterChain filterChain)
+            throws ServletException, IOException {
+        // 不再进行任何统计操作，直接放行
+        filterChain.doFilter(request, response);
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/interceptor/ApiLogInterceptor.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/interceptor/ApiLogInterceptor.java
new file mode 100644
index 0000000..dba1f33
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/interceptor/ApiLogInterceptor.java
@@ -0,0 +1,253 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.interceptor
+ * @ClassName: ApiLogInterceptor
+ * @Description: API日志拦截器，拦截所有API请求并记录详细的调用信息到数据库
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.interceptor;
+
+import com.alibaba.fastjson2.JSON;
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.ApiLog;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.ApiLogRepository;
+import jakarta.servlet.http.HttpServletRequest;
+import jakarta.servlet.http.HttpServletResponse;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.stereotype.Component;
+import org.springframework.web.servlet.HandlerInterceptor;
+import org.springframework.web.servlet.ModelAndView;
+import org.springframework.web.util.ContentCachingRequestWrapper;
+import org.springframework.web.util.ContentCachingResponseWrapper;
+
+import java.nio.charset.StandardCharsets;
+import java.time.LocalDateTime;
+import java.util.HashMap;
+import java.util.Map;
+import java.util.concurrent.CompletableFuture;
+
+/**
+ * API调用日志拦截器
+ */
+@Component
+public class ApiLogInterceptor implements HandlerInterceptor {
+
+    @Autowired
+    private ApiLogRepository apiLogRepository;
+
+    // 接口名称映射
+    private static final Map<String, String> API_NAME_MAP = new HashMap<String, String>() {{
+        put("/api/v1/sso/login", "SSO登录接口");
+        put("/api/v1/statistics/total", "获取API总访问次数");
+        put("/api/v1/statistics/count", "获取特定接口访问次数");
+        put("/api/v1/files/uploadzip", "上传ZIP压缩包接口");
+        put("/health", "健康检查");
+        put("/actuator/health", "健康检查");
+        put("/", "首页");
+        put("/swagger-ui.html", "Swagger文档");
+        put("/api-docs", "API文档JSON");
+    }};
+
+    @Override
+    public boolean preHandle(HttpServletRequest request, HttpServletResponse response, Object handler) throws Exception {
+        long startTime = System.currentTimeMillis();
+        request.setAttribute("startTime", startTime);
+        return true;
+    }
+
+    @Override
+    public void postHandle(HttpServletRequest request, HttpServletResponse response, Object handler, ModelAndView modelAndView) throws Exception {
+        // 不做处理
+    }
+
+    @Override
+    public void afterCompletion(HttpServletRequest request, HttpServletResponse response, Object handler, Exception ex) throws Exception {
+        long startTime = (Long) request.getAttribute("startTime");
+        long endTime = System.currentTimeMillis();
+        long duration = endTime - startTime;
+
+        String path = request.getRequestURI();
+        String method = request.getMethod();
+
+        // 排除健康检查页面发起的请求（避免产生大量无意义的日志）
+        // 健康检查页面通常使用HEAD/OPTIONS请求来检查接口状态，不应该记录日志
+        if ("HEAD".equalsIgnoreCase(method) || "OPTIONS".equalsIgnoreCase(method)) {
+            return;
+        }
+
+        // 只记录API接口（以/api/开头的路径）和健康检查接口，排除静态资源
+        if (!path.startsWith("/api/") && 
+            !path.equals("/health") && 
+            !path.equals("/actuator/health") &&
+            !path.startsWith("/swagger") &&
+            !path.startsWith("/webjars") &&
+            !path.startsWith("/v3/api-docs")) {
+            return;
+        }
+
+        // 在异步执行前，同步获取所有需要的信息（避免响应对象被回收）
+        int statusCode = response.getStatus();
+        String clientIp = getClientIp(request);
+        String apiName = getApiName(path);
+        String requestParams = getRequestParams(request);
+
+        // 获取请求体和响应体的内容（需要在响应对象被回收前获取）
+        String requestBody = null;
+        String responseBody = null;
+        
+        if (request instanceof ContentCachingRequestWrapper) {
+            ContentCachingRequestWrapper wrappedRequest = (ContentCachingRequestWrapper) request;
+            requestBody = getRequestBody(wrappedRequest);
+        }
+        
+        if (response instanceof ContentCachingResponseWrapper) {
+            ContentCachingResponseWrapper wrappedResponse = (ContentCachingResponseWrapper) response;
+            try {
+                responseBody = getResponseBody(wrappedResponse);
+            } catch (Exception e) {
+                // 如果读取响应体失败，只记录警告，不阻塞
+                System.err.println("读取响应体失败: " + e.getMessage());
+            }
+        }
+
+        // 获取异常信息
+        String errorMsg = ex != null ? ex.getMessage() : null;
+
+        // 使用获取到的值创建最终变量（避免在异步线程中访问响应对象）
+        final int finalStatusCode = statusCode;
+        final String finalClientIp = clientIp;
+        final String finalApiName = apiName;
+        final String finalRequestParams = requestParams;
+        final String finalRequestBody = requestBody;
+        final String finalResponseBody = responseBody;
+        final String finalErrorMsg = errorMsg;
+        final String finalPath = path;
+        final String finalMethod = method;
+        final long finalDuration = duration;
+
+        // 异步保存日志，避免阻塞请求
+        CompletableFuture.runAsync(() -> {
+            try {
+                ApiLog apiLog = new ApiLog();
+                // 从请求参数中提取userId和userName（处理不同接口的参数命名差异）
+                String userId = null;
+                String userName = null;
+                if (finalRequestParams != null) {
+                    try {
+                        Map<String, Object> params = JSON.parseObject(finalRequestParams, Map.class);
+                        // 尝试多种参数命名方式
+                        userId = (String) params.get("userId");
+                        if (userId == null) {
+                            userId = (String) params.get("user_id");
+                        }
+                        if (userId == null) {
+                            userId = (String) params.get("userld");
+                        }
+                        
+                        userName = (String) params.get("userName");
+                        if (userName == null) {
+                            userName = (String) params.get("user_name");
+                        }
+                    } catch (Exception e) {
+                        // 解析失败忽略
+                    }
+                }
+                
+                apiLog.setApiPath(finalPath);
+                apiLog.setApiName(finalApiName);
+                apiLog.setHttpMethod(finalMethod);
+                apiLog.setStatusCode(finalStatusCode);
+                apiLog.setDuration(finalDuration);
+                apiLog.setClientIp(finalClientIp);
+                apiLog.setUserId(userId);
+                apiLog.setUserName(userName);
+                apiLog.setCreateTime(LocalDateTime.now());
+                apiLog.setRequestParams(finalRequestParams);
+                apiLog.setRequestBody(finalRequestBody);
+                apiLog.setResponseBody(finalResponseBody);
+                apiLog.setErrorMsg(finalErrorMsg);
+
+                // 保存到数据库
+                apiLogRepository.save(apiLog);
+            } catch (Exception e) {
+                // 记录日志失败不影响主流程，只打印错误
+                System.err.println("保存API日志失败: " + e.getMessage());
+                e.printStackTrace();
+            }
+        });
+    }
+
+    /**
+     * 获取接口中文名称
+     */
+    private String getApiName(String path) {
+        return API_NAME_MAP.getOrDefault(path, path);
+    }
+
+    /**
+     * 获取请求参数
+     */
+    private String getRequestParams(HttpServletRequest request) {
+        Map<String, String[]> parameterMap = request.getParameterMap();
+        if (parameterMap.isEmpty()) {
+            return null;
+        }
+        
+        Map<String, Object> params = new HashMap<>();
+        for (Map.Entry<String, String[]> entry : parameterMap.entrySet()) {
+            String[] values = entry.getValue();
+            if (values.length == 1) {
+                params.put(entry.getKey(), values[0]);
+            } else {
+                params.put(entry.getKey(), values);
+            }
+        }
+        return params.isEmpty() ? null : JSON.toJSONString(params);
+    }
+
+    /**
+     * 获取请求体
+     */
+    private String getRequestBody(ContentCachingRequestWrapper request) {
+        byte[] content = request.getContentAsByteArray();
+        if (content.length == 0) {
+            return null;
+        }
+        String body = new String(content, StandardCharsets.UTF_8);
+        // 限制请求体长度，避免存储过大的数据
+        return body.length() > 10000 ? body.substring(0, 10000) + "...(truncated)" : body;
+    }
+
+    /**
+     * 获取响应体
+     */
+    private String getResponseBody(ContentCachingResponseWrapper response) {
+        byte[] content = response.getContentAsByteArray();
+        if (content.length == 0) {
+            return null;
+        }
+        String body = new String(content, StandardCharsets.UTF_8);
+        // 限制响应体长度，避免存储过大的数据
+        return body.length() > 10000 ? body.substring(0, 10000) + "...(truncated)" : body;
+    }
+
+    /**
+     * 获取客户端IP地址
+     */
+    private String getClientIp(HttpServletRequest request) {
+        String ip = request.getHeader("X-Forwarded-For");
+        if (ip == null || ip.isEmpty() || "unknown".equalsIgnoreCase(ip)) {
+            ip = request.getHeader("Proxy-Client-IP");
+        }
+        if (ip == null || ip.isEmpty() || "unknown".equalsIgnoreCase(ip)) {
+            ip = request.getHeader("WL-Proxy-Client-IP");
+        }
+        if (ip == null || ip.isEmpty() || "unknown".equalsIgnoreCase(ip)) {
+            ip = request.getRemoteAddr();
+        }
+        return ip;
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/ApiLogRepository.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/ApiLogRepository.java
new file mode 100644
index 0000000..57d8ca7
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/ApiLogRepository.java
@@ -0,0 +1,42 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.repository.mysql
+ * @ClassName: ApiLogRepository
+ * @Description: API调用日志Repository接口，提供API日志数据访问和统计方法
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.repository.mysql;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.ApiLog;
+import org.springframework.data.jpa.repository.JpaRepository;
+import org.springframework.data.jpa.repository.Query;
+import org.springframework.data.repository.query.Param;
+import org.springframework.stereotype.Repository;
+
+/**
+ * API调用日志 Repository
+ */
+@Repository
+public interface ApiLogRepository extends JpaRepository<ApiLog, Long> {
+
+    /**
+     * 获取总访问次数
+     *
+     * @return 总访问次数
+     */
+    @Query("SELECT COUNT(*) FROM ApiLog")
+    Long countTotal();
+
+    /**
+     * 获取特定接口的访问次数
+     *
+     * @param apiPath 接口路径
+     * @return 访问次数
+     */
+    @Query("SELECT COUNT(*) FROM ApiLog WHERE apiPath = :apiPath")
+    Long countByApiPath(@Param("apiPath") String apiPath);
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/ComparisonTaskRepository.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/ComparisonTaskRepository.java
new file mode 100644
index 0000000..74b7f76
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/ComparisonTaskRepository.java
@@ -0,0 +1,68 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.repository.mysql
+ * @ClassName: ComparisonTaskRepository
+ * @Description: 对比分析任务Repository接口
+ * @Author: 张志锋
+ * @Date: 2025-11-04
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.repository.mysql;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.ComparisonTask;
+import org.springframework.data.jpa.repository.JpaRepository;
+import org.springframework.data.jpa.repository.Query;
+import org.springframework.data.repository.query.Param;
+import org.springframework.stereotype.Repository;
+
+import java.util.List;
+
+/**
+ * 对比分析任务 Repository
+ */
+@Repository
+public interface ComparisonTaskRepository extends JpaRepository<ComparisonTask, Long> {
+
+    /**
+     * 根据任务ID查找任务
+     *
+     * @param taskId 任务ID
+     * @return 对比分析任务
+     */
+    ComparisonTask findByTaskId(String taskId);
+
+    /**
+     * 根据项目ID和用户ID查找任务列表
+     *
+     * @param projectId 项目ID
+     * @param userId 用户ID
+     * @return 任务列表
+     */
+    @Query("SELECT t FROM ComparisonTask t WHERE t.projectId = :projectId AND t.userId = :userId ORDER BY t.createTime DESC")
+    List<ComparisonTask> findByProjectIdAndUserIdOrderByCreateTimeDesc(@Param("projectId") String projectId, @Param("userId") String userId);
+
+    /**
+     * 根据项目ID查找任务列表
+     *
+     * @param projectId 项目ID
+     * @return 任务列表
+     */
+    List<ComparisonTask> findByProjectIdOrderByCreateTimeDesc(String projectId);
+
+    /**
+     * 根据用户ID查找任务列表
+     *
+     * @param userId 用户ID
+     * @return 任务列表
+     */
+    List<ComparisonTask> findByUserIdOrderByCreateTimeDesc(String userId);
+
+    /**
+     * 根据状态查找任务列表
+     *
+     * @param status 任务状态
+     * @return 任务列表
+     */
+    List<ComparisonTask> findByStatusOrderByCreateTimeDesc(String status);
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/FileDirectoryRepository.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/FileDirectoryRepository.java
new file mode 100644
index 0000000..5842009
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/FileDirectoryRepository.java
@@ -0,0 +1,66 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.repository.mysql
+ * @ClassName: FileDirectoryRepository
+ * @Description: 文件目录Repository
+ * @Author: 张志锋
+ * @Date: 2025-11-03
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.repository.mysql;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.FileDirectory;
+import org.springframework.data.jpa.repository.JpaRepository;
+import org.springframework.data.jpa.repository.Modifying;
+import org.springframework.data.jpa.repository.Query;
+import org.springframework.data.repository.query.Param;
+import org.springframework.stereotype.Repository;
+
+import java.util.List;
+
+/**
+ * 文件目录Repository
+ */
+@Repository
+public interface FileDirectoryRepository extends JpaRepository<FileDirectory, Long> {
+
+    /**
+     * 根据文件ID查找目录结构
+     */
+    List<FileDirectory> findByFileIdOrderBySortOrderAsc(String fileId);
+
+    /**
+     * 根据文件ID和父章节ID查找子章节
+     */
+    List<FileDirectory> findByFileIdAndParentChapterIdOrderBySortOrderAsc(String fileId, String parentChapterId);
+
+    /**
+     * 根据文件ID和章节ID查找章节
+     */
+    FileDirectory findByFileIdAndChapterId(String fileId, String chapterId);
+
+    /**
+     * 根据文件ID删除所有目录
+     */
+    @Modifying
+    @Query("DELETE FROM FileDirectory fd WHERE fd.fileId = :fileId")
+    void deleteByFileId(@Param("fileId") String fileId);
+
+    /**
+     * 根据文件ID统计章节数量
+     */
+    long countByFileId(String fileId);
+
+    /**
+     * 根据文件ID查找顶级章节（无父章节）
+     */
+    List<FileDirectory> findByFileIdAndParentChapterIdIsNullOrderBySortOrderAsc(String fileId);
+
+    /**
+     * 查询摘要为空或为null的目录记录
+     * 用于定时任务扫描需要生成摘要的章节
+     */
+    @Query("SELECT fd FROM FileDirectory fd WHERE fd.summary IS NULL OR fd.summary = '' ORDER BY fd.createTime ASC")
+    List<FileDirectory> findBySummaryIsNullOrEmpty();
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/FileRepository.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/FileRepository.java
new file mode 100644
index 0000000..3747e66
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/FileRepository.java
@@ -0,0 +1,134 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.repository.mysql
+ * @ClassName: FileRepository
+ * @Description: 文件Repository接口
+ * @Author: 张志锋
+ * @Date: 2025-11-01
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.repository.mysql;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.File;
+import org.springframework.data.jpa.repository.JpaRepository;
+import org.springframework.stereotype.Repository;
+
+import java.util.List;
+import java.util.Optional;
+
+/**
+ * 文件 Repository
+ */
+@Repository
+public interface FileRepository extends JpaRepository<File, Long> {
+
+    /**
+     * 根据文件ID查找文件
+     *
+     * @param fileId 文件ID
+     * @return 文件对象
+     */
+    Optional<File> findByFileId(String fileId);
+
+    /**
+     * 根据项目ID查找所有文件
+     *
+     * @param projectId 项目ID
+     * @return 文件列表
+     */
+    List<File> findByProjectId(String projectId);
+
+    /**
+     * 根据供应商ID查找所有文件
+     *
+     * @param supplierId 供应商ID
+     * @return 文件列表
+     */
+    List<File> findBySupplierId(String supplierId);
+
+    /**
+     * 根据文件路径查找文件
+     *
+     * @param filePath 文件路径
+     * @return 文件对象
+     */
+    Optional<File> findByFilePath(String filePath);
+
+    /**
+     * 根据项目ID和供应商ID查找所有文件
+     *
+     * @param projectId 项目ID
+     * @param supplierId 供应商ID
+     * @return 文件列表
+     */
+    List<File> findByProjectIdAndSupplierId(String projectId, String supplierId);
+
+    /**
+     * 根据项目ID和是否已向量化查找文件
+     *
+     * @param projectId 项目ID
+     * @param isVectorized 是否已向量化
+     * @return 文件列表
+     */
+    List<File> findByProjectIdAndIsVectorized(String projectId, Boolean isVectorized);
+
+    /**
+     * 根据项目ID和文件名查找文件
+     *
+     * @param projectId 项目ID
+     * @param fileName 文件名
+     * @return 文件列表
+     */
+    List<File> findByProjectIdAndFileName(String projectId, String fileName);
+
+    /**
+     * 根据目录摘取状态和更新时间查找文件（用于查找卡住的任务）
+     *
+     * @param status 目录摘取状态
+     * @param updateTime 更新时间截止点
+     * @return 文件列表
+     */
+    List<File> findByCatalogExtractionStatusAndUpdateTimeBefore(String status, java.time.LocalDateTime updateTime);
+
+    /**
+     * 根据目录摘取状态查找文件
+     */
+    List<File> findByCatalogExtractionStatus(String status);
+
+    /**
+     * 根据项目ID和目录摘取状态查找文件
+     */
+    List<File> findByProjectIdAndCatalogExtractionStatus(String projectId, String status);
+
+    /**
+     * 根据向量化状态查找文件
+     */
+    List<File> findByVectorizationStatus(String status);
+
+    /**
+     * 根据项目ID和向量化状态查找文件
+     */
+    List<File> findByProjectIdAndVectorizationStatus(String projectId, String status);
+
+    /**
+     * 统计目录摘取状态的文件数量
+     */
+    long countByCatalogExtractionStatus(String status);
+
+    /**
+     * 统计项目下目录摘取状态的文件数量
+     */
+    long countByProjectIdAndCatalogExtractionStatus(String projectId, String status);
+
+    /**
+     * 统计向量化状态的文件数量
+     */
+    long countByVectorizationStatus(String status);
+
+    /**
+     * 统计项目下向量化状态的文件数量
+     */
+    long countByProjectIdAndVectorizationStatus(String projectId, String status);
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/FileTypeRepository.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/FileTypeRepository.java
new file mode 100644
index 0000000..348b2cd
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/FileTypeRepository.java
@@ -0,0 +1,57 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.repository.mysql
+ * @ClassName: FileTypeRepository
+ * @Description: 文件类型Repository接口
+ * @Author: 张志锋
+ * @Date: 2025-11-01
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.repository.mysql;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.FileType;
+import org.springframework.data.jpa.repository.JpaRepository;
+import org.springframework.stereotype.Repository;
+
+import java.util.List;
+import java.util.Optional;
+
+/**
+ * 文件类型 Repository
+ */
+@Repository
+public interface FileTypeRepository extends JpaRepository<FileType, Long> {
+
+    /**
+     * 根据类型代码查找文件类型
+     *
+     * @param typeCode 类型代码
+     * @return 文件类型对象
+     */
+    Optional<FileType> findByTypeCode(String typeCode);
+
+    /**
+     * 根据类型名称查找文件类型
+     *
+     * @param typeName 类型名称
+     * @return 文件类型对象
+     */
+    Optional<FileType> findByTypeName(String typeName);
+
+    /**
+     * 查找所有启用的文件类型，按优先级降序排列
+     *
+     * @param enabled 是否启用
+     * @return 文件类型列表
+     */
+    List<FileType> findByEnabledOrderByPriorityDesc(Boolean enabled);
+
+    /**
+     * 查找所有启用的文件类型
+     *
+     * @return 文件类型列表
+     */
+    List<FileType> findByEnabledTrue();
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/ModelRequestLogRepository.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/ModelRequestLogRepository.java
new file mode 100644
index 0000000..34e656d
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/ModelRequestLogRepository.java
@@ -0,0 +1,11 @@
+package com.zhpb.gdyd_zhpb_zgf.repository.mysql;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.ModelRequestLog;
+import org.springframework.data.jpa.repository.JpaRepository;
+import org.springframework.stereotype.Repository;
+
+@Repository
+public interface ModelRequestLogRepository extends JpaRepository<ModelRequestLog, Long> {
+}
+
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/OcrResultRepository.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/OcrResultRepository.java
new file mode 100644
index 0000000..5bae8e3
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/OcrResultRepository.java
@@ -0,0 +1,82 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.repository.mysql
+ * @ClassName: OcrResultRepository
+ * @Description: OCR结果Repository接口
+ * @Author: 张志锋
+ * @Date: 2025-12-09
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.repository.mysql;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.OcrResult;
+import org.springframework.data.jpa.repository.JpaRepository;
+import org.springframework.data.jpa.repository.Query;
+import org.springframework.data.repository.query.Param;
+import org.springframework.stereotype.Repository;
+
+import java.util.List;
+import java.util.Optional;
+
+/**
+ * OCR结果 Repository
+ */
+@Repository
+public interface OcrResultRepository extends JpaRepository<OcrResult, Long> {
+
+    /**
+     * 根据文件ID查找OCR结果
+     *
+     * @param fileId 文件ID
+     * @return OCR结果
+     */
+    Optional<OcrResult> findByFileId(Long fileId);
+
+    /**
+     * 根据项目ID查找所有OCR结果
+     *
+     * @param projectId 项目ID
+     * @return OCR结果列表
+     */
+    List<OcrResult> findByProjectId(String projectId);
+
+    /**
+     * 根据项目ID和供应商ID查找OCR结果
+     *
+     * @param projectId 项目ID
+     * @param supplierId 供应商ID
+     * @return OCR结果列表
+     */
+    List<OcrResult> findByProjectIdAndSupplierId(String projectId, String supplierId);
+
+    /**
+     * 根据状态查找OCR结果
+     *
+     * @param status 状态
+     * @return OCR结果列表
+     */
+    List<OcrResult> findByStatus(String status);
+
+    /**
+     * 查找失败的OCR任务（用于重试）
+     *
+     * @return OCR结果列表
+     */
+    @Query("SELECT o FROM OcrResult o WHERE o.status = 'failed' AND o.errorMessage IS NOT NULL")
+    List<OcrResult> findFailedOcrTasks();
+
+    /**
+     * 统计项目OCR完成情况
+     *
+     * @param projectId 项目ID
+     * @return 统计结果 [总文件数, 成功数, 失败数, 处理中数]
+     */
+    @Query("SELECT COUNT(f), " +
+           "SUM(CASE WHEN o.status = 'success' THEN 1 ELSE 0 END), " +
+           "SUM(CASE WHEN o.status = 'failed' THEN 1 ELSE 0 END), " +
+           "SUM(CASE WHEN o.status = 'processing' THEN 1 ELSE 0 END) " +
+           "FROM File f LEFT JOIN OcrResult o ON f.id = o.fileId " +
+           "WHERE f.projectId = :projectId AND f.fileExtension = 'pdf'")
+    List<Object[]> getProjectOcrStatistics(@Param("projectId") String projectId);
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/ProjectRepository.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/ProjectRepository.java
new file mode 100644
index 0000000..105619b
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/ProjectRepository.java
@@ -0,0 +1,31 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.repository.mysql
+ * @ClassName: ProjectRepository
+ * @Description: 项目Repository接口，提供项目数据访问方法
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.repository.mysql;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.Project;
+import org.springframework.data.jpa.repository.JpaRepository;
+import org.springframework.stereotype.Repository;
+
+/**
+ * 项目 Repository
+ */
+@Repository
+public interface ProjectRepository extends JpaRepository<Project, Long> {
+
+    /**
+     * 根据项目ID查找项目
+     *
+     * @param projectId 项目ID
+     * @return 项目对象
+     */
+    Project findByProjectId(String projectId);
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/PromptTemplateRepository.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/PromptTemplateRepository.java
new file mode 100644
index 0000000..6014062
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/PromptTemplateRepository.java
@@ -0,0 +1,68 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.repository.mysql
+ * @ClassName: PromptTemplateRepository
+ * @Description: 提示词模板Repository接口
+ * @Author: 张志锋
+ * @Date: 2025-11-03
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.repository.mysql;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.PromptTemplate;
+import org.springframework.data.jpa.repository.JpaRepository;
+import org.springframework.data.jpa.repository.Query;
+import org.springframework.data.repository.query.Param;
+import org.springframework.stereotype.Repository;
+
+import java.util.List;
+
+/**
+ * 提示词模板 Repository
+ */
+@Repository
+public interface PromptTemplateRepository extends JpaRepository<PromptTemplate, Long> {
+
+    /**
+     * 根据模板ID查找模板
+     *
+     * @param templateId 模板ID
+     * @return 提示词模板
+     */
+    PromptTemplate findByTemplateId(String templateId);
+
+    /**
+     * 根据模板类型查找启用的模板
+     *
+     * @param templateType 模板类型
+     * @return 模板列表
+     */
+    @Query("SELECT t FROM PromptTemplate t WHERE t.templateType = :templateType AND t.isActive = true ORDER BY t.sortOrder ASC, t.updateTime DESC")
+    List<PromptTemplate> findActiveByTemplateType(@Param("templateType") String templateType);
+
+    /**
+     * 根据模板类型查找所有模板（包括未启用的）
+     *
+     * @param templateType 模板类型
+     * @return 模板列表
+     */
+    @Query("SELECT t FROM PromptTemplate t WHERE t.templateType = :templateType ORDER BY t.sortOrder ASC, t.updateTime DESC")
+    List<PromptTemplate> findAllByTemplateType(@Param("templateType") String templateType);
+
+    /**
+     * 根据模板名称查找模板
+     *
+     * @param templateName 模板名称
+     * @return 提示词模板
+     */
+    PromptTemplate findByTemplateName(String templateName);
+
+    /**
+     * 根据模板类型统计模板数量
+     *
+     * @param templateType 模板类型
+     * @return 模板数量
+     */
+    long countByTemplateType(String templateType);
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/QaMessageRepository.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/QaMessageRepository.java
new file mode 100644
index 0000000..27e91df
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/QaMessageRepository.java
@@ -0,0 +1,61 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.repository.mysql
+ * @ClassName: QaMessageRepository
+ * @Description: QA消息Repository
+ * @Author: 张志锋
+ * @Date: 2025-11-03
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.repository.mysql;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.QaMessage;
+import org.springframework.data.jpa.repository.JpaRepository;
+import org.springframework.data.jpa.repository.Query;
+import org.springframework.data.repository.query.Param;
+import org.springframework.stereotype.Repository;
+
+import java.util.List;
+
+/**
+ * QA消息Repository
+ */
+@Repository
+public interface QaMessageRepository extends JpaRepository<QaMessage, Long> {
+
+    /**
+     * 根据消息ID查找消息
+     */
+    java.util.Optional<QaMessage> findByMessageId(String messageId);
+
+    /**
+     * 根据会话ID查找消息列表，按时间戳排序
+     */
+    List<QaMessage> findBySessionIdOrderByTimestampAsc(String sessionId);
+
+    /**
+     * 根据会话ID和发送者查找消息列表
+     */
+    List<QaMessage> findBySessionIdAndSenderOrderByTimestampAsc(String sessionId, String sender);
+
+    /**
+     * 根据发送者查找消息列表
+     */
+    List<QaMessage> findBySenderOrderByTimestampDesc(String sender);
+
+    /**
+     * 统计会话的消息数量
+     */
+    long countBySessionId(String sessionId);
+
+    /**
+     * 删除会话的所有消息
+     */
+    void deleteBySessionId(String sessionId);
+    /**
+     * 查找会话中最后一条包含文件的消息
+     */
+    @Query(value = "SELECT * FROM t_qa_message WHERE session_id = :sessionId AND file_ids IS NOT NULL AND file_ids != '[]' AND file_ids != '' ORDER BY timestamp DESC LIMIT 1", nativeQuery = true)
+    QaMessage findLastMessageWithFiles(@Param("sessionId") String sessionId);
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/QaSessionRepository.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/QaSessionRepository.java
new file mode 100644
index 0000000..bea7ef4
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/QaSessionRepository.java
@@ -0,0 +1,57 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.repository.mysql
+ * @ClassName: QaSessionRepository
+ * @Description: QA会话Repository
+ * @Author: 张志锋
+ * @Date: 2025-11-03
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.repository.mysql;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.QaSession;
+import org.springframework.data.jpa.repository.JpaRepository;
+import org.springframework.data.jpa.repository.Query;
+import org.springframework.data.repository.query.Param;
+import org.springframework.stereotype.Repository;
+
+import java.util.List;
+import java.util.Optional;
+
+/**
+ * QA会话Repository
+ */
+@Repository
+public interface QaSessionRepository extends JpaRepository<QaSession, Long> {
+
+    /**
+     * 根据会话ID查找会话
+     */
+    Optional<QaSession> findBySessionId(String sessionId);
+
+    /**
+     * 根据项目ID和用户ID查找会话列表
+     */
+    List<QaSession> findByProjectIdAndUserIdOrderByStartTimeDesc(String projectId, String userId);
+
+    /**
+     * 根据项目ID查找所有会话
+     */
+    List<QaSession> findByProjectIdOrderByStartTimeDesc(String projectId);
+
+    /**
+     * 根据用户ID查找所有会话
+     */
+    List<QaSession> findByUserIdOrderByStartTimeDesc(String userId);
+
+    /**
+     * 统计项目的会话数量
+     */
+    long countByProjectId(String projectId);
+
+    /**
+     * 统计用户的会话数量
+     */
+    long countByUserId(String userId);
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/SsoLogRepository.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/SsoLogRepository.java
new file mode 100644
index 0000000..8faead6
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/SsoLogRepository.java
@@ -0,0 +1,24 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.repository.mysql
+ * @ClassName: SsoLogRepository
+ * @Description: SSO登录日志Repository接口，提供SSO日志数据访问方法
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.repository.mysql;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.SsoLog;
+import org.springframework.data.jpa.repository.JpaRepository;
+import org.springframework.stereotype.Repository;
+
+/**
+ * SSO登录日志 Repository
+ */
+@Repository
+public interface SsoLogRepository extends JpaRepository<SsoLog, Long> {
+    
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/SupplierRepository.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/SupplierRepository.java
new file mode 100644
index 0000000..487530b
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/SupplierRepository.java
@@ -0,0 +1,47 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.repository.mysql
+ * @ClassName: SupplierRepository
+ * @Description: 供应商Repository接口
+ * @Author: 张志锋
+ * @Date: 2025-11-01
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.repository.mysql;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.Supplier;
+import org.springframework.data.jpa.repository.JpaRepository;
+import org.springframework.stereotype.Repository;
+
+/**
+ * 供应商 Repository
+ */
+@Repository
+public interface SupplierRepository extends JpaRepository<Supplier, Long> {
+
+    /**
+     * 根据供应商ID查找供应商
+     *
+     * @param supplierId 供应商ID
+     * @return 供应商对象
+     */
+    Supplier findBySupplierId(String supplierId);
+
+    /**
+     * 根据供应商代码查找供应商
+     *
+     * @param supplierCode 供应商代码
+     * @return 供应商对象
+     */
+    Supplier findBySupplierCode(String supplierCode);
+
+    /**
+     * 根据供应商名称查找供应商
+     *
+     * @param supplierName 供应商名称
+     * @return 供应商对象
+     */
+    Supplier findBySupplierName(String supplierName);
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/SysLogRepository.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/SysLogRepository.java
new file mode 100644
index 0000000..10b420a
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/SysLogRepository.java
@@ -0,0 +1,84 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.repository.mysql
+ * @ClassName: SysLogRepository
+ * @Description: 系统日志Repository接口，提供系统日志数据访问方法
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.repository.mysql;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.SysLog;
+import org.springframework.data.domain.Page;
+import org.springframework.data.domain.Pageable;
+import org.springframework.data.jpa.repository.JpaRepository;
+import org.springframework.data.jpa.repository.JpaSpecificationExecutor;
+import org.springframework.data.jpa.repository.Query;
+import org.springframework.data.repository.query.Param;
+import org.springframework.stereotype.Repository;
+
+import java.time.LocalDateTime;
+import java.util.List;
+
+/**
+ * 系统日志Repository接口
+ */
+@Repository
+public interface SysLogRepository extends JpaRepository<SysLog, Long>, JpaSpecificationExecutor<SysLog> {
+
+    /**
+     * 根据日志级别查询
+     *
+     * @param logLevel 日志级别
+     * @return 日志列表
+     */
+    List<SysLog> findByLogLevel(String logLevel);
+
+    /**
+     * 根据时间范围查询
+     *
+     * @param startTime 开始时间
+     * @param endTime 结束时间
+     * @return 日志列表
+     */
+    List<SysLog> findByCreateTimeBetween(LocalDateTime startTime, LocalDateTime endTime);
+
+    /**
+     * 根据日志级别和时间范围查询（分页）
+     *
+     * @param logLevel 日志级别
+     * @param startTime 开始时间
+     * @param endTime 结束时间
+     * @param pageable 分页参数
+     * @return 分页日志列表
+     */
+    Page<SysLog> findByLogLevelAndCreateTimeBetween(
+            String logLevel, LocalDateTime startTime, LocalDateTime endTime, Pageable pageable);
+
+    /**
+     * 根据日志来源查询
+     *
+     * @param loggerName 日志来源
+     * @param pageable 分页参数
+     * @return 分页日志列表
+     */
+    Page<SysLog> findByLoggerName(String loggerName, Pageable pageable);
+
+    /**
+     * 根据关键字搜索日志内容
+     *
+     * @param keyword 关键字
+     * @param pageable 分页参数
+     * @return 分页日志列表
+     */
+    @Query("SELECT s FROM SysLog s WHERE s.logContent LIKE %:keyword% OR s.fullMessage LIKE %:keyword%")
+    Page<SysLog> searchByKeyword(@Param("keyword") String keyword, Pageable pageable);
+
+    /**
+     * 获取最新一条日志
+     */
+    SysLog findTopByOrderByCreateTimeDesc();
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/UserLogRepository.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/UserLogRepository.java
new file mode 100644
index 0000000..048afcb
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/UserLogRepository.java
@@ -0,0 +1,77 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.repository.mysql
+ * @ClassName: UserLogRepository
+ * @Description: 用户操作日志Repository接口，提供用户操作日志数据访问方法
+ * @Author: 
+ * @Date: 2025-11-28
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.repository.mysql;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.UserLog;
+import org.springframework.data.jpa.repository.JpaRepository;
+import org.springframework.data.jpa.repository.Query;
+import org.springframework.data.repository.query.Param;
+import org.springframework.stereotype.Repository;
+
+import java.util.List;
+
+/**
+ * 用户操作日志 Repository
+ */
+@Repository
+public interface UserLogRepository extends JpaRepository<UserLog, Long> {
+
+    /**
+     * 根据用户ID查询操作日志
+     *
+     * @param userId 用户ID
+     * @return 操作日志列表
+     */
+    List<UserLog> findByUserIdOrderByCreateTimeDesc(String userId);
+
+    /**
+     * 根据项目ID查询操作日志
+     *
+     * @param projectId 项目ID
+     * @return 操作日志列表
+     */
+    List<UserLog> findByProjectIdOrderByCreateTimeDesc(String projectId);
+
+    /**
+     * 根据操作类型查询操作日志
+     *
+     * @param operationType 操作类型
+     * @return 操作日志列表
+     */
+    List<UserLog> findByOperationTypeOrderByCreateTimeDesc(String operationType);
+
+    /**
+     * 根据用户ID和操作类型查询操作日志
+     *
+     * @param userId 用户ID
+     * @param operationType 操作类型
+     * @return 操作日志列表
+     */
+    List<UserLog> findByUserIdAndOperationTypeOrderByCreateTimeDesc(String userId, String operationType);
+
+    /**
+     * 统计指定用户的操作次数
+     *
+     * @param userId 用户ID
+     * @return 操作次数
+     */
+    @Query("SELECT COUNT(*) FROM UserLog WHERE userId = :userId")
+    Long countByUserId(@Param("userId") String userId);
+
+    /**
+     * 统计指定操作类型的次数
+     *
+     * @param operationType 操作类型
+     * @return 操作次数
+     */
+    @Query("SELECT COUNT(*) FROM UserLog WHERE operationType = :operationType")
+    Long countByOperationType(@Param("operationType") String operationType);
+}
\ No newline at end of file
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/UserRepository.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/UserRepository.java
new file mode 100644
index 0000000..a1807e8
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/mysql/UserRepository.java
@@ -0,0 +1,25 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.repository.mysql
+ * @ClassName: UserRepository
+ * @Description: MySQL用户Repository接口，提供用户数据访问方法
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.repository.mysql;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.User;
+import org.springframework.data.jpa.repository.JpaRepository;
+import org.springframework.stereotype.Repository;
+
+/**
+ * MySQL 用户 Repository
+ */
+@Repository
+public interface UserRepository extends JpaRepository<User, Long> {
+    
+    User findByUserId(String userId);
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/pgsql/PdfVectorRepository.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/pgsql/PdfVectorRepository.java
new file mode 100644
index 0000000..e0a42bc
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/pgsql/PdfVectorRepository.java
@@ -0,0 +1,76 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.repository.pgsql
+ * @ClassName: PdfVectorRepository
+ * @Description: PDF向量Repository接口，提供向量数据访问方法，支持向量相似度搜索
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.repository.pgsql;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.pgsql.PdfVector;
+import org.springframework.data.jpa.repository.JpaRepository;
+import org.springframework.data.jpa.repository.Query;
+import org.springframework.data.repository.query.Param;
+import org.springframework.stereotype.Repository;
+
+import java.util.List;
+
+/**
+ * PDF向量Repository接口
+ */
+@Repository
+public interface PdfVectorRepository extends JpaRepository<PdfVector, Long> {
+
+    /**
+     * 根据文件路径查找向量
+     *
+     * @param filePath 文件路径
+     * @return PDF向量对象
+     */
+    PdfVector findByFilePath(String filePath);
+
+    /**
+     * 根据项目ID查找所有向量
+     *
+     * @param projectId 项目ID
+     * @return PDF向量列表
+     */
+    List<PdfVector> findByProjectId(String projectId);
+
+    /**
+     * 根据文件ID查找向量记录
+     *
+     * @param fileId 文件ID
+     * @return PDF向量列表
+     */
+    List<PdfVector> findByFileId(String fileId);
+
+    /**
+     * 统计文件ID对应的向量化记录数量
+     *
+     * @param fileId 文件ID
+     * @return 向量化记录数量
+     */
+    long countByFileId(String fileId);
+
+    /**
+     * 向量相似度搜索（使用pgvector的余弦相似度）
+     * 返回相似度最高的前N条记录
+     * 
+     * 注意：此方法需要传入向量字符串，格式：[0.1,0.2,0.3,...]
+     * 使用pgvector的<=>操作符进行余弦距离计算
+     * 
+     * @param vectorStr 查询向量字符串（格式：[0.1,0.2,...]）
+     * @param limit 返回数量限制
+     * @return 相似度最高的向量列表
+     */
+    @Query(value = "SELECT *, 1 - (vector <=> CAST(:vectorStr AS vector)) AS similarity " +
+            "FROM t_pdf_vector " +
+            "ORDER BY vector <=> CAST(:vectorStr AS vector) " +
+            "LIMIT :limit", nativeQuery = true)
+    List<PdfVector> findSimilarVectors(@Param("vectorStr") String vectorStr, @Param("limit") int limit);
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/pgsql/ProductRepository.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/pgsql/ProductRepository.java
new file mode 100644
index 0000000..706043e
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/repository/pgsql/ProductRepository.java
@@ -0,0 +1,25 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.repository.pgsql
+ * @ClassName: ProductRepository
+ * @Description: PostgreSQL产品Repository接口，提供产品数据访问方法
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.repository.pgsql;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.pgsql.Product;
+import org.springframework.data.jpa.repository.JpaRepository;
+import org.springframework.stereotype.Repository;
+
+/**
+ * PostgreSQL 产品 Repository
+ */
+@Repository
+public interface ProductRepository extends JpaRepository<Product, Long> {
+    
+    Product findByName(String name);
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/.DS_Store b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/.DS_Store
new file mode 100644
index 0000000..c6fe1bd
Binary files /dev/null and b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/.DS_Store differ
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/AIModelService.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/AIModelService.java
new file mode 100644
index 0000000..236a3ca
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/AIModelService.java
@@ -0,0 +1,36 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.service
+ * @ClassName: AIModelService
+ * @Description: AI模型服务接口
+ * @Author: 张志锋
+ * @Date: 2025-11-05
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.service;
+
+/**
+ * AI模型服务接口
+ */
+public interface AIModelService {
+
+    /**
+     * 调用AI模型生成回答
+     *
+     * @param systemPrompt 系统提示词
+     * @param userMessage 用户消息
+     * @return AI回复内容
+     */
+    String generateResponse(String systemPrompt, String userMessage);
+
+    /**
+     * 调用AI模型生成回答（支持流式）
+     *
+     * @param systemPrompt 系统提示词
+     * @param userMessage 用户消息
+     * @param stream 是否流式输出
+     * @return AI回复内容
+     */
+    String generateResponse(String systemPrompt, String userMessage, boolean stream);
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/ApiStatisticsService.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/ApiStatisticsService.java
new file mode 100644
index 0000000..d441c2b
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/ApiStatisticsService.java
@@ -0,0 +1,41 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.service
+ * @ClassName: ApiStatisticsService
+ * @Description: API访问统计服务接口，提供API调用次数统计查询功能
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.service;
+
+/**
+ * API访问统计服务接口
+ */
+public interface ApiStatisticsService {
+
+    /**
+     * 增加接口访问次数（已废弃，现在由日志表自动统计）
+     *
+     * @param path 接口路径
+     */
+    @Deprecated
+    void incrementApiCount(String path);
+
+    /**
+     * 获取总访问次数（从数据库日志表查询）
+     *
+     * @return 总访问次数
+     */
+    Long getTotalApiCount();
+
+    /**
+     * 获取特定接口的访问次数（从数据库日志表查询）
+     *
+     * @param path 接口路径
+     * @return 访问次数
+     */
+    Long getApiCount(String path);
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/CatalogExtractionRetryService.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/CatalogExtractionRetryService.java
new file mode 100644
index 0000000..90c2b09
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/CatalogExtractionRetryService.java
@@ -0,0 +1,111 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.service
+ * @ClassName: CatalogExtractionRetryService
+ * @Description: 目录摘取重试定时任务服务 - 扫描卡住的文件并进行重试
+ * @Author: 张志锋
+ * @Date: 2025-12-11
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.service;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.File;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.FileRepository;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.scheduling.annotation.Scheduled;
+import org.springframework.stereotype.Service;
+import org.springframework.transaction.annotation.Transactional;
+
+import java.time.LocalDateTime;
+import java.util.List;
+
+/**
+ * 目录摘取重试定时任务服务
+ * 每5分钟扫描一次状态为processing且超过30分钟未更新的文件，进行重试
+ * 重试3次后标记为失败
+ */
+@Service
+public class CatalogExtractionRetryService {
+
+    private static final Logger logger = LoggerFactory.getLogger(CatalogExtractionRetryService.class);
+
+    /** 卡住超时时间：30分钟 */
+    private static final int STUCK_TIMEOUT_MINUTES = 30;
+    
+    /** 最大重试次数 */
+    private static final int MAX_RETRY_COUNT = 3;
+
+    @Autowired
+    private FileRepository fileRepository;
+
+    @Autowired
+    private FileService fileService;
+
+    /**
+     * 定时任务：每5分钟执行一次，扫描卡住的目录摘取任务
+     */
+    @Scheduled(fixedRate = 5 * 60 * 1000, initialDelay = 60 * 1000) // 5分钟执行一次，启动后1分钟开始
+    @Transactional
+    public void scanAndRetryStuckCatalogExtraction() {
+        logger.info("[目录摘取重试] ========== 开始扫描卡住的目录摘取任务 ==========");
+        
+        try {
+            // 计算30分钟前的时间点
+            LocalDateTime cutoffTime = LocalDateTime.now().minusMinutes(STUCK_TIMEOUT_MINUTES);
+            
+            // 查询状态为processing且更新时间超过30分钟的文件
+            List<File> stuckFiles = fileRepository.findByCatalogExtractionStatusAndUpdateTimeBefore("processing", cutoffTime);
+            
+            if (stuckFiles.isEmpty()) {
+                logger.info("[目录摘取重试] 没有发现卡住的文件");
+                return;
+            }
+            
+            logger.info("[目录摘取重试] 发现 {} 个卡住的文件，开始处理", stuckFiles.size());
+            
+            int retriedCount = 0;
+            int failedCount = 0;
+            
+            for (File file : stuckFiles) {
+                try {
+                    int retryCount = file.getCatalogExtractionRetryCount() != null ? file.getCatalogExtractionRetryCount() : 0;
+                    
+                    if (retryCount >= MAX_RETRY_COUNT) {
+                        // 已达到最大重试次数，标记为失败
+                        file.setCatalogExtractionStatus("failed");
+                        fileRepository.save(file);
+                        failedCount++;
+                        logger.warn("[目录摘取重试] 文件已达最大重试次数({}次)，标记为失败: fileId={}, fileName={}", 
+                            MAX_RETRY_COUNT, file.getFileId(), file.getFileName());
+                    } else {
+                        // 重置状态为 not_started，等待下次正常处理流程
+                        file.setCatalogExtractionStatus("not_started");
+                        fileRepository.save(file);
+                        retriedCount++;
+                        logger.info("[目录摘取重试] 重置文件状态为not_started，等待重新处理: fileId={}, fileName={}, 当前重试次数={}",
+                            file.getFileId(), file.getFileName(), retryCount);
+                    }
+                } catch (Exception e) {
+                    logger.error("[目录摘取重试] 处理文件失败: fileId={}", file.getFileId(), e);
+                }
+            }
+            
+            logger.info("[目录摘取重试] ========== 扫描完成: 重置{}个, 标记失败{}个 ==========", retriedCount, failedCount);
+            
+        } catch (Exception e) {
+            logger.error("[目录摘取重试] 扫描任务执行失败", e);
+        }
+    }
+    
+    /**
+     * 手动触发重试所有卡住的文件（可通过API调用）
+     */
+    public int manualRetryStuckFiles() {
+        logger.info("[目录摘取重试] 手动触发重试所有卡住文件");
+        scanAndRetryStuckCatalogExtraction();
+        return 0;
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/ChapterContentResolver.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/ChapterContentResolver.java
new file mode 100644
index 0000000..3370aa2
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/ChapterContentResolver.java
@@ -0,0 +1,128 @@
+package com.zhpb.gdyd_zhpb_zgf.service;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.File;
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.FileDirectory;
+import com.zhpb.gdyd_zhpb_zgf.entity.pgsql.PdfVector;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.FileRepository;
+import com.zhpb.gdyd_zhpb_zgf.repository.pgsql.PdfVectorRepository;
+import com.zhpb.gdyd_zhpb_zgf.service.FilePathResolver;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.stereotype.Service;
+import org.springframework.util.CollectionUtils;
+import org.springframework.util.StringUtils;
+
+import java.nio.file.Path;
+import java.util.Comparator;
+import java.util.List;
+import java.util.Objects;
+import java.util.Optional;
+import java.util.stream.Collectors;
+
+/**
+ * 章节内容解析组件，统一从向量数据库中读取章节正文并构建用户提示。
+ * 优化：使用PdfContentCache避免重复读取PDF文件
+ */
+@Service
+public class ChapterContentResolver {
+
+    private static final Logger logger = LoggerFactory.getLogger(ChapterContentResolver.class);
+
+    private static final int MAX_SEGMENTS = 5;
+    private static final int MAX_CONTENT_LENGTH = 2000;
+
+    private final PdfVectorRepository pdfVectorRepository;
+    private final FileRepository fileRepository;
+    private final FilePathResolver filePathResolver;
+    private final PdfContentCache pdfContentCache;
+
+    public ChapterContentResolver(PdfVectorRepository pdfVectorRepository,
+                                  FileRepository fileRepository,
+                                  FilePathResolver filePathResolver,
+                                  PdfContentCache pdfContentCache) {
+        this.pdfVectorRepository = pdfVectorRepository;
+        this.fileRepository = fileRepository;
+        this.filePathResolver = filePathResolver;
+        this.pdfContentCache = pdfContentCache;
+    }
+
+    /**
+     * 根据章节目录信息构建发送给大模型的 user 消息。
+     */
+    public String buildUserMessage(FileDirectory directory) {
+        String title = directory != null && StringUtils.hasText(directory.getTitle())
+            ? directory.getTitle().trim()
+            : "未命名章节";
+        String resolvedContent = resolveChapterContent(directory);
+        if (!StringUtils.hasText(resolvedContent)) {
+            resolvedContent = "未能提取该章节正文，请参考原始文档。";
+        }
+        return "章节标题：" + title + "\n章节内容：\n" + resolvedContent;
+    }
+
+    /**
+     * 从向量库中拼接章节正文。
+     */
+    public String resolveChapterContent(FileDirectory directory) {
+        if (directory == null || !StringUtils.hasText(directory.getFileId())) {
+            return "";
+        }
+        List<PdfVector> vectors = pdfVectorRepository.findByFileId(directory.getFileId());
+        if (CollectionUtils.isEmpty(vectors)) {
+            return loadContentFromDisk(directory);
+        }
+        String merged = vectors.stream()
+            .sorted(Comparator.comparing(PdfVector::getId))
+            .map(PdfVector::getContent)
+            .filter(Objects::nonNull)
+            .map(String::trim)
+            .filter(StringUtils::hasText)
+            .limit(MAX_SEGMENTS)
+            .collect(Collectors.joining("\n\n"));
+
+        if (!StringUtils.hasText(merged)) {
+            return "";
+        }
+        if (merged.length() <= MAX_CONTENT_LENGTH) {
+            return merged;
+        }
+        return merged.substring(0, MAX_CONTENT_LENGTH) + "……（正文已截断）";
+    }
+
+    /**
+     * 当向量库没有内容时，降级从原始PDF读取文本（使用缓存）
+     */
+    private String loadContentFromDisk(FileDirectory directory) {
+        try {
+            Optional<File> fileOpt = fileRepository.findByFileId(directory.getFileId());
+            if (fileOpt.isEmpty()) {
+                return "";
+            }
+            File file = fileOpt.get();
+            Optional<Path> resolvedPath = filePathResolver.resolveExistingPath(
+                file.getFilePath(), file.getProjectId(), file.getFileName());
+            if (resolvedPath.isEmpty()) {
+                return "";
+            }
+            
+            // 使用缓存读取PDF内容（优化：避免重复打开PDF文件）
+            String fullContent = pdfContentCache.getFullContent(resolvedPath.get().toString(), 5);
+            
+            if (!StringUtils.hasText(fullContent)) {
+                return "";
+            }
+            
+            fullContent = fullContent.trim();
+            if (fullContent.length() > MAX_CONTENT_LENGTH) {
+                return fullContent.substring(0, MAX_CONTENT_LENGTH) + "……（正文已截断）";
+            }
+            return fullContent;
+            
+        } catch (Exception e) {
+            logger.warn("降级读取PDF正文失败: fileId={}, error={}", directory.getFileId(), e.getMessage());
+            return "";
+        }
+    }
+}
+
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/ComparisonAsyncService.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/ComparisonAsyncService.java
new file mode 100644
index 0000000..d245c0b
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/ComparisonAsyncService.java
@@ -0,0 +1,445 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.service
+ * @ClassName: ComparisonAsyncService
+ * @Description: 供应商对比分析异步处理服务
+ * @Author: 张志锋
+ * @Date: 2025-12-02
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.service;
+
+import com.alibaba.fastjson2.JSON;
+import com.zhpb.gdyd_zhpb_zgf.context.ModelRequestContextHolder;
+import com.zhpb.gdyd_zhpb_zgf.dto.ComparisonRequest;
+import com.zhpb.gdyd_zhpb_zgf.dto.ComparisonResponse;
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.ComparisonTask;
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.File;
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.PromptTemplate;
+import com.zhpb.gdyd_zhpb_zgf.entity.pgsql.PdfVector;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.ComparisonTaskRepository;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.FileRepository;
+import com.zhpb.gdyd_zhpb_zgf.repository.pgsql.PdfVectorRepository;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.scheduling.annotation.Async;
+import org.springframework.stereotype.Service;
+
+import java.nio.file.Path;
+import java.time.LocalDateTime;
+import java.util.ArrayList;
+import java.util.Arrays;
+import java.util.List;
+import java.util.Objects;
+import java.util.Optional;
+import java.util.concurrent.CompletableFuture;
+import java.util.stream.Collectors;
+
+/**
+ * 供应商对比分析异步处理服务
+ * 专门处理耗时的异步任务，避免阻塞主线程
+ */
+@Service
+public class ComparisonAsyncService {
+
+    private static final Logger logger = LoggerFactory.getLogger(ComparisonAsyncService.class);
+
+    @Autowired
+    private ComparisonTaskRepository comparisonTaskRepository;
+
+    @Autowired
+    private FileRepository fileRepository;
+
+    @Autowired
+    private PromptTemplateService promptTemplateService;
+
+    @Autowired
+    private AIModelService aiModelService;
+
+    @Autowired
+    private com.zhpb.gdyd_zhpb_zgf.repository.mysql.SupplierRepository supplierRepository;
+
+    @Autowired
+    private PdfVectorRepository pdfVectorRepository;
+
+    @Autowired
+    private PdfContentCache pdfContentCache;
+
+    @Autowired
+    private FilePathResolver filePathResolver;
+
+    // 专用线程池用于并行处理维度和供应商对比
+    private final java.util.concurrent.ExecutorService comparisonExecutor = 
+        java.util.concurrent.Executors.newFixedThreadPool(
+            Math.max(4, Runtime.getRuntime().availableProcessors() * 2),
+            new java.util.concurrent.ThreadFactory() {
+                private final java.util.concurrent.atomic.AtomicInteger counter = new java.util.concurrent.atomic.AtomicInteger(0);
+                @Override
+                public Thread newThread(Runnable r) {
+                    Thread t = new Thread(r, "comp-parallel-" + counter.incrementAndGet());
+                    t.setDaemon(true);
+                    return t;
+                }
+            });
+
+    /**
+     * 异步处理供应商对比分析任务
+     * @param taskId 任务ID
+     */
+    @Async("comparisonTaskExecutor")
+    public void processComparisonAsync(String taskId) {
+        long startTime = System.currentTimeMillis();
+        logger.info("[横向对比] ========== 开始供应商对比分析任务: taskId={} ==========", taskId);
+
+        try {
+            ComparisonTask task = comparisonTaskRepository.findByTaskId(taskId);
+            if (task == null) {
+                logger.error("[横向对比] 对比任务不存在: {}", taskId);
+                return;
+            }
+
+            // 更新任务状态为处理中
+            task.setStatus("processing");
+            task.setProgress(10);
+            comparisonTaskRepository.save(task);
+            logger.info("[横向对比] 步骤1: 任务状态已更新为processing");
+
+            // 解析供应商ID和维度
+            List<String> vendorIds = JSON.parseArray(task.getVendorIds(), String.class);
+            List<ComparisonRequest.ComparisonDimension> dimensions =
+                JSON.parseArray(task.getDimensions(), ComparisonRequest.ComparisonDimension.class);
+
+            logger.info("[横向对比] 步骤2: 解析任务参数 - 供应商数量={}, 维度数量={}", vendorIds.size(), dimensions.size());
+            for (int i = 0; i < vendorIds.size(); i++) {
+                logger.info("[横向对比]   供应商{}: {}", i + 1, vendorIds.get(i));
+            }
+            for (int i = 0; i < dimensions.size(); i++) {
+                logger.info("[横向对比]   维度{}: {}", i + 1, dimensions.get(i).getValue());
+            }
+
+            // 构建对比结果矩阵 - 维度+供应商双层并行处理
+            final String projectId = task.getProjectId();
+            final int totalDimensions = dimensions.size();
+            final int totalVendors = vendorIds.size();
+            final java.util.concurrent.atomic.AtomicInteger completedDimensions = new java.util.concurrent.atomic.AtomicInteger(0);
+            
+            logger.info("[横向对比] 步骤3: 开始并行处理 {} 个维度", totalDimensions);
+            
+            // 并行处理所有维度
+            List<CompletableFuture<ComparisonResponse.ResultMatrixItem>> dimensionFutures = 
+                dimensions.stream()
+                    .map(dimension -> CompletableFuture.supplyAsync(() -> {
+                        long dimStartTime = System.currentTimeMillis();
+                        ComparisonResponse.ResultMatrixItem matrixItem = new ComparisonResponse.ResultMatrixItem();
+                        matrixItem.setDimension(dimension.getValue());
+                        matrixItem.setDimensionType(dimension.getType());
+
+                        logger.info("[横向对比] >>> 开始处理维度: '{}'", dimension.getValue());
+                        
+                        // 嵌套并行：每个维度内部再并行处理所有供应商
+                        List<CompletableFuture<ComparisonResponse.VendorComparison>> vendorFutures = 
+                            vendorIds.stream()
+                                .map(vendorId -> CompletableFuture.supplyAsync(() -> 
+                                    generateVendorComparison(projectId, vendorId, dimension), comparisonExecutor))
+                                .collect(Collectors.toList());
+                        
+                        // 等待该维度下所有供应商分析完成
+                        List<ComparisonResponse.VendorComparison> comparisons = vendorFutures.stream()
+                            .map(CompletableFuture::join)
+                            .collect(Collectors.toList());
+
+                        matrixItem.setComparisons(comparisons);
+                        int completed = completedDimensions.incrementAndGet();
+                        long dimDuration = System.currentTimeMillis() - dimStartTime;
+                        logger.info("[横向对比] <<< 维度 '{}' 分析完成 ({}/{})，耗时: {}ms", 
+                            dimension.getValue(), completed, totalDimensions, dimDuration);
+                        return matrixItem;
+                    }, comparisonExecutor))
+                    .collect(Collectors.toList());
+            
+            // 更新进度为50%表示并行处理已启动
+            task.setProgress(50);
+            comparisonTaskRepository.save(task);
+            
+            // 等待所有维度分析完成并汇总结果
+            List<ComparisonResponse.ResultMatrixItem> resultMatrix = dimensionFutures.stream()
+                .map(CompletableFuture::join)
+                .collect(Collectors.toList());
+            
+            logger.info("[横向对比] 步骤4: 所有 {} 个维度分析完成，开始保存结果", totalDimensions);
+
+            // 保存结果并更新任务状态
+            task.setResult(JSON.toJSONString(resultMatrix));
+            task.setStatus("completed");
+            task.setProgress(100);
+            task.setCompletedTime(LocalDateTime.now());
+            comparisonTaskRepository.save(task);
+
+            long totalDuration = System.currentTimeMillis() - startTime;
+            logger.info("[横向对比] ========== 任务完成: taskId={}, 维度数={}, 供应商数={}, 总耗时={}ms ({}秒) ==========", 
+                taskId, totalDimensions, totalVendors, totalDuration, String.format("%.1f", totalDuration / 1000.0));
+
+        } catch (Exception e) {
+            logger.error("供应商对比分析失败: taskId={}", taskId, e);
+
+            // 更新任务状态为失败
+            try {
+                ComparisonTask task = comparisonTaskRepository.findByTaskId(taskId);
+                if (task != null) {
+                    task.setStatus("failed");
+                    task.setErrorMessage(e.getMessage());
+                    comparisonTaskRepository.save(task);
+                }
+            } catch (Exception ex) {
+                logger.error("更新任务失败状态时出错", ex);
+            }
+        }
+    }
+
+    /**
+     * 生成供应商对比分析
+     */
+    private ComparisonResponse.VendorComparison generateVendorComparison(String projectId, String vendorId,
+                                                                       ComparisonRequest.ComparisonDimension dimension) {
+        long vendorStartTime = System.currentTimeMillis();
+        String vendorName = extractSupplierNameFromVendorId(vendorId);
+        logger.info("[横向对比]   -> 开始分析供应商: {} ({}), 维度: {}", vendorName, vendorId, dimension.getValue());
+        
+        ComparisonResponse.VendorComparison comparison = new ComparisonResponse.VendorComparison();
+        comparison.setVendorId(vendorId);
+        comparison.setVendorName(vendorName);
+
+        try {
+            // 获取供应商的文件信息
+            List<File> vendorFiles = fileRepository.findByProjectIdAndSupplierId(projectId, vendorId);
+            logger.info("[横向对比]      获取到 {} 个文件", vendorFiles.size());
+
+            // 获取对比提示词模板
+            PromptTemplate template = promptTemplateService.getActiveTemplate("comparison");
+            String promptTemplate = template != null ? template.getPromptContent() :
+                "请对供应商在{dimension}维度的表现进行分析。\n\n供应商文件：{files}\n\n请直接返回分析结果。";
+
+            // 构建供应商信息（包含实际文件内容）
+            StringBuilder supplierInfo = new StringBuilder();
+            supplierInfo.append("供应商名称: ").append(vendorName).append("\n");
+            supplierInfo.append("供应商ID: ").append(vendorId).append("\n");
+            supplierInfo.append("文件数量: ").append(vendorFiles.size()).append("\n\n");
+
+            // 每个文件内容的最大长度限制（避免超出token限制）
+            final int MAX_CONTENT_LENGTH_PER_FILE = 5000;
+            // 所有文件内容的总长度限制
+            final int MAX_TOTAL_CONTENT_LENGTH = 30000;
+            int totalContentLength = 0;
+
+            for (File file : vendorFiles) {
+                supplierInfo.append("## 文件: ").append(file.getFileName()).append("\n");
+                
+                // 获取文件实际内容
+                String content = getFileContent(file);
+                if (content != null && !content.trim().isEmpty()) {
+                    // 截取内容以控制长度
+                    if (content.length() > MAX_CONTENT_LENGTH_PER_FILE) {
+                        content = content.substring(0, MAX_CONTENT_LENGTH_PER_FILE) + "...(内容已截断)";
+                    }
+                    // 检查总长度限制
+                    if (totalContentLength + content.length() > MAX_TOTAL_CONTENT_LENGTH) {
+                        supplierInfo.append("(已达到内容长度限制，跳过后续文件内容)\n\n");
+                        break;
+                    }
+                    supplierInfo.append("内容摘要:\n").append(content).append("\n\n");
+                    totalContentLength += content.length();
+                    logger.debug("[横向对比]      文件 {} 内容长度: {} 字符", file.getFileName(), content.length());
+                } else {
+                    supplierInfo.append("(无法提取内容)\n\n");
+                    logger.warn("[横向对比]      文件 {} 无法提取内容", file.getFileName());
+                }
+            }
+            
+            logger.info("[横向对比]      供应商 {} 文件内容总长度: {} 字符", vendorName, totalContentLength);
+
+            // 替换模板变量
+            String prompt = promptTemplate
+                .replace("{comparison_dimension}", dimension.getValue())
+                .replace("{supplier_info}", supplierInfo.toString());
+            
+            // 强制要求纯文本格式
+            prompt += "\n\n请注意：请仅返回纯文本格式，不要使用Markdown、HTML或其他标记语言（如 **加粗**、# 标题等）。";
+
+            // 调用AI生成分析结果
+            String analysis = callDeepSeekForComparison(prompt);
+
+            comparison.setSummary(analysis);
+            comparison.setSimilarityScores(Arrays.asList(0.8, 0.7, 0.9)); // 模拟相似度分数
+
+            // 设置文件信息
+            List<ComparisonResponse.FileInfo> fileInfos = vendorFiles.stream()
+                .map(this::convertToFileInfo)
+                .collect(Collectors.toList());
+            comparison.setFileInfo(fileInfos);
+
+            long vendorDuration = System.currentTimeMillis() - vendorStartTime;
+            logger.info("[横向对比]   <- 供应商 {} 分析完成，耗时: {}ms", vendorName, vendorDuration);
+
+        } catch (Exception e) {
+            logger.error("[横向对比] 生成供应商对比分析失败: vendorId={}, error={}", vendorId, e.getMessage(), e);
+            comparison.setSummary("分析失败: " + e.getMessage());
+            comparison.setSimilarityScores(new ArrayList<>());
+            comparison.setFileInfo(new ArrayList<>());
+        }
+
+        return comparison;
+    }
+
+    /**
+     * 转换文件信息
+     */
+    private ComparisonResponse.FileInfo convertToFileInfo(File file) {
+        ComparisonResponse.FileInfo fileInfo = new ComparisonResponse.FileInfo();
+        fileInfo.setFileId(file.getFileId());
+        fileInfo.setFileName(file.getFileName());
+        fileInfo.setFileType(getFileTypeName(file.getFileTypeId()));
+        fileInfo.setFileSize(file.getFileSize());
+        return fileInfo;
+    }
+
+    /**
+     * 获取文件类型名称
+     */
+    private String getFileTypeName(Long fileTypeId) {
+        if (fileTypeId != null) {
+            switch (fileTypeId.intValue()) {
+                case 5: return "技术文件";
+                case 6: return "商务文件";
+                case 7: return "报价文件";
+                default: return "其他文件";
+            }
+        }
+        return "未知类型";
+    }
+
+    /**
+     * 从供应商ID推断供应商名称
+     */
+    private String extractSupplierNameFromVendorId(String vendorId) {
+        try {
+            com.zhpb.gdyd_zhpb_zgf.entity.mysql.Supplier supplier = supplierRepository.findBySupplierId(vendorId);
+            if (supplier != null && supplier.getSupplierName() != null) {
+                return supplier.getSupplierName();
+            }
+        } catch (Exception e) {
+            logger.warn("查询供应商名称失败: vendorId={}", vendorId);
+        }
+        return vendorId;
+    }
+
+    /**
+     * 调用DeepSeek进行供应商对比分析（带详细日志）
+     */
+    private String callDeepSeekForComparison(String prompt) {
+        long apiStartTime = System.currentTimeMillis();
+        String threadName = Thread.currentThread().getName();
+        
+        // 记录请求开始时间和参数摘要
+        logger.info("[AI调用] ========== 开始 [{}] ==========", threadName);
+        logger.info("[AI调用] 请求时间: {}", java.time.LocalDateTime.now());
+        logger.info("[AI调用] 提示词长度: {} 字符", prompt.length());
+        // 打印提示词前200字符作为摘要
+        String promptSummary = prompt.length() > 200 ? prompt.substring(0, 200) + "..." : prompt;
+        logger.debug("[AI调用] 提示词摘要: {}", promptSummary);
+        
+        try {
+            ModelRequestContextHolder.setRequestSource("COMPARISON_ANALYSIS");
+            String result;
+            try {
+                result = aiModelService.generateResponse("", prompt);
+            } finally {
+                ModelRequestContextHolder.clear();
+            }
+
+            long apiDuration = System.currentTimeMillis() - apiStartTime;
+            
+            if (result != null && !result.trim().isEmpty()) {
+                logger.info("[AI调用] 返回成功 [{}] - 返回长度: {} 字符, 耗时: {}ms ({:.1f}秒)", 
+                    threadName, result.length(), apiDuration, apiDuration / 1000.0);
+                // 打印返回值前100字符作为摘要
+                String resultSummary = result.length() > 100 ? result.substring(0, 100) + "..." : result;
+                logger.debug("[AI调用] 返回摘要: {}", resultSummary);
+                logger.info("[AI调用] ========== 完成 [{}] 耗时: {}ms ==========", threadName, apiDuration);
+                return result.trim();
+            }
+
+            logger.warn("[AI调用] 返回为空 [{}] - 耗时: {}ms", threadName, apiDuration);
+            return "AI分析结果暂不可用";
+
+        } catch (Exception e) {
+            long apiDuration = System.currentTimeMillis() - apiStartTime;
+            logger.error("[AI调用] 调用失败 [{}] - 错误: {}, 耗时: {}ms", threadName, e.getMessage(), apiDuration, e);
+            return "AI分析失败: " + e.getMessage();
+        }
+    }
+
+    /**
+     * 获取文件实际内容
+     * 优先从向量库获取，如无则从原文件读取
+     *
+     * @param file 文件实体
+     * @return 文件内容，如果无法获取则返回空字符串
+     */
+    private String getFileContent(File file) {
+        if (file == null || file.getFileId() == null) {
+            return "";
+        }
+
+        try {
+            // 1. 首先尝试从向量库获取内容
+            List<PdfVector> vectors = pdfVectorRepository.findByFileId(file.getFileId());
+            if (vectors != null && !vectors.isEmpty()) {
+                // 合并所有向量块的内容
+                String content = vectors.stream()
+                    .map(PdfVector::getContent)
+                    .filter(Objects::nonNull)
+                    .map(String::trim)
+                    .filter(s -> !s.isEmpty())
+                    .collect(Collectors.joining("\n\n"));
+                
+                if (!content.isEmpty()) {
+                    logger.debug("[横向对比] 从向量库获取文件内容: fileId={}, 内容长度={}", 
+                        file.getFileId(), content.length());
+                    return content;
+                }
+            }
+
+            // 2. 向量库没有内容，降级从原文件读取
+            logger.debug("[横向对比] 向量库无内容，尝试从原文件读取: fileId={}", file.getFileId());
+            
+            // 解析文件路径
+            Optional<Path> resolvedPath = filePathResolver.resolveExistingPath(
+                file.getFilePath(), file.getProjectId(), file.getFileName());
+            
+            if (resolvedPath.isEmpty()) {
+                logger.warn("[横向对比] 无法解析文件路径: fileId={}, filePath={}", 
+                    file.getFileId(), file.getFilePath());
+                return "";
+            }
+
+            // 使用 PdfContentCache 读取内容（带缓存，最多读取50页）
+            String pdfPath = resolvedPath.get().toString();
+            String content = pdfContentCache.getFullContent(pdfPath, 50);
+            
+            if (content != null && !content.isEmpty()) {
+                logger.debug("[横向对比] 从原文件读取内容: fileId={}, 内容长度={}", 
+                    file.getFileId(), content.length());
+                return content;
+            }
+
+            return "";
+
+        } catch (Exception e) {
+            logger.error("[横向对比] 获取文件内容失败: fileId={}, error={}", 
+                file.getFileId(), e.getMessage(), e);
+            return "";
+        }
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/ComparisonService.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/ComparisonService.java
new file mode 100644
index 0000000..f8159f7
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/ComparisonService.java
@@ -0,0 +1,53 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.service
+ * @ClassName: ComparisonService
+ * @Description: 供应商对比分析服务接口
+ * @Author: 张志锋
+ * @Date: 2025-11-04
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.service;
+
+import com.zhpb.gdyd_zhpb_zgf.dto.ComparisonRequest;
+import com.zhpb.gdyd_zhpb_zgf.dto.ComparisonResponse;
+
+/**
+ * 供应商对比分析服务接口
+ */
+public interface ComparisonService {
+
+    /**
+     * 获取指定项目的供应商数据
+     *
+     * @param projectId 项目ID
+     * @return 供应商数据
+     */
+    ComparisonResponse.VendorsData getProjectVendors(String projectId);
+
+    /**
+     * 启动供应商对比分析任务
+     *
+     * @param request 对比分析请求
+     * @return 任务启动结果
+     */
+    ComparisonResponse.StartTaskData startComparison(ComparisonRequest request);
+
+    /**
+     * 获取用户在指定项目的对比任务历史
+     *
+     * @param projectId 项目ID
+     * @param userId 用户ID
+     * @return 任务历史列表
+     */
+    ComparisonResponse.TaskHistoryData getComparisonHistory(String projectId, String userId);
+
+    /**
+     * 获取对比任务结果
+     *
+     * @param taskId 任务ID
+     * @return 对比结果
+     */
+    ComparisonResponse.ComparisonResultData getComparisonResult(String taskId);
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/FilePathResolver.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/FilePathResolver.java
new file mode 100644
index 0000000..34014e9
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/FilePathResolver.java
@@ -0,0 +1,182 @@
+package com.zhpb.gdyd_zhpb_zgf.service;
+
+import jakarta.annotation.PostConstruct;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.beans.factory.annotation.Value;
+import org.springframework.stereotype.Component;
+
+import java.io.IOException;
+import java.nio.file.*;
+import java.util.Locale;
+import java.util.Optional;
+import java.util.stream.Stream;
+
+/**
+ * 统一管理文件存储路径，负责在不同运行环境下解析/归一化文件路径。
+ */
+@Component
+public class FilePathResolver {
+
+    private static final Logger logger = LoggerFactory.getLogger(FilePathResolver.class);
+
+    @Value("${file.upload-path:./uploads}")
+    private String uploadPathConfig;
+
+    @Value("${file.extract-path:./extracts}")
+    private String extractPathConfig;
+
+    private Path absoluteUploadPath;
+    private Path absoluteExtractPath;
+
+    @PostConstruct
+    public void init() {
+        this.absoluteUploadPath = ensureAbsolutePath(uploadPathConfig);
+        this.absoluteExtractPath = ensureAbsolutePath(extractPathConfig);
+        logger.info("文件上传根目录: {}", absoluteUploadPath);
+        logger.info("文件解压根目录: {}", absoluteExtractPath);
+    }
+
+    public Path getAbsoluteUploadPath() {
+        return absoluteUploadPath;
+    }
+
+    public Path getAbsoluteExtractPath() {
+        return absoluteExtractPath;
+    }
+
+    /**
+     * 将绝对路径归一化为可存储的字符串。
+     * 如果位于解压目录下，则返回相对路径，便于跨平台迁移。
+     */
+    public String normalizeForStorage(Path absolutePath) {
+        if (absolutePath == null) {
+            return null;
+        }
+        Path normalized = absolutePath.toAbsolutePath().normalize();
+        if (isUnder(normalized, absoluteExtractPath)) {
+            Path relative = absoluteExtractPath.relativize(normalized);
+            return toForwardSlash(relative.toString());
+        }
+        return toForwardSlash(normalized.toString());
+    }
+
+    /**
+     * 将数据库中存储的路径转换为当前环境下的绝对路径（不保证文件存在）。
+     */
+    public Path resolveStoredPath(String storedPath) {
+        if (storedPath == null || storedPath.trim().isEmpty()) {
+            return null;
+        }
+        String trimmed = storedPath.trim();
+        Path candidate = Paths.get(trimmed);
+        if (candidate.isAbsolute()) {
+            return candidate.toAbsolutePath().normalize();
+        }
+        return absoluteExtractPath.resolve(trimmed).normalize();
+    }
+
+    /**
+     * 综合多种策略解析真实存在的文件路径。
+     * 如果 storedPath 对应的文件不存在，则尝试：
+     * 1. 用当前解压根目录替换旧的绝对路径前缀
+     * 2. 在解压目录下按项目ID查找同名文件
+     */
+    public Optional<Path> resolveExistingPath(String storedPath, String projectId, String fileName) {
+        Path direct = resolveStoredPath(storedPath);
+        if (direct != null && Files.exists(direct)) {
+            return Optional.of(direct);
+        }
+
+        Optional<Path> suffixCandidate = tryResolveByExtractSuffix(storedPath);
+        if (suffixCandidate.isPresent()) {
+            Path candidate = suffixCandidate.get();
+            if (Files.exists(candidate)) {
+                return Optional.of(candidate);
+            }
+        }
+
+        return searchInProject(projectId, fileName);
+    }
+
+    private Optional<Path> searchInProject(String projectId, String fileName) {
+        if (projectId == null || projectId.isBlank() || fileName == null || fileName.isBlank()) {
+            return Optional.empty();
+        }
+        try (DirectoryStream<Path> projectDirs = Files.newDirectoryStream(absoluteExtractPath, projectId + "*")) {
+            for (Path projectDir : projectDirs) {
+                if (!Files.isDirectory(projectDir)) {
+                    continue;
+                }
+                Optional<Path> match = findFileRecursively(projectDir, fileName);
+                if (match.isPresent()) {
+                    return match;
+                }
+            }
+        } catch (IOException e) {
+            logger.warn("搜索项目目录失败: {}", e.getMessage());
+        }
+        return Optional.empty();
+    }
+
+    private Optional<Path> findFileRecursively(Path dir, String fileName) throws IOException {
+        try (Stream<Path> stream = Files.walk(dir, 12)) {
+            return stream.filter(Files::isRegularFile)
+                    .filter(path -> path.getFileName().toString().equalsIgnoreCase(fileName))
+                    .findFirst();
+        }
+    }
+
+    private Optional<Path> tryResolveByExtractSuffix(String storedPath) {
+        if (storedPath == null) {
+            return Optional.empty();
+        }
+        String lower = storedPath.toLowerCase(Locale.ROOT);
+        int idx = lower.indexOf("extracts");
+        if (idx == -1) {
+            return Optional.empty();
+        }
+        int start = idx + "extracts".length();
+        while (start < storedPath.length() &&
+                (storedPath.charAt(start) == '\\' || storedPath.charAt(start) == '/')) {
+            start++;
+        }
+        if (start >= storedPath.length()) {
+            return Optional.empty();
+        }
+        String relative = storedPath.substring(start);
+        Path candidate = absoluteExtractPath.resolve(relative).normalize();
+        return Optional.of(candidate);
+    }
+
+    private Path ensureAbsolutePath(String configPath) {
+        Path path = Paths.get(configPath);
+        if (!path.isAbsolute()) {
+            String baseDir = System.getProperty("user.dir");
+            if (baseDir == null || baseDir.isEmpty()) {
+                baseDir = System.getProperty("user.home");
+            }
+            path = Paths.get(baseDir, configPath);
+        }
+        Path normalized = path.toAbsolutePath().normalize();
+        try {
+            Files.createDirectories(normalized);
+        } catch (IOException e) {
+            logger.warn("创建目录失败: {}", normalized, e);
+        }
+        return normalized;
+    }
+
+    private boolean isUnder(Path child, Path root) {
+        try {
+            return child.startsWith(root);
+        } catch (Exception e) {
+            return false;
+        }
+    }
+
+    private String toForwardSlash(String input) {
+        return input.replace("\\", "/");
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/FilePreviewService.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/FilePreviewService.java
new file mode 100644
index 0000000..d30d867
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/FilePreviewService.java
@@ -0,0 +1,45 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.service
+ * @ClassName: FilePreviewService
+ * @Description: 文件预览服务接口
+ * @Author: 张志锋
+ * @Date: 2025-11-01
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.service;
+
+import java.io.InputStream;
+
+/**
+ * 文件预览服务接口
+ */
+public interface FilePreviewService {
+    
+    /**
+     * 获取PDF文件预览流
+     * 
+     * @param fileId 文件ID
+     * @return PDF文件输入流
+     * @throws Exception 预览失败
+     */
+    InputStream previewPdf(String fileId) throws Exception;
+    
+    /**
+     * 获取Word文档预览流（转换为PDF后返回）
+     * 
+     * @param fileId 文件ID
+     * @return PDF文件输入流（Word转换后的PDF）
+     * @throws Exception 预览失败
+     */
+    InputStream previewWord(String fileId) throws Exception;
+    
+    /**
+     * 生成文件预览（上传完成后调用，异步生成）
+     * 
+     * @param fileId 文件ID
+     */
+    void generatePreview(String fileId);
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/FileService.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/FileService.java
new file mode 100644
index 0000000..0dd2a50
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/FileService.java
@@ -0,0 +1,112 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.service
+ * @ClassName: FileService
+ * @Description: 文件服务接口，定义文件上传、ZIP解压、文件路径管理等功能
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.service;
+
+import org.springframework.web.multipart.MultipartFile;
+
+import java.io.IOException;
+import java.nio.file.Path;
+import java.util.List;
+
+/**
+ * 文件服务接口
+ */
+public interface FileService {
+
+    /**
+     * 上传ZIP文件并解压
+     *
+     * @param projectId 项目ID
+     * @param zipFile ZIP文件
+     * @return 上传结果，包含ZIP文件路径和解压后的文件路径列表
+     * @throws IOException IO异常
+     */
+    UploadResult uploadAndExtractZip(String projectId, MultipartFile zipFile) throws IOException;
+    
+    /**
+     * 上传结果
+     */
+    class UploadResult {
+        private String zipFilePath;
+        private String extractPath;
+        private List<String> extractedFilePaths;
+        
+        public UploadResult(String zipFilePath, String extractPath, List<String> extractedFilePaths) {
+            this.zipFilePath = zipFilePath;
+            this.extractPath = extractPath;
+            this.extractedFilePaths = extractedFilePaths;
+        }
+        
+        public String getZipFilePath() { return zipFilePath; }
+        public String getExtractPath() { return extractPath; }
+        public List<String> getExtractedFilePaths() { return extractedFilePaths; }
+    }
+
+    /**
+     * 创建或更新项目
+     *
+     * @param projectId 项目ID
+     * @param projectName 项目名称（可选）
+     * @return 项目对象
+     */
+    com.zhpb.gdyd_zhpb_zgf.entity.mysql.Project createOrUpdateProject(String projectId, String projectName);
+
+    /**
+     * 获取上传路径
+     *
+     * @param projectId 项目ID
+     * @return 上传路径
+     */
+    Path getUploadPath(String projectId);
+
+    /**
+     * 获取解压路径
+     *
+     * @param projectId 项目ID
+     * @return 解压路径
+     */
+    Path getExtractPath(String projectId);
+
+    /**
+     * 根据项目ID查询文件列表
+     *
+     * @param projectId 项目ID
+     * @return 项目文件信息
+     */
+    com.zhpb.gdyd_zhpb_zgf.dto.ProjectFilesResponse.ProjectFilesData getProjectFiles(String projectId);
+
+    /**
+     * 异步处理采购文件下载和处理
+     *
+     * @param projectId 项目ID
+     * @param packageId 包ID
+     * @param fileUrls 文件URL列表
+     * @param traceId 跟踪ID
+     */
+    void processProcurementFilesAsync(String projectId, String packageId, List<String> fileUrls, String traceId);
+
+    /**
+     * 异步处理应答文件下载和处理
+     *
+     * @param projectId 项目ID
+     * @param packageId 包ID
+     * @param supplierId 供应商ID
+     * @param fileUrls 文件URL列表
+     * @param fileType 文件类型
+     * @param jsonFileId JSON文件ID
+     * @param unitId 单位ID
+     * @param traceId 跟踪ID
+     */
+    void processResponseFilesAsync(String projectId, String packageId, String supplierId,
+                                 List<String> fileUrls, String fileType, String jsonFileId,
+                                 String unitId, String traceId);
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/FileTypeService.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/FileTypeService.java
new file mode 100644
index 0000000..f7b984e
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/FileTypeService.java
@@ -0,0 +1,46 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.service
+ * @ClassName: FileTypeService
+ * @Description: 文件类型服务接口
+ * @Author: 张志锋
+ * @Date: 2025-11-01
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.service;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.FileType;
+
+/**
+ * 文件类型服务接口
+ */
+public interface FileTypeService {
+
+    /**
+     * 识别文件类型（根据文件名、路径等自动识别）
+     *
+     * @param fileName 文件名
+     * @param filePath 文件路径
+     * @return 文件类型对象，如果无法识别则返回null
+     */
+    FileType identifyFileType(String fileName, String filePath);
+
+    /**
+     * 创建或获取文件类型
+     *
+     * @param typeName 类型名称
+     * @param description 类型描述
+     * @return 文件类型对象
+     */
+    FileType createOrGetFileType(String typeName, String description);
+
+    /**
+     * 根据类型代码获取文件类型
+     *
+     * @param typeCode 类型代码
+     * @return 文件类型对象
+     */
+    FileType getFileTypeByCode(String typeCode);
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/ModelRequestLogService.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/ModelRequestLogService.java
new file mode 100644
index 0000000..fdf64f9
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/ModelRequestLogService.java
@@ -0,0 +1,18 @@
+package com.zhpb.gdyd_zhpb_zgf.service;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.ModelRequestLog;
+
+/**
+ * 模型请求日志服务
+ */
+public interface ModelRequestLogService {
+
+    /**
+     * 持久化模型请求日志
+     *
+     * @param log 模型请求日志实体
+     */
+    void record(ModelRequestLog log);
+}
+
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/OcrService.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/OcrService.java
new file mode 100644
index 0000000..404220a
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/OcrService.java
@@ -0,0 +1,211 @@
+package com.zhpb.gdyd_zhpb_zgf.service;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.File;
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.OcrResult;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.FileRepository;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.OcrResultRepository;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.scheduling.annotation.Async;
+import org.springframework.stereotype.Service;
+import org.springframework.transaction.annotation.Transactional;
+
+import java.time.LocalDateTime;
+import java.util.List;
+import java.util.Map;
+import java.util.Optional;
+
+/**
+ * OCR服务
+ * 负责管理PDF文件的OCR识别任务
+ */
+@Service
+public class OcrService {
+
+    private static final Logger logger = LoggerFactory.getLogger(OcrService.class);
+
+    @Autowired
+    private UmiOcrClient umiOcrClient;
+
+    @Autowired
+    private OcrResultRepository ocrResultRepository;
+
+    @Autowired
+    private FileRepository fileRepository;
+
+    /**
+     * 异步处理PDF文件的OCR识别
+     */
+    @Async
+    public void processPdfOcrAsync(Long fileId, String filePath, String projectId, String supplierId) {
+        logger.info("🚀 开始异步处理PDF OCR: fileId={}, filePath={}, projectId={}, supplierId={}",
+                   fileId, filePath, projectId, supplierId);
+
+        try {
+            // 检查文件是否存在
+            java.io.File file = new java.io.File(filePath);
+            if (!file.exists()) {
+                logger.error("❌ OCR文件不存在: {}", filePath);
+                updateOcrResult(fileId, "failed", null, null, null, "文件不存在: " + filePath);
+                return;
+            }
+
+            logger.info("📁 OCR文件存在，大小: {} bytes", file.length());
+            // 检查是否已存在OCR结果
+            Optional<OcrResult> existingResult = ocrResultRepository.findByFileId(fileId);
+            if (existingResult.isPresent()) {
+                logger.info("文件已存在OCR结果，跳过处理: fileId={}", fileId);
+                return;
+            }
+
+            // 创建OCR结果记录
+            OcrResult ocrResult = new OcrResult();
+            ocrResult.setFileId(fileId);
+            ocrResult.setProjectId(projectId);
+            ocrResult.setSupplierId(supplierId);
+            ocrResult.setStatus("processing");
+            ocrResult.setStartTime(LocalDateTime.now());
+            ocrResult = ocrResultRepository.save(ocrResult);
+
+            // 执行OCR识别 - 使用简化的配置避免页数范围问题
+            Map<String, Object> options = Map.of(
+                "ocr.language", "models/config_chinese.txt"
+                // 不设置页数范围，让OCR服务自动处理
+            );
+
+            List<String> downloadFileTypes = List.of("pdfLayered", "txtPlain");
+
+            Map<String, Object> ocrResultData = umiOcrClient.runDocOcrSync(
+                filePath, options, downloadFileTypes, 5);
+
+            // 更新OCR结果
+            ocrResult.setStatus("success");
+            ocrResult.setOcrTaskId((String) ocrResultData.get("task_id"));
+            ocrResult.setRecognizedText((String) ocrResultData.get("recognized_text"));
+            ocrResult.setEndTime(LocalDateTime.now());
+            ocrResult.setUpdateTime(LocalDateTime.now());
+
+            @SuppressWarnings("unchecked")
+            Map<String, String> downloadInfo = (Map<String, String>) ocrResultData.get("download_info");
+            if (downloadInfo != null) {
+                ocrResult.setDownloadUrl(downloadInfo.get("download_url"));
+                ocrResult.setDownloadFileName(downloadInfo.get("file_name"));
+            }
+
+            // 解析页数信息（从响应中提取）
+            // 这里可以根据实际需要添加更多字段
+
+            ocrResultRepository.save(ocrResult);
+
+            logger.info("✅ PDF OCR处理完成: fileId={}, taskId={}, 识别文本长度={}",
+                       fileId, ocrResult.getOcrTaskId(),
+                       ocrResult.getRecognizedText() != null ? ocrResult.getRecognizedText().length() : 0);
+
+        } catch (Exception e) {
+            logger.error("❌ PDF OCR处理失败: fileId={}, error={}", fileId, e.getMessage(), e);
+            updateOcrResult(fileId, "failed", null, null, null, e.getMessage());
+        }
+    }
+
+    /**
+     * 根据文件ID获取OCR结果
+     */
+    public Optional<OcrResult> getOcrResultByFileId(Long fileId) {
+        return ocrResultRepository.findByFileId(fileId);
+    }
+
+    /**
+     * 根据项目ID获取OCR结果列表
+     */
+    public List<OcrResult> getOcrResultsByProject(String projectId) {
+        return ocrResultRepository.findByProjectId(projectId);
+    }
+
+    /**
+     * 根据项目和供应商获取OCR结果
+     */
+    public List<OcrResult> getOcrResultsByProjectAndSupplier(String projectId, String supplierId) {
+        return ocrResultRepository.findByProjectIdAndSupplierId(projectId, supplierId);
+    }
+
+    /**
+     * 获取失败的OCR任务（用于重试）
+     */
+    public List<OcrResult> getFailedOcrTasks() {
+        return ocrResultRepository.findFailedOcrTasks();
+    }
+
+    /**
+     * 重试失败的OCR任务
+     */
+    @Async
+    public void retryFailedOcrTask(Long ocrResultId) {
+        Optional<OcrResult> ocrResultOpt = ocrResultRepository.findById(ocrResultId);
+        if (!ocrResultOpt.isPresent()) {
+            logger.warn("OCR结果不存在: id={}", ocrResultId);
+            return;
+        }
+
+        OcrResult ocrResult = ocrResultOpt.get();
+        if (!"failed".equals(ocrResult.getStatus())) {
+            logger.info("OCR任务状态不是失败状态，无需重试: id={}", ocrResultId);
+            return;
+        }
+
+        // 获取对应的文件信息
+        Optional<File> fileOpt = fileRepository.findById(ocrResult.getFileId());
+        if (!fileOpt.isPresent()) {
+            logger.warn("对应的文件不存在: fileId={}", ocrResult.getFileId());
+            return;
+        }
+
+        File file = fileOpt.get();
+
+        // 重新执行OCR
+        logger.info("开始重试OCR任务: fileId={}, ocrResultId={}", file.getId(), ocrResultId);
+        processPdfOcrAsync(file.getId(), file.getFilePath(), file.getProjectId(), file.getSupplierId());
+    }
+
+    /**
+     * 获取项目OCR统计信息
+     */
+    public Map<String, Object> getProjectOcrStatistics(String projectId) {
+        List<Object[]> stats = ocrResultRepository.getProjectOcrStatistics(projectId);
+
+        Map<String, Object> result = Map.of(
+            "totalPdfFiles", stats.get(0)[0] != null ? ((Number) stats.get(0)[0]).intValue() : 0,
+            "successfulOcr", stats.get(0)[1] != null ? ((Number) stats.get(0)[1]).intValue() : 0,
+            "failedOcr", stats.get(0)[2] != null ? ((Number) stats.get(0)[2]).intValue() : 0,
+            "processingOcr", stats.get(0)[3] != null ? ((Number) stats.get(0)[3]).intValue() : 0
+        );
+
+        return result;
+    }
+
+    /**
+     * 更新OCR结果状态的辅助方法
+     */
+    private void updateOcrResult(Long fileId, String status, String recognizedText,
+                                String downloadUrl, String downloadFileName, String errorMessage) {
+        try {
+            Optional<OcrResult> resultOpt = ocrResultRepository.findByFileId(fileId);
+            if (resultOpt.isPresent()) {
+                OcrResult ocrResult = resultOpt.get();
+                ocrResult.setStatus(status);
+                if (recognizedText != null) ocrResult.setRecognizedText(recognizedText);
+                if (downloadUrl != null) ocrResult.setDownloadUrl(downloadUrl);
+                if (downloadFileName != null) ocrResult.setDownloadFileName(downloadFileName);
+                if (errorMessage != null) ocrResult.setErrorMessage(errorMessage);
+                ocrResult.setEndTime(LocalDateTime.now());
+                ocrResult.setUpdateTime(LocalDateTime.now());
+                ocrResultRepository.save(ocrResult);
+                logger.info("✅ 更新OCR结果状态: fileId={}, status={}", fileId, status);
+            } else {
+                logger.warn("⚠️ 未找到OCR结果记录: fileId={}", fileId);
+            }
+        } catch (Exception e) {
+            logger.error("❌ 更新OCR结果状态失败: fileId={}", fileId, e);
+        }
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/PdfContentCache.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/PdfContentCache.java
new file mode 100644
index 0000000..25ade4c
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/PdfContentCache.java
@@ -0,0 +1,299 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.service
+ * @ClassName: PdfContentCache
+ * @Description: PDF内容缓存服务，避免重复读取PDF文件
+ * @Author: 张志锋
+ * @Date: 2025-12-11
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.service;
+
+import org.apache.pdfbox.Loader;
+import org.apache.pdfbox.pdmodel.PDDocument;
+import org.apache.pdfbox.text.PDFTextStripper;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.stereotype.Service;
+
+import java.io.File;
+import java.nio.file.Path;
+import java.util.Map;
+import java.util.concurrent.*;
+
+/**
+ * PDF内容缓存服务
+ * 使用LRU缓存策略，避免重复读取PDF文件
+ */
+@Service
+public class PdfContentCache {
+
+    private static final Logger logger = LoggerFactory.getLogger(PdfContentCache.class);
+
+    // LRU缓存：文件路径 -> 全文内容
+    private final Map<String, CachedContent> contentCache;
+    
+    // 按页缓存：文件路径 -> (页码 -> 页内容)
+    private final Map<String, ConcurrentMap<Integer, String>> pageCache;
+    
+    // 缓存配置
+    private static final int MAX_CACHE_ENTRIES = 20;  // 最多缓存20个PDF
+    private static final long CACHE_EXPIRE_MS = 10 * 60 * 1000;  // 10分钟过期
+    
+    // 并行读取配置
+    private static final int PARALLEL_PAGE_THRESHOLD = 50;  // 超过50页启用并行读取
+    private static final int PAGES_PER_BATCH = 20;  // 每批读取20页
+    
+    private final ExecutorService pdfReadExecutor;
+    
+    public PdfContentCache() {
+        // 使用LinkedHashMap实现LRU缓存
+        this.contentCache = new java.util.LinkedHashMap<String, CachedContent>(MAX_CACHE_ENTRIES, 0.75f, true) {
+            @Override
+            protected boolean removeEldestEntry(Map.Entry<String, CachedContent> eldest) {
+                if (size() > MAX_CACHE_ENTRIES) {
+                    logger.debug("移除最久未使用的PDF缓存: {}", eldest.getKey());
+                    return true;
+                }
+                return false;
+            }
+        };
+        this.pageCache = new ConcurrentHashMap<>();
+        
+        // PDF读取线程池
+        this.pdfReadExecutor = Executors.newFixedThreadPool(3, r -> {
+            Thread t = new Thread(r, "pdf-reader");
+            t.setDaemon(true);
+            return t;
+        });
+        
+        logger.info("PDF内容缓存服务初始化完成，最大缓存: {} 个文件，过期时间: {} 分钟", 
+            MAX_CACHE_ENTRIES, CACHE_EXPIRE_MS / 60000);
+    }
+    
+    /**
+     * 获取PDF全文内容（带缓存）
+     */
+    public String getFullContent(String filePath) {
+        return getFullContent(filePath, 100);  // 默认最多读取100页
+    }
+    
+    /**
+     * 获取PDF全文内容（带缓存）
+     */
+    public synchronized String getFullContent(String filePath, int maxPages) {
+        // 检查缓存
+        CachedContent cached = contentCache.get(filePath);
+        if (cached != null && !cached.isExpired()) {
+            logger.debug("PDF缓存命中: {}", filePath);
+            return cached.content;
+        }
+        
+        // 缓存未命中，读取PDF
+        logger.info("PDF缓存未命中，开始读取: {}", filePath);
+        long startTime = System.currentTimeMillis();
+        
+        String content = readPdfContent(filePath, maxPages);
+        
+        // 存入缓存
+        contentCache.put(filePath, new CachedContent(content, System.currentTimeMillis()));
+        
+        long elapsed = System.currentTimeMillis() - startTime;
+        logger.info("PDF读取完成: {}, 耗时: {}ms, 内容长度: {} 字符", 
+            filePath, elapsed, content.length());
+        
+        return content;
+    }
+    
+    /**
+     * 获取PDF指定页范围的内容（带缓存）
+     */
+    public String getPageContent(String filePath, int startPage, int endPage) {
+        String cacheKey = filePath;
+        pageCache.computeIfAbsent(cacheKey, k -> new ConcurrentHashMap<>());
+        
+        StringBuilder result = new StringBuilder();
+        
+        for (int page = startPage; page <= endPage; page++) {
+            ConcurrentMap<Integer, String> pages = pageCache.get(cacheKey);
+            String pageContent = pages.get(page);
+            
+            if (pageContent == null) {
+                // 页面未缓存，需要读取
+                pageContent = readSinglePage(filePath, page);
+                pages.put(page, pageContent);
+            }
+            
+            result.append(pageContent).append("\n");
+        }
+        
+        return result.toString();
+    }
+    
+    /**
+     * 预加载PDF内容（异步）
+     */
+    public CompletableFuture<String> preloadAsync(String filePath, int maxPages) {
+        return CompletableFuture.supplyAsync(() -> getFullContent(filePath, maxPages), pdfReadExecutor);
+    }
+    
+    /**
+     * 并行读取大PDF（按页范围分批）
+     */
+    public String readLargePdfParallel(String filePath) {
+        File file = new File(filePath);
+        if (!file.exists()) {
+            logger.warn("PDF文件不存在: {}", filePath);
+            return "";
+        }
+        
+        try (PDDocument document = Loader.loadPDF(file)) {
+            int totalPages = document.getNumberOfPages();
+            
+            if (totalPages <= PARALLEL_PAGE_THRESHOLD) {
+                // 小PDF直接读取
+                return readPdfContent(filePath, totalPages);
+            }
+            
+            logger.info("大PDF并行读取: {}, 总页数: {}", filePath, totalPages);
+            long startTime = System.currentTimeMillis();
+            
+            // 计算批次
+            int batchCount = (totalPages + PAGES_PER_BATCH - 1) / PAGES_PER_BATCH;
+            
+            // 并行读取各批次
+            CompletableFuture<String>[] futures = new CompletableFuture[batchCount];
+            
+            for (int i = 0; i < batchCount; i++) {
+                int batchStart = i * PAGES_PER_BATCH + 1;
+                int batchEnd = Math.min((i + 1) * PAGES_PER_BATCH, totalPages);
+                
+                final int start = batchStart;
+                final int end = batchEnd;
+                
+                futures[i] = CompletableFuture.supplyAsync(() -> {
+                    try {
+                        return readPageRange(filePath, start, end);
+                    } catch (Exception e) {
+                        logger.error("读取PDF页面范围失败: pages={}-{}", start, end, e);
+                        return "";
+                    }
+                }, pdfReadExecutor);
+            }
+            
+            // 等待所有批次完成并合并结果
+            CompletableFuture.allOf(futures).join();
+            
+            StringBuilder fullContent = new StringBuilder();
+            for (CompletableFuture<String> future : futures) {
+                fullContent.append(future.get());
+            }
+            
+            long elapsed = System.currentTimeMillis() - startTime;
+            logger.info("大PDF并行读取完成: {}, 耗时: {}ms", filePath, elapsed);
+            
+            return fullContent.toString();
+            
+        } catch (Exception e) {
+            logger.error("并行读取大PDF失败: {}", filePath, e);
+            return readPdfContent(filePath, 100);  // 降级到普通读取
+        }
+    }
+    
+    /**
+     * 清除指定文件的缓存
+     */
+    public synchronized void evict(String filePath) {
+        contentCache.remove(filePath);
+        pageCache.remove(filePath);
+        logger.debug("清除PDF缓存: {}", filePath);
+    }
+    
+    /**
+     * 清除所有缓存
+     */
+    public synchronized void clearAll() {
+        contentCache.clear();
+        pageCache.clear();
+        logger.info("清除所有PDF缓存");
+    }
+    
+    /**
+     * 获取缓存统计信息
+     */
+    public Map<String, Object> getCacheStats() {
+        Map<String, Object> stats = new ConcurrentHashMap<>();
+        stats.put("fullContentCacheSize", contentCache.size());
+        stats.put("pageCacheSize", pageCache.size());
+        stats.put("maxCacheEntries", MAX_CACHE_ENTRIES);
+        stats.put("cacheExpireMinutes", CACHE_EXPIRE_MS / 60000);
+        return stats;
+    }
+    
+    // ==================== 私有方法 ====================
+    
+    private String readPdfContent(String filePath, int maxPages) {
+        File file = new File(filePath);
+        if (!file.exists()) {
+            logger.warn("PDF文件不存在: {}", filePath);
+            return "";
+        }
+        
+        try (PDDocument document = Loader.loadPDF(file)) {
+            PDFTextStripper stripper = new PDFTextStripper();
+            stripper.setStartPage(1);
+            stripper.setEndPage(Math.min(document.getNumberOfPages(), maxPages));
+            stripper.setSortByPosition(true);
+            
+            String text = stripper.getText(document);
+            return text != null ? text.trim() : "";
+            
+        } catch (Exception e) {
+            logger.error("读取PDF内容失败: {}", filePath, e);
+            return "";
+        }
+    }
+    
+    private String readPageRange(String filePath, int startPage, int endPage) {
+        File file = new File(filePath);
+        if (!file.exists()) {
+            return "";
+        }
+        
+        try (PDDocument document = Loader.loadPDF(file)) {
+            PDFTextStripper stripper = new PDFTextStripper();
+            stripper.setStartPage(startPage);
+            stripper.setEndPage(endPage);
+            stripper.setSortByPosition(true);
+            
+            String text = stripper.getText(document);
+            return text != null ? text.trim() : "";
+            
+        } catch (Exception e) {
+            logger.error("读取PDF页面范围失败: {}, pages={}-{}", filePath, startPage, endPage, e);
+            return "";
+        }
+    }
+    
+    private String readSinglePage(String filePath, int page) {
+        return readPageRange(filePath, page, page);
+    }
+    
+    /**
+     * 缓存内容包装类
+     */
+    private static class CachedContent {
+        final String content;
+        final long timestamp;
+        
+        CachedContent(String content, long timestamp) {
+            this.content = content;
+            this.timestamp = timestamp;
+        }
+        
+        boolean isExpired() {
+            return System.currentTimeMillis() - timestamp > CACHE_EXPIRE_MS;
+        }
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/PdfReadService.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/PdfReadService.java
new file mode 100644
index 0000000..4acad11
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/PdfReadService.java
@@ -0,0 +1,28 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.service
+ * @ClassName: PdfReadService
+ * @Description: PDF文件读取服务接口，提供异步读取PDF文件内容的功能
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.service;
+
+import java.util.List;
+
+/**
+ * PDF文件读取服务接口
+ */
+public interface PdfReadService {
+
+    /**
+     * 异步读取PDF文件内容并进行向量化
+     * 
+     * @param pdfFilePaths PDF文件路径列表
+     * @param projectId 项目ID（可选）
+     */
+    void readPdfFilesAsync(List<String> pdfFilePaths, String projectId);
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/PromptTemplateService.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/PromptTemplateService.java
new file mode 100644
index 0000000..c6a71e3
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/PromptTemplateService.java
@@ -0,0 +1,81 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.service
+ * @ClassName: PromptTemplateService
+ * @Description: 提示词模板服务接口
+ * @Author: 张志锋
+ * @Date: 2025-11-03
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.service;
+
+import com.zhpb.gdyd_zhpb_zgf.dto.PromptTemplateDto;
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.PromptTemplate;
+
+import java.util.List;
+
+/**
+ * 提示词模板服务接口
+ */
+public interface PromptTemplateService {
+
+    /**
+     * 获取所有提示词模板
+     *
+     * @param templateType 模板类型，可为空
+     * @return 模板列表
+     */
+    List<PromptTemplateDto> getAllTemplates(String templateType);
+
+    /**
+     * 根据ID获取模板
+     *
+     * @param templateId 模板ID
+     * @return 模板信息
+     */
+    PromptTemplateDto getTemplateById(String templateId);
+
+    /**
+     * 创建模板
+     *
+     * @param dto 模板信息
+     * @param userId 操作用户ID
+     * @return 创建的模板
+     */
+    PromptTemplateDto createTemplate(PromptTemplateDto dto, String userId);
+
+    /**
+     * 更新模板
+     *
+     * @param templateId 模板ID
+     * @param dto 模板信息
+     * @param userId 操作用户ID
+     * @return 更新后的模板
+     */
+    PromptTemplateDto updateTemplate(String templateId, PromptTemplateDto dto, String userId);
+
+    /**
+     * 删除模板
+     *
+     * @param templateId 模板ID
+     */
+    void deleteTemplate(String templateId);
+
+    /**
+     * 启用/禁用模板
+     *
+     * @param templateId 模板ID
+     * @param isActive 是否启用
+     * @param userId 操作用户ID
+     */
+    void toggleTemplateStatus(String templateId, Boolean isActive, String userId);
+
+    /**
+     * 获取启用的模板
+     *
+     * @param templateType 模板类型
+     * @return 启用的模板
+     */
+    PromptTemplate getActiveTemplate(String templateType);
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/QaService.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/QaService.java
new file mode 100644
index 0000000..c808e48
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/QaService.java
@@ -0,0 +1,46 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.service
+ * @ClassName: QaService
+ * @Description: 智能问答服务接口
+ * @Author: 张志锋
+ * @Date: 2025-11-03
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.service;
+
+import com.zhpb.gdyd_zhpb_zgf.dto.QaHistoryResponse;
+import com.zhpb.gdyd_zhpb_zgf.dto.QaRequest;
+import com.zhpb.gdyd_zhpb_zgf.dto.QaResponse;
+
+/**
+ * 智能问答服务接口
+ */
+public interface QaService {
+
+    /**
+     * 发送智能问答消息
+     *
+     * @param request 问答请求
+     * @return 问答响应
+     */
+    QaResponse.QaData sendMessage(QaRequest request);
+
+    /**
+     * 查询历史会话
+     *
+     * @param projectId 项目ID
+     * @param userId 用户ID
+     * @return 会话列表
+     */
+    QaHistoryResponse.SessionsData getSessionHistory(String projectId, String userId);
+
+    /**
+     * 查询会话消息历史
+     *
+     * @param sessionId 会话ID
+     * @return 消息列表
+     */
+    QaHistoryResponse.MessagesData getMessageHistory(String sessionId);
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/SummaryGenerationScheduleService.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/SummaryGenerationScheduleService.java
new file mode 100644
index 0000000..0ce074b
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/SummaryGenerationScheduleService.java
@@ -0,0 +1,182 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.service
+ * @ClassName: SummaryGenerationScheduleService
+ * @Description: 摘要总结定时扫描服务 - 每3分钟扫描摘要为空的章节并调用模型生成摘要
+ * @Author: 张志锋
+ * @Date: 2025-12-13
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.service;
+
+import com.zhpb.gdyd_zhpb_zgf.context.ModelRequestContextHolder;
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.FileDirectory;
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.PromptTemplate;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.FileDirectoryRepository;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.scheduling.annotation.Scheduled;
+import org.springframework.stereotype.Service;
+
+import java.util.List;
+import java.util.concurrent.atomic.AtomicBoolean;
+
+/**
+ * 摘要总结定时扫描服务
+ * 每3分钟扫描一次摘要为空的章节记录，调用模型生成摘要
+ * 使用单线程处理，避免资源竞争
+ */
+@Service
+public class SummaryGenerationScheduleService {
+
+    private static final Logger logger = LoggerFactory.getLogger(SummaryGenerationScheduleService.class);
+
+    /** 每批处理的最大数量 */
+    private static final int BATCH_SIZE = 20;
+
+    /** 标记是否正在执行，确保只有一个线程在处理 */
+    private final AtomicBoolean isRunning = new AtomicBoolean(false);
+
+    @Autowired
+    private FileDirectoryRepository fileDirectoryRepository;
+
+    @Autowired
+    private PromptTemplateService promptTemplateService;
+
+    @Autowired
+    private AIModelService aiModelService;
+
+    @Autowired
+    private ChapterContentResolver chapterContentResolver;
+
+    /**
+     * 定时任务：每3分钟执行一次，扫描摘要为空的章节
+     */
+    @Scheduled(fixedRate = 3 * 60 * 1000, initialDelay = 2 * 60 * 1000) // 3分钟执行一次，启动后2分钟开始
+    public void scanAndGenerateMissingSummaries() {
+        // 使用原子操作确保只有一个线程在执行
+        if (!isRunning.compareAndSet(false, true)) {
+            logger.debug("[摘要定时扫描] 上一次任务尚未完成，跳过本次执行");
+            return;
+        }
+
+        try {
+            logger.info("[摘要定时扫描] ========== 开始扫描摘要为空的章节 ==========");
+            
+            // 查询摘要为空的章节记录
+            List<FileDirectory> emptyDirectories = fileDirectoryRepository.findBySummaryIsNullOrEmpty();
+            
+            if (emptyDirectories.isEmpty()) {
+                logger.info("[摘要定时扫描] 没有发现需要生成摘要的章节");
+                return;
+            }
+            
+            logger.info("[摘要定时扫描] 发现 {} 个需要生成摘要的章节，开始处理（最多处理{}个）", 
+                emptyDirectories.size(), BATCH_SIZE);
+            
+            // 获取摘要提示词模板
+            PromptTemplate summaryTemplate = promptTemplateService.getActiveTemplate("summary");
+            if (summaryTemplate == null) {
+                logger.error("[摘要定时扫描] 未找到激活的文档摘要提示词模板，请启用或创建summary类型的提示词模板");
+                return;
+            }
+            
+            String promptTemplate = summaryTemplate.getPromptContent();
+            logger.info("[摘要定时扫描] 使用提示词模板: {} (ID: {})", 
+                summaryTemplate.getTemplateName(), summaryTemplate.getTemplateId());
+            
+            int successCount = 0;
+            int failCount = 0;
+            int processedCount = 0;
+            
+            // 逐个处理，使用单线程避免资源竞争
+            for (FileDirectory directory : emptyDirectories) {
+                if (processedCount >= BATCH_SIZE) {
+                    logger.info("[摘要定时扫描] 已达到本批次处理上限({})，剩余 {} 个章节将在下次执行时处理", 
+                        BATCH_SIZE, emptyDirectories.size() - processedCount);
+                    break;
+                }
+                
+                try {
+                    processedCount++;
+                    logger.info("[摘要定时扫描] 处理章节 [{}/{}]: fileId={}, chapterId={}, title={}", 
+                        processedCount, Math.min(emptyDirectories.size(), BATCH_SIZE),
+                        directory.getFileId(), directory.getChapterId(), directory.getTitle());
+                    
+                    // 生成摘要
+                    String summary = generateSummary(directory, promptTemplate);
+                    
+                    if (summary != null && !summary.trim().isEmpty()) {
+                        // 保存摘要（不修改目录抽取状态）
+                        directory.setSummary(summary.trim());
+                        fileDirectoryRepository.save(directory);
+                        successCount++;
+                        logger.info("[摘要定时扫描] 摘要生成成功: chapterId={}, 摘要长度={}", 
+                            directory.getChapterId(), summary.length());
+                    } else {
+                        failCount++;
+                        logger.warn("[摘要定时扫描] 摘要生成返回空: chapterId={}", directory.getChapterId());
+                    }
+                    
+                } catch (Exception e) {
+                    failCount++;
+                    logger.error("[摘要定时扫描] 处理章节失败: chapterId={}, error={}", 
+                        directory.getChapterId(), e.getMessage(), e);
+                }
+            }
+            
+            logger.info("[摘要定时扫描] ========== 扫描完成: 处理{}个, 成功{}个, 失败{}个 ==========", 
+                processedCount, successCount, failCount);
+            
+        } catch (Exception e) {
+            logger.error("[摘要定时扫描] 扫描任务执行失败", e);
+        } finally {
+            isRunning.set(false);
+        }
+    }
+
+    /**
+     * 使用AI模型生成章节摘要
+     */
+    private String generateSummary(FileDirectory directory, String promptTemplate) {
+        try {
+            // 使用模板替换变量
+            String prompt = promptTemplate.replace("{title}", 
+                directory.getTitle() != null ? directory.getTitle() : "未命名章节");
+            
+            // 获取章节内容
+            String userMessage = chapterContentResolver.buildUserMessage(directory);
+            
+            // 调用模型生成摘要
+            ModelRequestContextHolder.setRequestSource("DOC_SUMMARY_SCHEDULED");
+            try {
+                return aiModelService.generateResponse(prompt, userMessage);
+            } finally {
+                ModelRequestContextHolder.clear();
+            }
+            
+        } catch (Exception e) {
+            logger.error("[摘要定时扫描] 生成摘要失败: chapterId={}", directory.getChapterId(), e);
+            return null;
+        }
+    }
+
+    /**
+     * 手动触发摘要生成（可通过API调用）
+     * @return 处理的章节数量
+     */
+    public int manualTriggerSummaryGeneration() {
+        logger.info("[摘要定时扫描] 手动触发摘要生成");
+        scanAndGenerateMissingSummaries();
+        return 0;
+    }
+
+    /**
+     * 获取当前待处理的摘要数量
+     */
+    public long getPendingSummaryCount() {
+        return fileDirectoryRepository.findBySummaryIsNullOrEmpty().size();
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/SysLogService.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/SysLogService.java
new file mode 100644
index 0000000..8fd95f8
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/SysLogService.java
@@ -0,0 +1,29 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.service
+ * @ClassName: SysLogService
+ * @Description: 系统日志服务接口，提供日志记录功能
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.service;
+
+/**
+ * 系统日志服务接口
+ */
+public interface SysLogService {
+
+    /**
+     * 记录系统日志
+     *
+     * @param logLevel 日志级别
+     * @param loggerName 日志来源
+     * @param logContent 日志内容
+     * @param fullMessage 完整消息
+     * @param exceptionStack 异常堆栈
+     */
+    void log(String logLevel, String loggerName, String logContent, String fullMessage, String exceptionStack);
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/UmiOcrClient.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/UmiOcrClient.java
new file mode 100644
index 0000000..02138df
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/UmiOcrClient.java
@@ -0,0 +1,241 @@
+package com.zhpb.gdyd_zhpb_zgf.service;
+
+import com.fasterxml.jackson.databind.JsonNode;
+import com.fasterxml.jackson.databind.ObjectMapper;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.http.*;
+import org.springframework.stereotype.Service;
+import org.springframework.util.LinkedMultiValueMap;
+import org.springframework.util.MultiValueMap;
+import org.springframework.web.client.RestTemplate;
+
+import java.io.File;
+import java.nio.file.Files;
+import java.time.LocalDateTime;
+import java.util.HashMap;
+import java.util.List;
+import java.util.Map;
+
+/**
+ * Umi-OCR HTTP API 客户端
+ * 封装了文档识别的完整异步流程
+ */
+@Service
+public class UmiOcrClient {
+
+    private static final Logger logger = LoggerFactory.getLogger(UmiOcrClient.class);
+
+    private final String baseUrl;
+    private final RestTemplate restTemplate;
+    private final ObjectMapper objectMapper;
+
+    public UmiOcrClient(String baseUrl, RestTemplate restTemplate, ObjectMapper objectMapper) {
+        this.baseUrl = baseUrl.replaceAll("/$", "");
+        this.restTemplate = restTemplate;
+        this.objectMapper = objectMapper;
+        logger.info("UmiOcrClient 初始化完成。目标地址: {}", this.baseUrl);
+    }
+
+    /**
+     * 发送HTTP请求的通用内部方法
+     */
+    private JsonNode sendRequest(String method, String endpoint, Map<String, Object> data,
+                                MultiValueMap<String, Object> formData) {
+        String url = baseUrl + endpoint;
+        HttpHeaders headers = new HttpHeaders();
+
+        try {
+            HttpEntity<?> entity;
+            if (formData != null) {
+                headers.setContentType(MediaType.MULTIPART_FORM_DATA);
+                entity = new HttpEntity<>(formData, headers);
+            } else {
+                headers.setContentType(MediaType.APPLICATION_JSON);
+                String jsonData = objectMapper.writeValueAsString(data != null ? data : Map.of());
+                entity = new HttpEntity<>(jsonData, headers);
+            }
+
+            ResponseEntity<String> response;
+            if ("GET".equalsIgnoreCase(method)) {
+                response = restTemplate.exchange(url, HttpMethod.GET, entity, String.class);
+            } else if ("POST".equalsIgnoreCase(method)) {
+                response = restTemplate.exchange(url, HttpMethod.POST, entity, String.class);
+            } else {
+                throw new IllegalArgumentException("不支持的HTTP方法: " + method);
+            }
+
+            if (response.getStatusCode() != HttpStatus.OK) {
+                throw new RuntimeException("HTTP请求失败: " + response.getStatusCode());
+            }
+
+            String responseBody = response.getBody();
+            if (responseBody == null || responseBody.trim().isEmpty()) {
+                // 返回空的成功响应
+                return objectMapper.createObjectNode()
+                    .put("code", 100)
+                    .put("data", "Success (No content)");
+            }
+
+            return objectMapper.readTree(responseBody);
+
+        } catch (Exception e) {
+            logger.error("请求 {} 失败", url, e);
+            throw new RuntimeException("请求错误: " + e.getMessage(), e);
+        }
+    }
+
+    // ==========================================================================
+    // 📚 文档识别（PDF等文件）接口封装
+    // ==========================================================================
+
+    /**
+     * 0. 准备工作：参数查询 (/api/doc/get_options)
+     */
+    public JsonNode docGetOptions() {
+        return sendRequest("GET", "/api/doc/get_options", null, null);
+    }
+
+    /**
+     * 1. 上传待识别文档 (/api/doc/upload)，返回任务ID
+     */
+    public String docUpload(String filePath, Map<String, Object> options) throws Exception {
+        logger.info("🚀 上传文件: {}", filePath);
+
+        File file = new File(filePath);
+        if (!file.exists()) {
+            throw new RuntimeException("文件不存在: " + filePath);
+        }
+
+        // 清理可能有问题的选项
+        Map<String, Object> cleanOptions = new HashMap<>();
+        if (options != null) {
+            cleanOptions.putAll(options);
+            // 移除可能导致问题的页数范围设置，让OCR服务自动处理
+            cleanOptions.remove("pageRangeEnd");
+            cleanOptions.remove("pageRangeStart");
+        }
+
+        MultiValueMap<String, Object> formData = new LinkedMultiValueMap<>();
+        formData.add("json", objectMapper.writeValueAsString(cleanOptions));
+        formData.add("file", new org.springframework.core.io.FileSystemResource(file));
+
+        JsonNode response = sendRequest("POST", "/api/doc/upload", null, formData);
+
+        if (response.get("code").asInt() == 100) {
+            return response.get("data").asText();
+        } else {
+            throw new RuntimeException("文件上传失败: " + response.get("data").asText());
+        }
+    }
+
+    /**
+     * 2. 查询任务状态 (/api/doc/result)
+     */
+    public JsonNode docQueryResult(String taskId, boolean isData, boolean isUnread, String dataFormat) {
+        Map<String, Object> requestData = new HashMap<>();
+        requestData.put("id", taskId);
+        requestData.put("is_data", isData);
+        requestData.put("is_unread", isUnread);
+        requestData.put("format", dataFormat);
+
+        return sendRequest("POST", "/api/doc/result", requestData, null);
+    }
+
+    /**
+     * 3. 获取结果下载链接 (/api/doc/download)
+     */
+    public Map<String, String> docGetDownloadLink(String taskId, List<String> fileTypes, boolean ignoreBlank) {
+        Map<String, Object> requestData = new HashMap<>();
+        requestData.put("id", taskId);
+        requestData.put("file_types", fileTypes);
+        requestData.put("ignore_blank", ignoreBlank);
+
+        JsonNode response = sendRequest("POST", "/api/doc/download", requestData, null);
+
+        if (response.get("code").asInt() == 100) {
+            Map<String, String> result = new HashMap<>();
+            result.put("download_url", baseUrl + response.get("data").asText());
+            result.put("file_name", response.get("name").asText());
+            return result;
+        } else {
+            throw new RuntimeException("生成下载链接失败: " + response.get("data").asText());
+        }
+    }
+
+    /**
+     * 5. 任务清理 (/api/doc/clear/<id>)
+     */
+    public boolean docClearTask(String taskId) {
+        JsonNode response = sendRequest("GET", "/api/doc/clear/" + taskId, null, null);
+
+        if (response.get("code").asInt() == 100) {
+            logger.info("🧹 任务ID {} 清理成功", taskId);
+            return true;
+        } else {
+            logger.warn("⚠️ 任务ID {} 清理失败或不存在: {}", taskId, response.get("data").asText());
+            return false;
+        }
+    }
+
+    /**
+     * 同步执行整个文档OCR流程 (上传 -> 轮询 -> 下载链接 -> 清理)
+     */
+    public Map<String, Object> runDocOcrSync(String filePath, Map<String, Object> options,
+                                           List<String> downloadFileTypes, int interval) throws Exception {
+        String taskId = null;
+        try {
+            // 1. 上传文件，获取任务ID
+            taskId = docUpload(filePath, options);
+            logger.info("📝 任务启动成功，任务ID: {}", taskId);
+
+            // 2. 轮询任务状态
+            logger.info("⏳ 正在轮询任务状态...");
+            while (true) {
+                JsonNode statusRes = docQueryResult(taskId, false, true, "dict");
+
+                boolean isDone = statusRes.get("is_done").asBoolean(false);
+                String state = statusRes.get("state").asText("unknown");
+                int processed = statusRes.get("processed_count").asInt(0);
+                String total = statusRes.get("pages_count").asText("?");
+
+                logger.info("   - 状态: {} ({}/{} 页已处理)", state.toUpperCase(), processed, total);
+
+                if (isDone) {
+                    if ("success".equals(state)) {
+                        logger.info("✅ 任务完成！");
+                        break;
+                    } else {
+                        String message = statusRes.get("message").asText("未知错误");
+                        throw new RuntimeException("任务执行失败 (State: " + state + "): " + message);
+                    }
+                }
+
+                Thread.sleep(interval * 1000L);
+            }
+
+            // 3. 获取下载链接
+            Map<String, String> downloadInfo = docGetDownloadLink(taskId, downloadFileTypes, true);
+            logger.info("🔗 下载链接生成成功。文件: {}", downloadInfo.get("file_name"));
+
+            // 4. 获取识别文本（可选步骤，通过查询获取）
+            JsonNode textResult = docQueryResult(taskId, true, true, "text");
+            String recognizedText = textResult.get("data").asText();
+
+            Map<String, Object> result = new HashMap<>();
+            result.put("code", 100);
+            result.put("task_id", taskId);
+            result.put("download_info", downloadInfo);
+            result.put("recognized_text", recognizedText);
+            return result;
+
+        } catch (Exception e) {
+            throw new RuntimeException("文档识别流程失败: " + e.getMessage(), e);
+        } finally {
+            // 5. 任务清理
+            if (taskId != null) {
+                docClearTask(taskId);
+            }
+        }
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/UserLogService.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/UserLogService.java
new file mode 100644
index 0000000..19a1c63
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/UserLogService.java
@@ -0,0 +1,98 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.service
+ * @ClassName: UserLogService
+ * @Description: 用户操作日志服务接口
+ * @Author: 
+ * @Date: 2025-11-28
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.service;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.UserLog;
+import jakarta.servlet.http.HttpServletRequest;
+import java.util.List;
+
+/**
+ * 用户操作日志服务接口
+ */
+public interface UserLogService {
+
+    /**
+     * 记录用户操作日志
+     *
+     * @param userId 用户ID
+     * @param userName 用户名
+     * @param projectId 项目ID
+     * @param operationType 操作类型
+     * @param operationDetail 操作详情
+     * @param relatedId 关联ID
+     * @param request HTTP请求对象（用于获取IP地址和User-Agent）
+     */
+    void logUserOperation(String userId, String userName, String projectId, 
+                         String operationType, String operationDetail, 
+                         String relatedId, HttpServletRequest request);
+
+    /**
+     * 记录用户操作日志（简化版）
+     *
+     * @param userId 用户ID
+     * @param userName 用户名
+     * @param operationType 操作类型
+     * @param operationDetail 操作详情
+     * @param request HTTP请求对象
+     */
+    void logUserOperation(String userId, String userName, 
+                         String operationType, String operationDetail, 
+                         HttpServletRequest request);
+
+    /**
+     * 根据用户ID查询操作日志
+     *
+     * @param userId 用户ID
+     * @return 操作日志列表
+     */
+    List<UserLog> getUserLogsByUserId(String userId);
+
+    /**
+     * 根据项目ID查询操作日志
+     *
+     * @param projectId 项目ID
+     * @return 操作日志列表
+     */
+    List<UserLog> getUserLogsByProjectId(String projectId);
+
+    /**
+     * 根据操作类型查询操作日志
+     *
+     * @param operationType 操作类型
+     * @return 操作日志列表
+     */
+    List<UserLog> getUserLogsByOperationType(String operationType);
+
+    /**
+     * 根据用户ID和操作类型查询操作日志
+     *
+     * @param userId 用户ID
+     * @param operationType 操作类型
+     * @return 操作日志列表
+     */
+    List<UserLog> getUserLogsByUserIdAndOperationType(String userId, String operationType);
+
+    /**
+     * 统计指定用户的操作次数
+     *
+     * @param userId 用户ID
+     * @return 操作次数
+     */
+    Long countUserOperations(String userId);
+
+    /**
+     * 统计指定操作类型的次数
+     *
+     * @param operationType 操作类型
+     * @return 操作次数
+     */
+    Long countOperationsByType(String operationType);
+}
\ No newline at end of file
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/UserService.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/UserService.java
new file mode 100644
index 0000000..f0af6b6
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/UserService.java
@@ -0,0 +1,55 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.service
+ * @ClassName: UserService
+ * @Description: 用户服务接口，定义用户查询、创建、更新等功能
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.service;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.User;
+
+/**
+ * 用户服务接口
+ */
+public interface UserService {
+    
+    /**
+     * 根据用户ID查找用户
+     * 
+     * @param userId 用户ID
+     * @return 用户对象
+     */
+    User findByUserId(String userId);
+    
+    /**
+     * 创建新用户
+     * 
+     * @param userId 用户ID
+     * @param name 用户名称
+     * @return 用户对象
+     */
+    User createUser(String userId, String name);
+    
+    /**
+     * 创建新用户（包含项目ID）
+     * 
+     * @param userId 用户ID
+     * @param name 用户名称
+     * @param projectId 项目ID
+     * @return 用户对象
+     */
+    User createUser(String userId, String name, String projectId);
+    
+    /**
+     * 更新用户信息
+     * 
+     * @param user 用户对象
+     * @return 更新后的用户对象
+     */
+    User updateUser(User user);
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/VectorizationService.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/VectorizationService.java
new file mode 100644
index 0000000..39377a9
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/VectorizationService.java
@@ -0,0 +1,42 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.service
+ * @ClassName: VectorizationService
+ * @Description: 向量化服务接口
+ * @Author: 张志锋
+ * @Date: 2025-11-05
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.service;
+
+import java.util.List;
+
+/**
+ * 向量化服务接口
+ */
+public interface VectorizationService {
+
+    /**
+     * 向量化单个文本
+     *
+     * @param text 待向量化的文本
+     * @return 向量数组
+     */
+    float[] vectorizeText(String text);
+
+    /**
+     * 批量向量化文本
+     *
+     * @param texts 待向量化的文本列表
+     * @return 向量数组列表
+     */
+    List<float[]> vectorizeTexts(List<String> texts);
+
+    /**
+     * 获取向量维度
+     *
+     * @return 向量维度
+     */
+    int getDimension();
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/ApiStatisticsServiceImpl.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/ApiStatisticsServiceImpl.java
new file mode 100644
index 0000000..c061ff2
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/ApiStatisticsServiceImpl.java
@@ -0,0 +1,43 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.service.impl
+ * @ClassName: ApiStatisticsServiceImpl
+ * @Description: API访问统计服务实现类，基于数据库日志表统计API调用次数
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.service.impl;
+
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.ApiLogRepository;
+import com.zhpb.gdyd_zhpb_zgf.service.ApiStatisticsService;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.stereotype.Service;
+
+/**
+ * API访问统计服务实现类（基于数据库日志表）
+ */
+@Service
+public class ApiStatisticsServiceImpl implements ApiStatisticsService {
+
+    @Autowired
+    private ApiLogRepository apiLogRepository;
+
+    @Override
+    public void incrementApiCount(String path) {
+        // 不再需要此方法，统计由日志表自动完成
+        // 保留接口是为了兼容性
+    }
+
+    @Override
+    public Long getTotalApiCount() {
+        return apiLogRepository.countTotal();
+    }
+
+    @Override
+    public Long getApiCount(String path) {
+        return apiLogRepository.countByApiPath(path);
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/ComparisonServiceImpl.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/ComparisonServiceImpl.java
new file mode 100644
index 0000000..1d40229
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/ComparisonServiceImpl.java
@@ -0,0 +1,368 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.service.impl
+ * @ClassName: ComparisonServiceImpl
+ * @Description: 供应商对比分析服务实现
+ * @Author: 张志锋
+ * @Date: 2025-11-04
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.service.impl;
+
+import com.zhpb.gdyd_zhpb_zgf.dto.ComparisonRequest;
+import com.zhpb.gdyd_zhpb_zgf.context.ModelRequestContextHolder;
+import com.zhpb.gdyd_zhpb_zgf.dto.ComparisonResponse;
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.ComparisonTask;
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.File;
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.PromptTemplate;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.ComparisonTaskRepository;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.FileRepository;
+import com.zhpb.gdyd_zhpb_zgf.service.AIModelService;
+import com.zhpb.gdyd_zhpb_zgf.service.ComparisonAsyncService;
+import com.zhpb.gdyd_zhpb_zgf.service.ComparisonService;
+import com.zhpb.gdyd_zhpb_zgf.service.PromptTemplateService;
+import com.zhpb.gdyd_zhpb_zgf.utils.HttpClientUtils;
+import com.zhpb.gdyd_zhpb_zgf.utils.TaskStatusUtils;
+import com.alibaba.fastjson2.JSON;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.scheduling.annotation.Async;
+import org.springframework.stereotype.Service;
+import org.springframework.transaction.annotation.Transactional;
+
+import java.time.LocalDateTime;
+import java.time.format.DateTimeFormatter;
+import java.util.*;
+import java.util.stream.Collectors;
+
+/**
+ * 供应商对比分析服务实现
+ */
+@Service
+public class ComparisonServiceImpl implements ComparisonService {
+
+    private static final Logger logger = LoggerFactory.getLogger(ComparisonServiceImpl.class);
+
+    @Autowired
+    private ComparisonTaskRepository comparisonTaskRepository;
+
+    @Autowired
+    private FileRepository fileRepository;
+
+    @Autowired
+    private PromptTemplateService promptTemplateService;
+
+    @Autowired
+    private AIModelService aiModelService;
+
+    @Autowired
+    private org.springframework.core.env.Environment environment;
+
+    @Autowired
+    private ComparisonAsyncService comparisonAsyncService;
+
+    @Autowired
+    private com.zhpb.gdyd_zhpb_zgf.repository.mysql.SupplierRepository supplierRepository;
+
+    @Override
+    public ComparisonResponse.VendorsData getProjectVendors(String projectId) {
+        logger.info("获取项目供应商数据: projectId={}", projectId);
+
+        ComparisonResponse.VendorsData result = new ComparisonResponse.VendorsData();
+        List<ComparisonResponse.VendorInfo> vendors = new ArrayList<>();
+
+        try {
+            // 获取项目中的所有文件
+            List<File> files = fileRepository.findByProjectId(projectId);
+            logger.info("项目 {} 中共有 {} 个文件", projectId, files.size());
+
+            // 按供应商分组
+            Map<String, List<File>> filesBySupplier = files.stream()
+                .filter(file -> file.getSupplierId() != null && !file.getSupplierId().trim().isEmpty())
+                .collect(Collectors.groupingBy(File::getSupplierId));
+
+            logger.info("按供应商分组后共有 {} 个供应商", filesBySupplier.size());
+
+            // 为每个供应商构建信息
+            for (Map.Entry<String, List<File>> entry : filesBySupplier.entrySet()) {
+                String supplierId = entry.getKey();
+                List<File> supplierFiles = entry.getValue();
+
+                ComparisonResponse.VendorInfo vendorInfo = new ComparisonResponse.VendorInfo();
+                vendorInfo.setVendorId(supplierId);
+
+                // 直接从数据库查询供应商名称
+                String supplierName = getSupplierNameFromDatabase(supplierId);
+                vendorInfo.setVendorName(supplierName);
+                vendorInfo.setVendorType("供应商");
+
+                // 构建文件信息列表
+                List<ComparisonResponse.FileInfo> fileInfos = supplierFiles.stream()
+                    .map(this::convertToFileInfo)
+                    .collect(Collectors.toList());
+
+                vendorInfo.setFiles(fileInfos);
+                vendors.add(vendorInfo);
+
+                logger.debug("供应商 {} ({}) 有 {} 个文件", supplierName, supplierId, fileInfos.size());
+            }
+
+            result.setVendors(vendors);
+            logger.info("成功获取 {} 个供应商的信息", vendors.size());
+
+        } catch (Exception e) {
+            logger.error("获取项目供应商数据失败", e);
+            throw new RuntimeException("获取供应商数据失败: " + e.getMessage());
+        }
+
+        return result;
+    }
+
+    /**
+     * 从数据库查询供应商名称
+     * 如果供应商名称为空或不存在，则返回supplierId
+     */
+    private String getSupplierNameFromDatabase(String supplierId) {
+        try {
+            com.zhpb.gdyd_zhpb_zgf.entity.mysql.Supplier supplier = supplierRepository.findBySupplierId(supplierId);
+            if (supplier != null && supplier.getSupplierName() != null && !supplier.getSupplierName().trim().isEmpty()) {
+                return supplier.getSupplierName();
+            }
+        } catch (Exception e) {
+            logger.warn("查询供应商名称失败: supplierId={}, error={}", supplierId, e.getMessage());
+        }
+        // 如果查不到或名称为空，返回supplierId
+        return supplierId;
+    }
+
+    @Override
+    @Transactional
+    public ComparisonResponse.StartTaskData startComparison(ComparisonRequest request) {
+        logger.info("启动供应商对比分析任务: projectId={}, userId={}, vendorCount={}, dimensionCount={}",
+            request.getProjectId(), request.getUserId(), request.getVendorIds().size(), request.getDimensions().size());
+
+        try {
+            // 生成任务ID
+            String taskId = "task-" + UUID.randomUUID().toString().replace("-", "").substring(0, 12);
+
+            // 创建对比任务
+            ComparisonTask task = new ComparisonTask();
+            task.setTaskId(taskId);
+            task.setProjectId(request.getProjectId());
+            task.setUserId(request.getUserId());
+            task.setVendorIds(JSON.toJSONString(request.getVendorIds()));
+            task.setDimensions(JSON.toJSONString(request.getDimensions()));
+            task.setStatus("pending");
+            task.setProgress(0);
+
+            // 保存任务
+            ComparisonTask savedTask = comparisonTaskRepository.save(task);
+            logger.info("对比分析任务已创建: taskId={}", taskId);
+
+            // 使用事务同步机制，确保事务提交后再调用异步方法
+            // 这解决了异步线程查询不到刚保存数据的问题
+            org.springframework.transaction.support.TransactionSynchronizationManager.registerSynchronization(
+                new org.springframework.transaction.support.TransactionSynchronization() {
+                    @Override
+                    public void afterCommit() {
+                        logger.info("事务已提交，开始触发异步对比分析任务: taskId={}", taskId);
+                        comparisonAsyncService.processComparisonAsync(taskId);
+                    }
+                });
+
+            // 返回任务启动信息
+            ComparisonResponse.StartTaskData result = new ComparisonResponse.StartTaskData();
+            result.setTaskId(taskId);
+            // 使用数据库中的实际状态
+            result.setStatus(TaskStatusUtils.toChineseStatus(savedTask.getStatus()));
+            result.setCreatedTime(savedTask.getCreateTime().format(DateTimeFormatter.ISO_LOCAL_DATE_TIME));
+
+            return result;
+
+        } catch (Exception e) {
+            logger.error("启动对比分析任务失败", e);
+            throw new RuntimeException("启动对比分析任务失败: " + e.getMessage());
+        }
+    }
+
+    @Override
+    @Transactional(readOnly = true)
+    public ComparisonResponse.TaskHistoryData getComparisonHistory(String projectId, String userId) {
+        logger.info("获取对比任务历史: projectId={}, userId={}", projectId, userId);
+
+        ComparisonResponse.TaskHistoryData result = new ComparisonResponse.TaskHistoryData();
+        List<ComparisonResponse.TaskInfo> tasks = new ArrayList<>();
+
+        try {
+            List<ComparisonTask> comparisonTasks = comparisonTaskRepository
+                .findByProjectIdAndUserIdOrderByCreateTimeDesc(projectId, userId);
+
+            int totalTasks = comparisonTasks.size();
+            for (int i = 0; i < totalTasks; i++) {
+                ComparisonTask task = comparisonTasks.get(i);
+                ComparisonResponse.TaskInfo taskInfo = new ComparisonResponse.TaskInfo();
+                
+                // 设置任务名称：对比任务N (最新的任务编号最大)
+                taskInfo.setTaskName("对比任务" + (totalTasks - i));
+                
+                taskInfo.setTaskId(task.getTaskId());
+                taskInfo.setProjectId(task.getProjectId());
+                taskInfo.setUserId(task.getUserId());
+                // 转换为中文状态
+                taskInfo.setStatus(TaskStatusUtils.toChineseStatus(task.getStatus()));
+
+                // 解析供应商ID和名称
+                List<String> vendorNames = new ArrayList<>();
+                try {
+                    List<String> vendorIds = JSON.parseArray(task.getVendorIds(), String.class);
+                    taskInfo.setVendorCount(vendorIds.size());
+                    // 查询供应商真实名称
+                    for (String vendorId : vendorIds) {
+                        com.zhpb.gdyd_zhpb_zgf.entity.mysql.Supplier supplier = 
+                            supplierRepository.findBySupplierId(vendorId);
+                        if (supplier != null && supplier.getSupplierName() != null) {
+                            vendorNames.add(supplier.getSupplierName());
+                        } else {
+                            // 如果查不到供应商，使用vendorId作为备用
+                            vendorNames.add(vendorId);
+                        }
+                    }
+                } catch (Exception e) {
+                    taskInfo.setVendorCount(0);
+                }
+                taskInfo.setVendors(vendorNames);
+
+                // 解析维度
+                List<String> dimensionNames = new ArrayList<>();
+                try {
+                    List<ComparisonRequest.ComparisonDimension> dimensions =
+                        JSON.parseArray(task.getDimensions(), ComparisonRequest.ComparisonDimension.class);
+                    taskInfo.setDimensionCount(dimensions.size());
+                    // 提取维度名称
+                    dimensionNames = dimensions.stream()
+                        .map(ComparisonRequest.ComparisonDimension::getValue)
+                        .collect(Collectors.toList());
+                } catch (Exception e) {
+                    taskInfo.setDimensionCount(0);
+                }
+                taskInfo.setDimensions(dimensionNames);
+
+                taskInfo.setCreatedTime(task.getCreateTime().format(DateTimeFormatter.ISO_LOCAL_DATE_TIME));
+                if (task.getCompletedTime() != null) {
+                    taskInfo.setCompletedTime(task.getCompletedTime().format(DateTimeFormatter.ISO_LOCAL_DATE_TIME));
+                    // 计算耗时（秒）
+                    long timeInSeconds = java.time.Duration.between(
+                        task.getCreateTime(), task.getCompletedTime()).getSeconds();
+                    taskInfo.setTime(timeInSeconds);
+                }
+
+                tasks.add(taskInfo);
+            }
+
+            result.setTasks(tasks);
+            logger.info("成功获取 {} 个对比任务的历史记录", tasks.size());
+
+        } catch (Exception e) {
+            logger.error("获取对比任务历史失败", e);
+            throw new RuntimeException("获取对比任务历史失败: " + e.getMessage());
+        }
+
+        return result;
+    }
+
+    @Override
+    @Transactional(readOnly = true)
+    public ComparisonResponse.ComparisonResultData getComparisonResult(String taskId) {
+        logger.info("获取对比任务结果: taskId={}", taskId);
+
+        try {
+            ComparisonTask task = comparisonTaskRepository.findByTaskId(taskId);
+            if (task == null) {
+                throw new RuntimeException("对比任务不存在: " + taskId);
+            }
+
+            ComparisonResponse.ComparisonResultData result = new ComparisonResponse.ComparisonResultData();
+            result.setTaskId(task.getTaskId());
+            // 转换为中文状态
+            result.setStatus(TaskStatusUtils.toChineseStatus(task.getStatus()));
+
+            // 如果任务已完成，解析结果
+            if ("completed".equals(task.getStatus()) && task.getResult() != null) {
+                try {
+                    List<ComparisonResponse.ResultMatrixItem> resultMatrix =
+                        JSON.parseArray(task.getResult(), ComparisonResponse.ResultMatrixItem.class);
+                    result.setResultMatrix(resultMatrix);
+                } catch (Exception e) {
+                    logger.warn("解析对比结果失败", e);
+                    result.setResultMatrix(new ArrayList<>());
+                }
+            } else {
+                result.setResultMatrix(new ArrayList<>());
+            }
+
+            return result;
+
+        } catch (Exception e) {
+            logger.error("获取对比任务结果失败", e);
+            throw new RuntimeException("获取对比任务结果失败: " + e.getMessage());
+        }
+    }
+
+    /**
+     * 从文件信息推断供应商名称
+     */
+    private String extractSupplierNameFromFile(File file) {
+        // 从文件路径中提取供应商名称
+        String filePath = file.getFilePath();
+        if (filePath != null) {
+            // 简单的路径解析逻辑，这里可以根据实际的文件路径结构进行调整
+            String[] parts = filePath.split("[\\\\/]");
+            for (String part : parts) {
+                // 优先匹配包含公司名称的路径段
+                if (part.contains("公司") || part.contains("有限公司") || part.contains("股份有限公司")) {
+                    // 清理路径中的多余信息，如"_虚拟标段_标包1_文件"
+                    String cleanName = part.replaceAll("(_虚拟标段)?(_标包\\d+)?(_文件)?$", "");
+                    return cleanName;
+                }
+            }
+            // 如果没找到公司名称，尝试其他关键词
+            for (String part : parts) {
+                if (part.contains("供应商") || part.contains("投标") || part.length() > 10) {
+                    return part.replaceAll("[_\\d]+$", ""); // 移除末尾的数字和下划线
+                }
+            }
+        }
+        return file.getSupplierId(); // 如果无法推断，返回供应商ID
+    }
+
+    /**
+     * 转换文件信息
+     */
+    private ComparisonResponse.FileInfo convertToFileInfo(File file) {
+        ComparisonResponse.FileInfo fileInfo = new ComparisonResponse.FileInfo();
+        fileInfo.setFileId(file.getFileId());
+        fileInfo.setFileName(file.getFileName());
+        fileInfo.setFileType(getFileTypeName(file.getFileTypeId()));
+        fileInfo.setFileSize(file.getFileSize());
+        return fileInfo;
+    }
+
+    /**
+     * 获取文件类型名称
+     */
+    private String getFileTypeName(Long fileTypeId) {
+        // 这里可以根据实际的文件类型表进行查询
+        // 简化处理，返回默认类型
+        if (fileTypeId != null) {
+            switch (fileTypeId.intValue()) {
+                case 5: return "技术文件";
+                case 6: return "商务文件";
+                case 7: return "报价文件";
+                default: return "其他文件";
+            }
+        }
+        return "未知类型";
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/DeepSeekServiceImpl.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/DeepSeekServiceImpl.java
new file mode 100644
index 0000000..9b3d6f1
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/DeepSeekServiceImpl.java
@@ -0,0 +1,174 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.service.impl
+ * @ClassName: DeepSeekServiceImpl
+ * @Description: DeepSeek AI模型服务实现
+ * @Author: 张志锋
+ * @Date: 2025-11-05
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.service.impl;
+
+import com.alibaba.fastjson2.JSON;
+import com.alibaba.fastjson2.JSONObject;
+import com.zhpb.gdyd_zhpb_zgf.context.ModelRequestContextHolder;
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.ModelRequestLog;
+import com.zhpb.gdyd_zhpb_zgf.service.AIModelService;
+import com.zhpb.gdyd_zhpb_zgf.service.ModelRequestLogService;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.beans.factory.annotation.Value;
+import org.springframework.http.*;
+import org.springframework.stereotype.Service;
+import org.springframework.web.client.RestTemplate;
+import org.springframework.web.client.HttpStatusCodeException;
+
+import java.util.Arrays;
+import java.util.HashMap;
+import java.util.Map;
+
+/**
+ * DeepSeek AI模型服务实现
+ */
+@Service
+public class DeepSeekServiceImpl implements AIModelService {
+
+    private static final Logger logger = LoggerFactory.getLogger(DeepSeekServiceImpl.class);
+
+    @Value("${deepseek.api-key}")
+    private String apiKey;
+
+    @Value("${deepseek.api-url}")
+    private String apiUrl;
+
+    @Value("${deepseek.model}")
+    private String model;
+
+    @Value("${deepseek.temperature:0.7}")
+    private double temperature;
+
+    // max_tokens不再限制，让模型自由决定输出长度
+
+    private final RestTemplate restTemplate;
+    private final ModelRequestLogService modelRequestLogService;
+
+    /**
+     * 创建带超时配置的RestTemplate
+     */
+    private static RestTemplate createRestTemplateWithTimeout() {
+        org.springframework.http.client.SimpleClientHttpRequestFactory factory = 
+            new org.springframework.http.client.SimpleClientHttpRequestFactory();
+        factory.setConnectTimeout(30000);  // 连接超时30秒
+        factory.setReadTimeout(180000);    // 读取超时180秒（AI模型响应可能较慢，特别是QwQ推理模型）
+        return new RestTemplate(factory);
+    }
+
+    public DeepSeekServiceImpl(ModelRequestLogService modelRequestLogService) {
+        this.restTemplate = createRestTemplateWithTimeout();
+        this.modelRequestLogService = modelRequestLogService;
+    }
+
+    @Override
+    public String generateResponse(String systemPrompt, String userMessage) {
+        return generateResponse(systemPrompt, userMessage, false);
+    }
+
+    @Override
+    public String generateResponse(String systemPrompt, String userMessage, boolean stream) {
+        long startTime = System.currentTimeMillis();
+        String requestJson = null;
+        String responseBody = null;
+        boolean success = false;
+        String errorMessage = null;
+        Integer promptTokens = null;
+        Integer completionTokens = null;
+        Integer totalTokens = null;
+        try {
+            logger.info("调用DeepSeek API - 模型: {}, 系统提示长度: {}, 用户消息长度: {}",
+                       model, systemPrompt.length(), userMessage.length());
+
+            // 构建请求体
+            Map<String, Object> requestBody = new HashMap<>();
+            requestBody.put("model", model);
+            requestBody.put("temperature", temperature);
+            // 不设置max_tokens限制，让模型自由决定输出长度
+            requestBody.put("stream", stream);
+
+            // 构建消息
+            Map<String, String> systemMessage = new HashMap<>();
+            systemMessage.put("role", "system");
+            systemMessage.put("content", systemPrompt);
+
+            Map<String, String> userMessageMap = new HashMap<>();
+            userMessageMap.put("role", "user");
+            userMessageMap.put("content", userMessage);
+
+            requestBody.put("messages", Arrays.asList(systemMessage, userMessageMap));
+            requestJson = JSON.toJSONString(requestBody);
+
+            // 设置请求头
+            HttpHeaders headers = new HttpHeaders();
+            headers.setContentType(MediaType.APPLICATION_JSON);
+            headers.setBearerAuth(apiKey);
+
+            HttpEntity<Map<String, Object>> requestEntity = new HttpEntity<>(requestBody, headers);
+
+            // 发送请求
+            ResponseEntity<String> responseEntity = restTemplate.exchange(
+                apiUrl, HttpMethod.POST, requestEntity, String.class);
+
+            if (responseEntity.getStatusCode() == HttpStatus.OK) {
+                responseBody = responseEntity.getBody();
+                JSONObject jsonResponse = JSON.parseObject(responseBody);
+
+                // 提取回复内容
+                String content = jsonResponse.getJSONArray("choices")
+                    .getJSONObject(0)
+                    .getJSONObject("message")
+                    .getString("content");
+
+                JSONObject usage = jsonResponse.getJSONObject("usage");
+                if (usage != null) {
+                    promptTokens = usage.getInteger("prompt_tokens");
+                    completionTokens = usage.getInteger("completion_tokens");
+                    totalTokens = usage.getInteger("total_tokens");
+                }
+
+                success = true;
+                logger.info("DeepSeek API调用成功，响应长度: {}", content.length());
+                return content;
+            } else {
+                logger.error("DeepSeek API调用失败，状态码: {}", responseEntity.getStatusCode());
+                throw new RuntimeException("DeepSeek API调用失败: " + responseEntity.getStatusCode());
+            }
+
+        } catch (HttpStatusCodeException e) {
+            responseBody = e.getResponseBodyAsString();
+            errorMessage = e.getMessage();
+            logger.error("调用DeepSeek API异常: {}", e.getMessage());
+            throw new RuntimeException("调用AI模型失败: " + e.getMessage(), e);
+        } catch (Exception e) {
+            errorMessage = e.getMessage();
+            logger.error("调用DeepSeek API异常", e);
+            throw new RuntimeException("调用AI模型失败: " + e.getMessage(), e);
+        } finally {
+            ModelRequestLog logEntry = new ModelRequestLog();
+            logEntry.setRequestSource(ModelRequestContextHolder.getRequestSourceOrDefault("AI_PUBLIC"));
+            logEntry.setRequestUrl(apiUrl);
+            logEntry.setRequestMethod("POST");
+            logEntry.setModelName(model);
+            logEntry.setApiKey(apiKey);
+            logEntry.setRequestPayload(requestJson);
+            logEntry.setResponsePayload(responseBody);
+            logEntry.setStatus(success ? "SUCCESS" : "FAILED");
+            logEntry.setDurationMs(System.currentTimeMillis() - startTime);
+            logEntry.setPromptTokens(promptTokens);
+            logEntry.setCompletionTokens(completionTokens);
+            logEntry.setTotalTokens(totalTokens);
+            logEntry.setErrorMessage(errorMessage);
+            modelRequestLogService.record(logEntry);
+            ModelRequestContextHolder.clear();
+        }
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/FilePreviewServiceImpl.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/FilePreviewServiceImpl.java
new file mode 100644
index 0000000..3029a44
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/FilePreviewServiceImpl.java
@@ -0,0 +1,234 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.service.impl
+ * @ClassName: FilePreviewServiceImpl
+ * @Description: 文件预览服务实现类
+ * @Author: 张志锋
+ * @Date: 2025-11-01
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.service.impl;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.File;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.FileRepository;
+import com.zhpb.gdyd_zhpb_zgf.service.FilePreviewService;
+import com.zhpb.gdyd_zhpb_zgf.service.FilePathResolver;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.beans.factory.annotation.Value;
+import org.springframework.scheduling.annotation.Async;
+import org.springframework.stereotype.Service;
+import org.springframework.transaction.annotation.Transactional;
+
+import java.io.IOException;
+import java.io.InputStream;
+import java.nio.file.Files;
+import java.nio.file.Path;
+import java.nio.file.Paths;
+import java.util.Optional;
+
+/**
+ * 文件预览服务实现类
+ */
+@Service
+public class FilePreviewServiceImpl implements FilePreviewService {
+
+    private static final Logger logger = LoggerFactory.getLogger(FilePreviewServiceImpl.class);
+
+    @Autowired
+    private FileRepository fileRepository;
+
+    @Autowired
+    private FilePathResolver filePathResolver;
+
+    @Value("${file.preview-path:./previews}")
+    private String previewPath;
+
+    @Override
+    public InputStream previewPdf(String fileId) throws Exception {
+        Optional<File> fileOpt = fileRepository.findByFileId(fileId);
+        if (!fileOpt.isPresent()) {
+            throw new IllegalArgumentException("文件不存在: " + fileId);
+        }
+
+        File file = fileOpt.get();
+        
+        Path previewCandidate = resolvePreviewFilePath(file.getPreviewPath());
+        if (previewCandidate != null) {
+            logger.debug("使用预览文件: {}", previewCandidate);
+            return Files.newInputStream(previewCandidate);
+        }
+        
+        Path originalPath = resolveOriginalFilePath(file);
+        if (originalPath == null || !Files.exists(originalPath)) {
+            throw new IOException("文件不存在: " + file.getFilePath());
+        }
+        
+        String extension = file.getFileExtension();
+        if (extension != null && extension.equalsIgnoreCase("pdf")) {
+            return Files.newInputStream(originalPath);
+        }
+        
+        throw new UnsupportedOperationException("不支持的文件类型: " + extension);
+    }
+
+    @Override
+    public InputStream previewWord(String fileId) throws Exception {
+        Optional<File> fileOpt = fileRepository.findByFileId(fileId);
+        if (!fileOpt.isPresent()) {
+            throw new IllegalArgumentException("文件不存在: " + fileId);
+        }
+
+        File file = fileOpt.get();
+        Path originalPath = resolveOriginalFilePath(file);
+        String filePath = originalPath != null ? originalPath.toString() : file.getFilePath();
+        
+        // 如果有预览路径（可能是URL或文件路径），如果是文件路径且存在则使用转换后的PDF
+        if (file.getPreviewPath() != null && !file.getPreviewPath().isEmpty()) {
+            String previewPathValue = file.getPreviewPath();
+            Path previewCandidate = resolvePreviewFilePath(previewPathValue);
+            if (previewCandidate != null) {
+                logger.debug("使用Word转换后的PDF预览文件: {}", previewCandidate);
+                return Files.newInputStream(previewCandidate);
+            }
+            // 如果是URL格式，说明预览接口已配置，需要查找转换后的PDF文件
+        }
+        
+        // 查找Word转换后的PDF文件（存储在预览目录）
+        String extension = file.getFileExtension();
+        if (extension != null && (extension.toLowerCase().equals("doc") || 
+                                  extension.toLowerCase().equals("docx"))) {
+            Path previewDir = Paths.get(previewPath);
+            String previewFileName = Paths.get(filePath).getFileName().toString()
+                .replaceAll("\\.(doc|docx)$", ".pdf");
+            Path convertedPdfPath = previewDir.resolve(fileId + "_" + previewFileName);
+            
+            if (Files.exists(convertedPdfPath)) {
+                logger.debug("使用已转换的PDF文件: {}", convertedPdfPath);
+                return Files.newInputStream(convertedPdfPath);
+            }
+            
+            // 如果还没有转换，抛出异常提示需要转换
+            throw new UnsupportedOperationException("Word转PDF功能暂未实现，文件类型: " + extension + "，需要先将Word转换为PDF");
+        }
+        
+        throw new UnsupportedOperationException("不支持的文件类型: " + extension);
+    }
+
+    @Override
+    @Async
+    @Transactional
+    public void generatePreview(String fileId) {
+        try {
+            // 先检查是否已经有预览URL（大部分文件在保存时已设置）
+            Optional<File> fileOpt = fileRepository.findByFileId(fileId);
+            if (!fileOpt.isPresent()) {
+                logger.warn("文件不存在，无法生成预览: {}", fileId);
+                return;
+            }
+
+            File file = fileOpt.get();
+            
+            // 如果已经有预览URL，直接跳过（大部分文件在保存时已设置）
+            if (file.getPreviewPath() != null && !file.getPreviewPath().isEmpty()) {
+                // 如果是URL格式，说明已配置，跳过
+                if (file.getPreviewPath().startsWith("/api/v1/files/preview/")) {
+                    logger.debug("预览URL已配置，跳过生成: {} -> {}", fileId, file.getPreviewPath());
+                    return;
+                }
+                // 如果是旧的文件路径格式，检查文件是否存在
+                String previewPathValue = file.getPreviewPath();
+                if (previewPathValue.contains(":\\") || 
+                    (previewPathValue.startsWith("/") && !previewPathValue.startsWith("/api/"))) {
+                    Path previewFilePath = Paths.get(previewPathValue);
+                    if (Files.exists(previewFilePath)) {
+                        logger.debug("预览文件已存在，跳过生成: {}", fileId);
+                        return;
+                    }
+                }
+            }
+            
+            // 只有没有预览URL的文件才需要设置（主要是旧数据或特殊情况）
+            String extension = file.getFileExtension();
+            if (extension == null) {
+                logger.debug("文件无扩展名，跳过预览生成: {}", fileId);
+                return;
+            }
+            
+            String ext = extension.toLowerCase();
+
+            // PDF文件：设置预览接口URL
+            if ("pdf".equals(ext)) {
+                String previewUrl = "/api/v1/files/stream/pdf/" + fileId;
+                file.setPreviewPath(previewUrl);
+                fileRepository.save(file);
+                logger.info("PDF文件预览URL已设置: {} -> {}", fileId, previewUrl);
+                return;
+            }
+
+            // Word文件：设置预览接口URL（转换为PDF后预览）
+            if ("doc".equals(ext) || "docx".equals(ext)) {
+                String previewUrl = "/api/v1/files/preview/word/" + fileId;
+                file.setPreviewPath(previewUrl);
+                fileRepository.save(file);
+                logger.info("Word文件预览URL已设置: {} -> {}", fileId, previewUrl);
+                
+                // TODO: 实际转换Word到PDF的功能可以在预览时异步处理
+                logger.debug("Word转PDF功能将在预览时异步处理: {}", fileId);
+            }
+
+        } catch (Exception e) {
+            logger.error("生成文件预览失败: {}, 错误: {}", fileId, e.getMessage(), e);
+        }
+    }
+
+    private Path resolveOriginalFilePath(File file) throws IOException {
+        try {
+            Optional<Path> resolvedOpt = filePathResolver.resolveExistingPath(
+                file.getFilePath(), file.getProjectId(), file.getFileName());
+            if (resolvedOpt.isEmpty()) {
+                return null;
+            }
+            Path resolvedPath = resolvedOpt.get();
+            String normalized = filePathResolver.normalizeForStorage(resolvedPath);
+            if (normalized != null && !normalized.equals(file.getFilePath())) {
+                file.setFilePath(normalized);
+                fileRepository.save(file);
+                logger.info("自动纠正文件路径: {} -> {}", file.getFileId(), normalized);
+            }
+            return resolvedPath;
+        } catch (Exception e) {
+            logger.warn("解析文件路径失败: fileId={}, error={}", file.getFileId(), e.getMessage());
+            return null;
+        }
+    }
+
+    private Path resolvePreviewFilePath(String previewPathValue) {
+        if (previewPathValue == null || previewPathValue.isBlank()) {
+            return null;
+        }
+        if (previewPathValue.startsWith("/api/")) {
+            return null;
+        }
+        try {
+            Path candidate = filePathResolver.resolveStoredPath(previewPathValue);
+            if (candidate != null && Files.exists(candidate)) {
+                return candidate;
+            }
+        } catch (Exception ignore) {
+            // ignore
+        }
+        try {
+            Path direct = Paths.get(previewPathValue);
+            if (Files.exists(direct)) {
+                return direct;
+            }
+        } catch (Exception ignore) {
+            // ignore
+        }
+        return null;
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/FileServiceImpl.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/FileServiceImpl.java
new file mode 100644
index 0000000..3411cbe
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/FileServiceImpl.java
@@ -0,0 +1,2435 @@
+
+/*
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.service.impl
+ * @ClassName: FileServiceImpl
+ * @Description: 文件服务实现类，处理文件上传、ZIP解压（支持递归解压嵌套ZIP）、文件路径管理等功能
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.service.impl;
+import com.zhpb.gdyd_zhpb_zgf.service.*;
+import org.apache.commons.compress.utils.IOUtils;
+import com.zhpb.gdyd_zhpb_zgf.context.ModelRequestContextHolder;
+import org.apache.commons.compress.archivers.zip.ZipArchiveInputStream;
+import org.apache.commons.compress.archivers.zip.ZipArchiveEntry;
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.File;
+import org.apache.commons.compress.archivers.zip.ZipArchiveInputStream;
+import org.apache.commons.compress.archivers.zip.ZipArchiveEntry;
+import org.apache.commons.compress.utils.IOUtils;
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.FileDirectory;
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.FileType;
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.Project;
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.Supplier;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.FileRepository;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.FileTypeRepository;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.ProjectRepository;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.SupplierRepository;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.FileDirectoryRepository;
+import com.zhpb.gdyd_zhpb_zgf.utils.HttpClientUtils;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.beans.factory.annotation.Value;
+import org.springframework.stereotype.Service;
+import org.springframework.transaction.annotation.Transactional;
+import org.springframework.transaction.annotation.Propagation;
+import org.springframework.transaction.support.TransactionTemplate;
+import org.springframework.web.multipart.MultipartFile;
+import org.springframework.scheduling.annotation.Async;
+
+import org.apache.pdfbox.pdmodel.PDDocument;
+import org.apache.pdfbox.pdmodel.interactive.documentnavigation.outline.PDDocumentOutline;
+import org.apache.pdfbox.pdmodel.interactive.documentnavigation.outline.PDOutlineItem;
+
+import jakarta.annotation.PostConstruct;
+import jakarta.annotation.PreDestroy;
+import com.alibaba.fastjson2.JSON;
+import java.io.*;
+import java.nio.file.Files;
+import java.nio.file.Path;
+import java.nio.file.Paths;
+import java.time.LocalDateTime;
+import java.time.format.DateTimeFormatter;
+import java.util.*;
+import java.util.regex.Matcher;
+import java.util.regex.Pattern;
+import java.util.concurrent.CompletableFuture;
+import java.util.concurrent.ExecutorService;
+import java.util.concurrent.Executors;
+import java.util.stream.Collectors;
+import java.util.Set;
+import java.util.UUID;
+import java.util.zip.ZipEntry;
+import java.util.zip.ZipFile;
+import java.util.zip.ZipInputStream;
+import java.util.Enumeration;
+import java.nio.charset.Charset;
+
+/**
+ * 文件服务实现类
+ */
+@Service
+public class FileServiceImpl implements FileService {
+
+    private static final Logger logger = LoggerFactory.getLogger(FileServiceImpl.class);
+
+    @Autowired
+    private ProjectRepository projectRepository;
+
+    @Autowired
+    private FileRepository fileRepository;
+
+    @Autowired
+    private SupplierRepository supplierRepository;
+
+    @Autowired
+    private FileTypeRepository fileTypeRepository;
+
+    @Autowired
+    private FileTypeService fileTypeService;
+
+    @Autowired
+    private PdfReadService pdfReadService;
+
+    @Autowired
+    private com.zhpb.gdyd_zhpb_zgf.service.FilePreviewService filePreviewService;
+
+    @Autowired
+    private com.zhpb.gdyd_zhpb_zgf.service.PromptTemplateService promptTemplateService;
+
+    @Autowired
+    private FileDirectoryRepository fileDirectoryRepository;
+
+    @Autowired
+    private TransactionTemplate transactionTemplate;
+
+    @Autowired
+    private AIModelService aiModelService;
+
+    @Autowired
+    private ChapterContentResolver chapterContentResolver;
+
+    @Autowired
+    private FilePathResolver filePathResolver;
+
+    @Autowired
+    private OcrService ocrService;
+
+    private ExecutorService directoryExecutor;
+
+    @Value("${vectorize.base-url:http://localhost:8001}")
+    private String vectorizeBaseUrl;
+
+    @Value("${deepseek.api-key:}")
+    private String deepseekApiKey;
+
+    @Value("${deepseek.api-url:https://api.deepseek.com/v1/chat/completions}")
+    private String deepseekApiUrl;
+
+    @Value("${deepseek.model:deepseek-chat}")
+    private String deepseekModel;
+
+    @Value("${deepseek.temperature:0.7}")
+    private double deepseekTemperature;
+
+    @Value("${deepseek.max-tokens:4096}")
+    private int deepseekMaxTokens;
+
+    @Value("${file.download-replace-source:}")
+    private String downloadReplaceSource;
+
+    @Value("${file.download-replace-target:}")
+    private String downloadReplaceTarget;
+
+    @Value("${umi-ocr.enabled:false}")
+    private boolean ocrEnabled;
+    
+    // 存储处理后的绝对路径
+    private String absoluteUploadPath;
+    private String absoluteExtractPath;
+    
+    // 时间格式
+    private static final DateTimeFormatter DATE_TIME_FORMATTER = DateTimeFormatter.ofPattern("yyyyMMdd_HHmmss");
+
+    // 7zip可用性标志
+    private boolean sevenZipAvailable = false;
+    private String sevenZipCommand = null;
+    
+    /**
+     * 初始化路径和7zip检测
+     */
+    @PostConstruct
+    public void init() {
+        absoluteUploadPath = filePathResolver.getAbsoluteUploadPath().toString();
+        absoluteExtractPath = filePathResolver.getAbsoluteExtractPath().toString();
+        logger.info("文件上传路径: {}", absoluteUploadPath);
+        logger.info("文件解压路径: {}", absoluteExtractPath);
+
+        // 检测7zip可用性
+        checkSevenZipAvailability();
+
+        // 目录抽取线程池大小设为3，避免数据库连接池耗尽
+        // 因为目录抽取涉及LLM调用且在事务中运行，持有连接时间较长
+        int poolSize = 3;
+        directoryExecutor = Executors.newFixedThreadPool(poolSize);
+        logger.info("目录抽取线程池初始化完成, 核心线程数: {}", poolSize);
+
+        // 记录OCR功能状态
+        logger.info("OCR功能状态: {}", ocrEnabled ? "已启用" : "已禁用");
+        if (ocrEnabled) {
+            logger.info("PDF文件上传后将自动进行OCR识别处理");
+        } else {
+            logger.info("PDF文件上传后将跳过OCR识别处理（可在配置文件中设置 umi-ocr.enabled=true 启用）");
+        }
+    }
+
+    @PreDestroy
+    public void shutdown() {
+        if (directoryExecutor != null) {
+            directoryExecutor.shutdown();
+        }
+    }
+
+    /**
+     * 检测7zip工具的可用性
+     */
+    private void checkSevenZipAvailability() {
+        logger.info("开始检测7-Zip工具可用性...");
+
+        String[] sevenZipCommands = {
+            "7z", "7za", "7zr",
+            "C:\\Program Files\\7-Zip\\7z.exe",
+            "C:\\Program Files (x86)\\7-Zip\\7z.exe",
+            "/usr/bin/7z", "/usr/local/bin/7z",
+            "/opt/7zip/7z", "/usr/lib/7zip/7z"
+        };
+
+        for (String cmd : sevenZipCommands) {
+            try {
+                Process process = new ProcessBuilder(cmd, "--help").start();
+                int exitCode = process.waitFor();
+
+                if (exitCode == 0) {
+                    // 验证这是真的7zip
+                    try (java.io.BufferedReader reader = new java.io.BufferedReader(
+                            new java.io.InputStreamReader(process.getInputStream()))) {
+                        String firstLine = reader.readLine();
+                        if (firstLine != null && firstLine.toLowerCase().contains("7-zip")) {
+                            sevenZipAvailable = true;
+                            sevenZipCommand = cmd;
+                            logger.info("✅ 7-Zip工具可用: {}", cmd);
+                            logger.info("   版本信息: {}", firstLine.trim());
+                            return;
+                        }
+                    }
+
+                    // 如果读取失败但退出码为0，也认为是可用的
+                    sevenZipAvailable = true;
+                    sevenZipCommand = cmd;
+                    logger.info("✅ 7-Zip工具可用: {}", cmd);
+                    return;
+                }
+            } catch (Exception e) {
+                // 继续尝试下一个命令
+            }
+        }
+
+        sevenZipAvailable = false;
+        sevenZipCommand = null;
+        logger.warn("❌ 未检测到7-Zip工具");
+        logger.warn("   这将影响分卷压缩文件的解压功能");
+        logger.warn("   请安装7-Zip并确保其在系统PATH中，或将其添加到上述检测路径之一");
+        logger.warn("   下载地址: https://www.7-zip.org/");
+    }
+
+    @Value("${file.allowed-types:pdf,doc,docx,txt}")
+    private String allowedTypes;
+
+    @Value("${file.filter.technical-only:true}")
+    private boolean filterTechnicalOnly;
+
+    // 支持的文件扩展名列表
+
+    @Override
+    public FileService.UploadResult uploadAndExtractZip(String projectId, MultipartFile zipFile) throws IOException {
+        // 1. 创建或更新项目记录
+        createOrUpdateProject(projectId, null);
+
+        // 2. 生成文件夹名称：项目id_时间_uuid
+        String timestamp = LocalDateTime.now().format(DATE_TIME_FORMATTER);
+        String uuid = UUID.randomUUID().toString().replace("-", "").substring(0, 8);
+        String folderName = projectId + "_" + timestamp + "_" + uuid;
+        
+        // 3. 创建上传目录（使用新的文件夹名称）
+        Path projectUploadPath = Paths.get(absoluteUploadPath, folderName);
+        Files.createDirectories(projectUploadPath);
+        logger.info("上传目录: {}", projectUploadPath);
+
+        // 4. 创建解压目录（使用新的文件夹名称）
+        Path projectExtractPath = Paths.get(absoluteExtractPath, folderName);
+        Files.createDirectories(projectExtractPath);
+        logger.info("解压目录: {}", projectExtractPath);
+
+        // 5. 保存ZIP文件
+        String originalFilename = zipFile.getOriginalFilename();
+        if (originalFilename == null || !originalFilename.toLowerCase().endsWith(".zip")) {
+            throw new IllegalArgumentException("文件必须是ZIP格式");
+        }
+
+        Path zipFilePath = projectUploadPath.resolve(originalFilename);
+        zipFile.transferTo(zipFilePath.toFile());
+        String zipFilePathStr = zipFilePath.toAbsolutePath().toString();
+        logger.info("ZIP文件已保存到: {}", zipFilePathStr);
+
+        // 6. 解压ZIP文件到指定的解压目录
+        logger.info("开始解压ZIP文件: {}", zipFilePathStr);
+        logger.info("解压目标目录: {}", projectExtractPath.toAbsolutePath());
+        List<String> extractedFilePaths;
+        try {
+            extractedFilePaths = extractZipFile(zipFilePath, projectExtractPath);
+        } catch (Exception e) {
+            logger.error("解压ZIP文件时发生异常: {}", e.getMessage(), e);
+            // 不要抛出异常，而是返回空列表，让用户知道解压失败
+            logger.error("ZIP文件解压失败，返回空文件列表。错误详情: {}", e.getMessage());
+            extractedFilePaths = new ArrayList<>();
+        }
+        String extractPathStr = projectExtractPath.toAbsolutePath().toString();
+
+        // 最终过滤：只过滤掉以点开头的临时文件（如 .DS_Store, .__MACOSX 等）
+        List<String> finalExtractedFilePaths = new ArrayList<>();
+        for (String filePath : extractedFilePaths) {
+            Path path = Paths.get(filePath);
+            String fileName = path.getFileName().toString();
+            
+            // 只过滤以点开头的文件（通常是隐藏文件或临时文件）
+            if (!fileName.startsWith(".")) {
+                finalExtractedFilePaths.add(filePath);
+            } else {
+                logger.warn("过滤掉临时/隐藏文件: {}", filePath);
+            }
+        }
+
+        logger.info("ZIP文件解压完成，共解压 {} 个文件，过滤后返回 {} 个内容文件到目录: {}",
+                extractedFilePaths.size(), finalExtractedFilePaths.size(), extractPathStr);
+        logger.info("最终解压文件列表: {}", finalExtractedFilePaths);
+
+        // 如果最终文件列表为空，记录更多调试信息
+        if (finalExtractedFilePaths.isEmpty()) {
+            logger.error("警告：没有解压出任何文件！");
+            logger.error("原始解压文件数量: {}", extractedFilePaths.size());
+            logger.error("ZIP文件路径: {}", zipFilePathStr);
+            logger.error("解压目录: {}", extractPathStr);
+
+            try {
+                if (Files.exists(zipFilePath)) {
+                    logger.error("ZIP文件存在，大小: {} 字节", Files.size(zipFilePath));
+                } else {
+                    logger.error("ZIP文件不存在！");
+                }
+
+                if (Files.exists(projectExtractPath)) {
+                    logger.error("解压目录存在");
+                    try (var paths = Files.walk(projectExtractPath)) {
+                        List<Path> allFiles = paths.filter(Files::isRegularFile).collect(java.util.stream.Collectors.toList());
+                        logger.error("解压目录中的所有文件 ({} 个):", allFiles.size());
+                        for (Path file : allFiles) {
+                            String ext = getFileExtension(file.getFileName().toString());
+                            logger.error("  - {} (扩展名: {})", file.getFileName(), ext);
+                        }
+                    }
+                } else {
+                    logger.error("解压目录不存在");
+                }
+            } catch (Exception e) {
+                logger.error("收集调试信息时出错: {}", e.getMessage());
+            }
+        }
+
+        if (finalExtractedFilePaths.isEmpty()) {
+            logger.warn("警告：ZIP文件中没有解压出任何文件，可能是ZIP文件为空");
+            // 检查ZIP文件内容
+            try {
+                long zipSize = Files.size(zipFilePath);
+                logger.warn("ZIP文件大小: {} 字节", zipSize);
+                if (zipSize == 0) {
+                    logger.error("ZIP文件为空文件！");
+                } else {
+                    logger.warn("ZIP文件不为空，但没有解压出任何文件，尝试列出ZIP内容...");
+                    listZipContents(zipFilePath);
+                }
+            } catch (Exception e) {
+                logger.error("检查ZIP文件大小时出错: {}", e.getMessage());
+            }
+        } else {
+            // 按文件扩展名统计
+            java.util.Map<String, Integer> extensionCount = new java.util.HashMap<>();
+            int zeroByteFileCount = 0;
+            for (String path : finalExtractedFilePaths) {
+                String ext = getFileExtension(path);
+                extensionCount.put(ext.isEmpty() ? "无扩展名" : ext, extensionCount.getOrDefault(ext, 0) + 1);
+                
+                // 统计0KB文件
+                try {
+                    Path filePath = Paths.get(path);
+                    if (Files.exists(filePath) && Files.size(filePath) == 0) {
+                        zeroByteFileCount++;
+                    }
+                } catch (Exception e) {
+                    // 忽略统计错误
+                }
+            }
+            logger.info("解压文件统计（按扩展名）: {}", extensionCount);
+            if (zeroByteFileCount > 0) {
+                logger.info("0KB文件统计: {} 个文件", zeroByteFileCount);
+            }
+            
+            // 统计包含"商务文件"、"技术文件"、"报价文件"的路径
+            long businessFileCount = extractedFilePaths.stream()
+                    .filter(p -> p.contains("商务文件")).count();
+            long technicalFileCount = extractedFilePaths.stream()
+                    .filter(p -> p.contains("技术文件")).count();
+            long quotationFileCount = extractedFilePaths.stream()
+                    .filter(p -> p.contains("报价文件")).count();
+            logger.info("文件路径统计 - 商务文件: {}, 技术文件: {}, 报价文件: {}", 
+                    businessFileCount, technicalFileCount, quotationFileCount);
+            
+            // 统计"广东迅力供应链管理有限公司"的文件
+            long xunliFileCount = extractedFilePaths.stream()
+                    .filter(p -> p.contains("广东迅力供应链管理有限公司")).count();
+            logger.info("广东迅力供应链管理有限公司的文件数: {}", xunliFileCount);
+            if (xunliFileCount > 0) {
+                List<String> xunliFiles = extractedFilePaths.stream()
+                        .filter(p -> p.contains("广东迅力供应链管理有限公司"))
+                        .collect(java.util.stream.Collectors.toList());
+                logger.info("广东迅力供应链管理有限公司的文件列表（前10个）: {}", 
+                        xunliFiles.stream().limit(10)
+                            .map(p -> java.nio.file.Paths.get(p).getFileName().toString())
+                            .collect(java.util.stream.Collectors.joining(", ")));
+            }
+        }
+        
+        // 7. 异步处理所有后续操作（数据库保存、向量化、目录抽取）
+        // 同步执行写库部分，确保接口返回前即可查询到数据
+        final List<String> extractedPaths = new ArrayList<>(extractedFilePaths);
+        final List<String> finalPaths = new ArrayList<>(finalExtractedFilePaths);
+
+        updateProjectFilePaths(projectId, zipFilePathStr, extractedPaths);
+        saveFilesToDatabase(projectId, extractedPaths, null, null);
+
+        CompletableFuture.runAsync(() -> {
+            try {
+                asyncProcessUploadedFiles(projectId, zipFilePathStr,
+                        new ArrayList<>(extractedPaths), new ArrayList<>(finalPaths), null, null);
+            } catch (Exception e) {
+                logger.error("异步处理上传文件异常", e);
+            }
+        });
+
+        return new FileService.UploadResult(zipFilePathStr, extractPathStr, finalExtractedFilePaths);
+    }
+
+    /**
+     * 处理上传后的所有操作（数据库保存、向量化、目录抽取）
+     */
+    public void asyncProcessUploadedFiles(String projectId, String zipFilePathStr,
+                                          List<String> extractedFilePaths,
+                                          List<String> finalExtractedFilePaths,
+                                          String supplierId,
+                                          Long explicitFileTypeId) {
+        try {
+            logger.info("开始异步处理上传文件，项目ID: {}, 供应商ID: {}, 文件类型ID: {}, 文件数量: {}", 
+                    projectId, supplierId, explicitFileTypeId, extractedFilePaths.size());
+        
+            // 1. 保存文件信息到数据库
+            saveFilesToDatabase(projectId, extractedFilePaths, supplierId, explicitFileTypeId);
+
+            // 2. 更新项目表中的文件路径
+            updateProjectFilePaths(projectId, zipFilePathStr, extractedFilePaths);
+
+            // 3. 筛选需要向量化的文件并异步读取内容
+        List<String> vectorizableFilePaths = filterVectorizableFiles(extractedFilePaths);
+        if (!vectorizableFilePaths.isEmpty()) {
+            logger.info("发现 {} 个需要向量化的文件，将异步读取内容并进行向量化", vectorizableFilePaths.size());
+
+            // 分别处理PDF和其他类型文件
+            List<String> pdfFilePaths = filterPdfFiles(vectorizableFilePaths);
+            if (!pdfFilePaths.isEmpty()) {
+                logger.info("发现 {} 个PDF文件，将异步读取内容并进行向量化", pdfFilePaths.size());
+                pdfReadService.readPdfFilesAsync(pdfFilePaths, projectId);
+
+                // 同时进行OCR识别处理
+                if (ocrEnabled) {
+                    logger.info("开始异步处理PDF文件的OCR识别");
+                } else {
+                    logger.info("OCR功能已关闭，跳过PDF文件的OCR识别");
+                }
+                processPdfOcrForFiles(pdfFilePaths, projectId, supplierId);
+            }
+
+            // 这里可以扩展为调用相应的服务来处理其他文件类型
+        } else {
+            logger.info("未发现需要向量化的文件");
+        }
+        
+            // 4. 异步处理文件目录抽取
+            if (!finalExtractedFilePaths.isEmpty()) {
+                logger.info("开始异步处理文件目录抽取，共 {} 个文件", finalExtractedFilePaths.size());
+                processFileDirectoriesAsync(finalExtractedFilePaths, projectId);
+            } else {
+                logger.info("没有文件需要进行目录抽取");
+            }
+
+            logger.info("异步处理上传文件完成，项目ID: {}", projectId);
+
+        } catch (Exception e) {
+            logger.error("异步处理上传文件失败，项目ID: {}", projectId, e);
+        }
+    }
+
+    /**
+     * 保存文件信息到数据库并更新项目文件路径（保留兼容性）
+     */
+    public void asyncSaveFilesAndUpdateProject(String projectId, String zipFilePathStr, List<String> extractedFilePaths) {
+        asyncProcessUploadedFiles(projectId, zipFilePathStr, extractedFilePaths, extractedFilePaths, null, null);
+    }
+
+    /**
+     * 从文件路径列表中筛选出PDF文件
+     *
+     * @param filePaths 文件路径列表
+     * @return PDF文件路径列表
+     */
+    private List<String> filterPdfFiles(List<String> filePaths) {
+        List<String> pdfFiles = new ArrayList<>();
+        for (String filePath : filePaths) {
+            if (filePath != null && filePath.toLowerCase().endsWith(".pdf")) {
+                pdfFiles.add(filePath);
+            }
+        }
+        return pdfFiles;
+    }
+
+    /**
+     * 处理PDF文件的OCR识别
+     *
+     * @param pdfFilePaths PDF文件路径列表
+     * @param projectId 项目ID
+     * @param supplierId 供应商ID
+     */
+    private void processPdfOcrForFiles(List<String> pdfFilePaths, String projectId, String supplierId) {
+        // 检查OCR功能是否启用
+        if (!ocrEnabled) {
+            logger.info("OCR功能已关闭，跳过PDF文件OCR处理: projectId={}, supplierId={}", projectId, supplierId);
+            return;
+        }
+
+        logger.info("准备处理 {} 个PDF文件的OCR: projectId={}, supplierId={}", pdfFilePaths.size(), projectId, supplierId);
+
+        for (String filePath : pdfFilePaths) {
+            try {
+                logger.debug("查找PDF文件的数据库记录: {}", filePath);
+
+                // 首先尝试通过存储的路径查找（normalized路径）
+                Path absolutePath = Paths.get(filePath).toAbsolutePath().normalize();
+                String storedPath = filePathResolver.normalizeForStorage(absolutePath);
+
+                java.util.Optional<File> fileOpt = fileRepository.findByFilePath(storedPath);
+                if (fileOpt.isPresent()) {
+                    File file = fileOpt.get();
+                    logger.info("✅ 通过存储路径找到PDF文件记录，开始OCR处理: {} (ID: {}, 扩展名: {})",
+                               storedPath, file.getId(), file.getFileExtension());
+                    ocrService.processPdfOcrAsync(file.getId(), filePath, projectId, supplierId);
+                } else {
+                    logger.warn("❌ 通过存储路径未找到PDF文件记录: {}", storedPath);
+
+                    // 尝试通过原始绝对路径查找
+                    fileOpt = fileRepository.findByFilePath(absolutePath.toString());
+                    if (fileOpt.isPresent()) {
+                        File file = fileOpt.get();
+                        logger.info("✅ 通过绝对路径找到PDF文件记录，开始OCR处理: {} (ID: {})", absolutePath, file.getId());
+                        ocrService.processPdfOcrAsync(file.getId(), filePath, projectId, supplierId);
+                    } else {
+                        logger.warn("❌ 通过绝对路径也未找到PDF文件记录: {}", absolutePath);
+
+                        // 最后尝试通过文件名查找（作为备用方案）
+                        String fileName = java.nio.file.Paths.get(filePath).getFileName().toString();
+                        logger.debug("尝试通过文件名查找: {}", fileName);
+                        java.util.List<File> filesByName = fileRepository.findByProjectIdAndFileName(projectId, fileName);
+                        if (!filesByName.isEmpty()) {
+                            File file = filesByName.get(0);
+                            logger.info("✅ 通过文件名找到PDF文件记录，开始OCR处理: {} (ID: {})", filePath, file.getId());
+                            ocrService.processPdfOcrAsync(file.getId(), filePath, projectId, supplierId);
+                        } else {
+                            logger.error("❌ 通过所有方式都未找到PDF文件记录: {}", fileName);
+                        }
+                    }
+                }
+            } catch (Exception e) {
+                logger.error("❌ 处理PDF文件OCR时出错: {}", filePath, e);
+            }
+        }
+    }
+
+    /**
+     * 过滤需要向量化的文件类型（PDF、Word、PPT、TXT、JSON）
+     * 只返回真正的内容文件，不包含压缩文件
+     *
+     * @param filePaths 文件路径列表
+     * @return 需要向量化的文件路径列表
+     */
+    private List<String> filterVectorizableFiles(List<String> filePaths) {
+        List<String> vectorizableFiles = new ArrayList<>();
+
+        // 只支持指定的文件类型：pdf, word, ppt, txt, json
+        Set<String> supportedExtensions = new HashSet<>(Arrays.asList(
+            ".pdf",    // PDF文件
+            ".doc", ".docx",  // Word文件
+            ".ppt", ".pptx",  // PowerPoint文件
+            ".txt",    // 文本文件
+            ".json"    // JSON文件
+        ));
+
+        for (String filePath : filePaths) {
+            if (filePath != null && !filePath.trim().isEmpty()) {
+                String lowerCasePath = filePath.toLowerCase();
+                for (String ext : supportedExtensions) {
+                    if (lowerCasePath.endsWith(ext)) {
+                        vectorizableFiles.add(filePath);
+                        break; // 找到匹配的扩展名就跳出内循环
+                    }
+                }
+            }
+        }
+
+        logger.info("过滤得到 {} 个需要向量化的文件 (pdf, word, ppt, txt, json)", vectorizableFiles.size());
+        return vectorizableFiles;
+    }
+    
+    /**
+     * 保存文件信息到文件表
+     */
+    @Transactional
+    private void saveFilesToDatabase(String projectId, List<String> filePaths, String explicitSupplierId, Long explicitFileTypeId) {
+        try {
+            for (String filePath : filePaths) {
+                if (filePath == null || filePath.trim().isEmpty()) {
+                    continue;
+                }
+
+                try {
+                    Path path = Paths.get(filePath);
+                    if (!Files.exists(path)) {
+                        logger.warn("文件不存在，跳过保存: {}", filePath);
+                        continue;
+                    }
+
+                    Path absolutePath = path.toAbsolutePath().normalize();
+                    path = absolutePath;
+                    String storedPath = filePathResolver.normalizeForStorage(absolutePath);
+
+                    java.util.Optional<File> existingFile = fileRepository.findByFilePath(storedPath);
+                    if (existingFile.isPresent()) {
+                        logger.debug("文件已存在于数据库，跳过: {}", storedPath);
+                        continue;
+                    }
+
+                    java.util.Optional<File> legacyFile = fileRepository.findByFilePath(absolutePath.toString());
+                    if (legacyFile.isPresent()) {
+                        File legacy = legacyFile.get();
+                        legacy.setFilePath(storedPath);
+                        fileRepository.save(legacy);
+                        logger.debug("更新已有文件记录路径: {}", storedPath);
+                        continue;
+                    }
+
+                    // 获取文件信息
+                    String fileName = path.getFileName().toString();
+                    String fileExtension = getFileExtension(fileName);
+                    long fileSize = Files.size(path);
+                    
+                    // 0KB文件也应该被保存，记录日志以便调试
+                    if (fileSize == 0) {
+                        logger.debug("检测到0KB文件: {} (路径: {})", fileName, filePath);
+                    }
+
+                    // 从路径识别文件类型（技术文件、商务文件、报价文件等）
+                    Long fileTypeId = explicitFileTypeId;
+                    if (fileTypeId == null) {
+                        fileTypeId = extractFileTypeIdFromPath(filePath);
+                    }
+                    if (fileTypeId == null) {
+                        // 如果无法从路径识别，尝试通过文件名识别
+                        FileType fileType = fileTypeService.identifyFileType(fileName, filePath);
+                        if (fileType != null) {
+                            fileTypeId = fileType.getId();
+                            logger.debug("通过文件名识别文件类型: {} -> {}", fileName, fileType.getTypeName());
+                        } else {
+                            // 如果无法识别，创建一个"未知类型"
+                            FileType unknownType = fileTypeService.createOrGetFileType("未知类型", "无法自动识别的文件类型");
+                            fileTypeId = unknownType.getId();
+                            logger.debug("使用未知类型: {}", fileName);
+                        }
+                    }
+
+                    // 从路径提取供应商信息
+                    String supplierId = explicitSupplierId;
+                    if (supplierId == null) {
+                        supplierId = extractSupplierIdFromPath(filePath);
+                    } else {
+                        // 如果明确指定了supplierId，确保它存在于数据库中
+                        try {
+                            if (supplierRepository.findBySupplierId(supplierId) == null) {
+                                Supplier newSupplier = new Supplier();
+                                newSupplier.setSupplierId(supplierId);
+                                newSupplier.setSupplierName(supplierId); // 暂无名称
+                                newSupplier.setStatus("active");
+                                supplierRepository.save(newSupplier);
+                            }
+                        } catch (Exception e) {
+                            logger.error("处理供应商信息时出错: {}", e.getMessage(), e);
+                        }
+                    }
+
+                    // 创建并保存File实体
+                    File file = new File();
+                    file.setFilePath(storedPath);
+                    file.setFileName(fileName);
+                    file.setFileExtension(fileExtension);
+                    file.setFileSize(fileSize);
+                    file.setProjectId(projectId);
+                    file.setFileTypeId(fileTypeId);
+                    file.setSupplierId(supplierId);
+                    file.setStatus("uploaded");
+                    
+                    // 计算MD5
+                    try {
+                        String md5 = calculateMD5(path);
+                        file.setFileMd5(md5);
+                    } catch (Exception e) {
+                        logger.warn("计算文件MD5失败: {}", e.getMessage());
+                    }
+                    
+                    // 设置目录抽取初始状态为processing，以便异步处理
+                    file.setCatalogExtractionStatus("processing");
+                    file.setCatalogExtractionRetryCount(0);
+                    
+                    fileRepository.save(file);
+                    logger.info("文件已保存到数据库: fileId={}, fileName={}, catalogStatus=processing", 
+                            file.getFileId(), fileName);
+                } catch (Exception e) {
+                    logger.error("保存文件信息到数据库时出错: {}", filePath, e);
+                }
+            }
+        } catch (Exception e) {
+            logger.error("批量保存文件信息到数据库失败: {}", e.getMessage(), e);
+        }
+    }
+                      
+    /**
+     * 计算文件MD5值
+     */
+    private String calculateMD5(Path filePath) throws IOException {
+        try {
+            long fileSize = Files.size(filePath);
+            
+            // 对于0KB文件，返回空字符串的MD5值
+            if (fileSize == 0) {
+                java.security.MessageDigest md = java.security.MessageDigest.getInstance("MD5");
+                byte[] digest = md.digest();
+                StringBuilder sb = new StringBuilder();
+                for (byte b : digest) {
+                    sb.append(String.format("%02x", b));
+                }
+                return sb.toString();
+            }
+            
+            // 对于非空文件，正常计算MD5
+            try (java.io.InputStream is = Files.newInputStream(filePath)) {
+                java.security.MessageDigest md = java.security.MessageDigest.getInstance("MD5");
+                byte[] buffer = new byte[8192];
+                int read;
+                while ((read = is.read(buffer)) > 0) {
+                    md.update(buffer, 0, read);
+                }
+                byte[] digest = md.digest();
+                StringBuilder sb = new StringBuilder();
+                for (byte b : digest) {
+                    sb.append(String.format("%02x", b));
+                }
+                return sb.toString();
+            }
+        } catch (java.security.NoSuchAlgorithmException e) {
+            throw new IOException("MD5算法不可用", e);
+        }
+    }
+
+    /**
+     * 更新项目表中的文件路径
+     */
+    @Transactional
+    private void updateProjectFilePaths(String projectId, String zipFilePath, List<String> extractedFilePaths) {
+        try {
+            Project project = projectRepository.findByProjectId(projectId);
+            if (project != null) {
+                // 构建文件路径列表
+                // 1. 上传目录中的原始ZIP文件路径（不删除，保留）
+                List<String> allFilePaths = new ArrayList<>();
+                allFilePaths.add(zipFilePath);
+                
+                // 2. 添加所有解压后的文件路径（不包括解压目录中已删除的ZIP文件）
+                // 注意：extractedFilePaths 中已经包含了所有成功解压的文件（不包括已删除的ZIP）
+                allFilePaths.addAll(extractedFilePaths);
+                
+                // 去重（避免重复路径）
+                List<String> uniqueFilePaths = new ArrayList<>(new java.util.LinkedHashSet<>(allFilePaths));
+                
+                // 保存为JSON格式
+                String filePathsJson = JSON.toJSONString(uniqueFilePaths);
+                project.setFilePaths(filePathsJson);
+                project.setUpdateTime(LocalDateTime.now());
+                projectRepository.save(project);
+                logger.info("项目 {} 的文件路径已更新，共 {} 个文件（原始ZIP: 1个，解压文件: {}个）", 
+                        projectId, uniqueFilePaths.size(), extractedFilePaths.size());
+                logger.debug("文件路径JSON: {}", filePathsJson);
+            } else {
+                logger.warn("项目不存在: {}", projectId);
+            }
+        } catch (Exception e) {
+            logger.error("更新项目文件路径失败: {}", e.getMessage(), e);
+            // 不抛出异常，避免影响主流程
+        }
+    }
+    
+    @Override
+    @Transactional
+    public Project createOrUpdateProject(String projectId, String projectName) {
+        Project project = projectRepository.findByProjectId(projectId);
+        
+        if (project == null) {
+            // 创建新项目
+            project = new Project();
+            project.setProjectId(projectId);
+            project.setProjectName(projectName != null ? projectName : projectId);
+            project.setStatus("ACTIVE");
+            project.setCreateTime(LocalDateTime.now());
+            project.setUpdateTime(LocalDateTime.now());
+            project = projectRepository.save(project);
+            logger.info("创建新项目: {}", projectId);
+        } else {
+            // 更新项目
+            if (projectName != null && !projectName.isEmpty()) {
+                project.setProjectName(projectName);
+            }
+            project.setUpdateTime(LocalDateTime.now());
+            project = projectRepository.save(project);
+            logger.info("更新项目: {}", projectId);
+        }
+        
+        return project;
+    }
+
+    /**
+     * 解压ZIP文件（带递归深度限制）
+     *
+     * @param zipFilePath ZIP文件路径
+     * @param extractPath 解压目标路径
+     * @return 解压后的文件路径列表
+     */
+    private List<String> extractZipFile(Path zipFilePath, Path extractPath) throws IOException {
+        return extractZipFile(zipFilePath, extractPath, 0);
+    }
+    
+    /**
+     * 解压ZIP文件（递归版本，支持多层嵌套解压）
+     *
+     * @param zipFilePath ZIP文件路径
+     * @param extractPath 解压目标路径
+     * @param depth 当前递归深度（用于防止无限递归，最大深度限制为10）
+     * @return 解压后的文件路径列表
+     */
+    private List<String> extractZipFile(Path zipFilePath, Path extractPath, int depth) throws IOException {
+        logger.info("=== 开始调用extractZipFile方法 ===");
+        logger.info("ZIP文件路径: {}", zipFilePath);
+        logger.info("解压目标路径: {}", extractPath);
+        logger.info("递归深度: {}", depth);
+
+        // 1. 检查文件是否真的是ZIP文件（通过Magic Bytes），防止HTML文件冒充
+        try (InputStream is = Files.newInputStream(zipFilePath)) {
+            byte[] buffer = new byte[4];
+            int read = is.read(buffer);
+            if (read >= 1 && buffer[0] == '<') {
+                // 很可能是HTML文件
+                 String preview = new String(buffer, 0, read);
+                // 再读一点内容以便确认
+                byte[] moreBuffer = new byte[100];
+                int moreRead = is.read(moreBuffer);
+                if (moreRead > 0) {
+                    preview += new String(moreBuffer, 0, moreRead);
+                }
+                String errorMsg = "下载的文件是HTML页面而不是ZIP文件 (预览: " + preview.replace("\n", " ").replace("\r", " ") + "...)";
+                logger.error(errorMsg);
+                throw new IOException(errorMsg);
+            }
+        } catch (IOException e) {
+             if (e.getMessage() != null && e.getMessage().contains("HTML页面")) {
+                throw e;
+            }
+            logger.warn("检查文件头失败: {}", e.getMessage());
+        }
+
+        // 防止无限递归，最大深度限制为10层
+        if (depth >= 10) {
+            logger.warn("ZIP文件嵌套深度超过限制（10层），停止递归解压: {}", zipFilePath);
+            return new ArrayList<>();
+        }
+        
+        List<String> extractedFiles = new ArrayList<>();
+        int totalEntries = 0;
+        int skippedEntries = 0;
+        // 记录被跳过的文件信息（用于调试）
+        List<String> skippedFiles = new ArrayList<>();
+        List<String> skippedReasons = new ArrayList<>();
+
+        logger.info("开始读取ZIP文件 [深度: {}]: {}", depth, zipFilePath);
+        
+        // 检查ZIP文件是否存在
+        if (!Files.exists(zipFilePath)) {
+            throw new IOException("ZIP文件不存在: " + zipFilePath);
+        }
+        
+        // 检查ZIP文件大小
+        long zipFileSize = Files.size(zipFilePath);
+        logger.info("ZIP文件大小: {} 字节", zipFileSize);
+        
+        if (zipFileSize == 0) {
+            logger.warn("ZIP文件为空");
+            return extractedFiles;
+        }
+        
+        // 快速检测明显的非ZIP文件（例如上游传了JSON/TXT）
+        String lowerName = zipFilePath.getFileName().toString().toLowerCase(Locale.ROOT);
+        try (InputStream detectIs = Files.newInputStream(zipFilePath)) {
+            byte[] head = new byte[4];
+            int read = detectIs.read(head);
+            boolean looksJson = read > 0 && (head[0] == '{' || head[0] == '[');
+            if (looksJson || lowerName.endsWith(".json") || lowerName.endsWith(".txt")) {
+                String magicBytes = read > 0
+                        ? String.format("%02X %02X %02X %02X", head[0], head.length > 1 ? head[1] : 0,
+                                head.length > 2 ? head[2] : 0, head.length > 3 ? head[3] : 0)
+                        : "unknown";
+                logger.info("检测到非ZIP文件，跳过解压: {} (Magic Bytes: {})", zipFilePath, magicBytes);
+                return extractedFiles;
+            }
+        } catch (Exception e) {
+            logger.warn("检测文件类型失败，继续尝试作为ZIP处理: {}", e.getMessage());
+        }
+        
+        // 先尝试使用ZipFile（更高效），如果失败则回退到ZipInputStream（更宽容）
+        boolean useZipInputStream = false;
+        
+        // 尝试1: 使用ZipFile（支持UTF-8编码）
+        ZipFile zipFile = null;
+        try {
+            // 尝试使用UTF-8编码打开ZIP文件
+            try {
+                zipFile = new ZipFile(zipFilePath.toFile(), Charset.forName("UTF-8"));
+            } catch (Exception e) {
+                // 如果UTF-8失败，尝试默认编码
+                logger.debug("使用UTF-8编码打开ZIP文件失败，尝试默认编码: {}", e.getMessage());
+                zipFile = new ZipFile(zipFilePath.toFile());
+            }
+            
+            logger.info("使用ZipFile方式读取ZIP文件条目...");
+            Enumeration<? extends ZipEntry> entries = zipFile.entries();
+            
+            while (entries.hasMoreElements()) {
+                ZipEntry entry = entries.nextElement();
+                totalEntries++;
+                
+                String entryName = entry.getName();
+                long entrySize = entry.getSize();
+                boolean isDirectory = entry.isDirectory() || entryName.endsWith("/");
+                
+                // 记录条目信息（包括预期大小，用于后续验证）
+                logger.info("处理ZIP条目 [{}]: {}, 类型: {}, ZIP中记录的大小: {} 字节", 
+                        totalEntries, entryName, isDirectory ? "目录" : "文件", 
+                        entrySize != -1 ? entrySize : "未知");
+
+                // 跳过目录（目录不算文件，所以不记录到skippedFiles，只计数）
+                if (isDirectory) {
+                    logger.debug("跳过目录: {}", entryName);
+                    skippedEntries++;
+                    continue;
+                }
+
+                // 规范化路径（统一使用正斜杠）
+                String normalizedEntryName = entryName.replace('\\', '/');
+
+                // 移除路径过滤逻辑，让所有文件都能被解压用于向量化
+                
+                // 检查文件扩展名
+                String extension = getFileExtension(entryName);
+                
+                // 不再检查文件类型，允许所有文件类型解压
+                // 文件类型过滤将在返回结果时进行
+                
+                // 处理嵌套的ZIP文件（分卷文件在ZIP解压完成后统一处理）
+                if ("zip".equalsIgnoreCase(extension)) {
+                    // 先解压嵌套的ZIP文件到临时位置
+                    Path nestedZipPath = extractPath.resolve(normalizedEntryName).normalize();
+                    Path normalizedExtractPath = extractPath.normalize();
+                    
+                    if (!nestedZipPath.startsWith(normalizedExtractPath)) {
+                        logger.warn("检测到路径遍历攻击，跳过嵌套ZIP文件: {} (目标路径: {})", entryName, nestedZipPath);
+                        skippedEntries++;
+                        skippedFiles.add(entryName);
+                        skippedReasons.add("路径遍历攻击：嵌套ZIP文件");
+                        continue;
+                    }
+                    
+                    // 创建父目录
+                    Path parentDir = nestedZipPath.getParent();
+                    if (parentDir != null) {
+                        Files.createDirectories(parentDir);
+                    }
+                    
+                    // 保存嵌套的ZIP文件
+                    try (InputStream entryInputStream = zipFile.getInputStream(entry);
+                         BufferedInputStream bis = new BufferedInputStream(entryInputStream);
+                         BufferedOutputStream bos = new BufferedOutputStream(
+                                 Files.newOutputStream(nestedZipPath))) {
+                        
+                        byte[] buffer = new byte[8192];
+                        int len;
+                        long totalBytes = 0;
+                        while ((len = bis.read(buffer)) > 0) {
+                            bos.write(buffer, 0, len);
+                            totalBytes += len;
+                        }
+                        // 确保文件被正确创建（即使是0KB文件）
+                        bos.flush();
+                        logger.info("嵌套ZIP文件已保存: {} (大小: {} 字节)", nestedZipPath, totalBytes);
+                    } catch (IOException e) {
+                        logger.error("保存嵌套ZIP文件失败: {}, 错误: {}", entryName, e.getMessage(), e);
+                        skippedEntries++;
+                        continue;
+                    }
+                    
+                    // 递归解压嵌套的ZIP文件
+                    try {
+                        // 创建嵌套ZIP的解压目录（使用文件名去掉.zip后缀）
+                        String nestedZipName = nestedZipPath.getFileName().toString();
+                        String nestedDirName = nestedZipName.substring(0, nestedZipName.lastIndexOf('.'));
+                        Path nestedExtractPath = nestedZipPath.getParent().resolve(nestedDirName);
+                        
+                        // 递归解压嵌套的ZIP文件
+                        List<String> nestedFiles = extractZipFile(nestedZipPath, nestedExtractPath, depth + 1);
+                        extractedFiles.addAll(nestedFiles);
+                        
+                        logger.info("嵌套ZIP文件已解压 [深度: {}]: {} -> {} (解压 {} 个文件)", 
+                                depth + 1, nestedZipPath, nestedExtractPath, nestedFiles.size());
+                        
+                        // 解压完成后，删除解压目录中的嵌套ZIP文件（保留上传目录中的原始ZIP）
+                        try {
+                            Files.deleteIfExists(nestedZipPath);
+                            logger.info("已删除解压目录中的嵌套ZIP文件: {}", nestedZipPath);
+                        } catch (IOException deleteEx) {
+                            logger.warn("删除嵌套ZIP文件失败: {}, 错误: {}", nestedZipPath, deleteEx.getMessage());
+                        }
+                        
+                    } catch (Exception e) {
+                        logger.error("递归解压嵌套ZIP文件失败 [深度: {}]: {}, 错误: {}", 
+                                depth + 1, entryName, e.getMessage(), e);
+                        skippedEntries++;
+                        skippedFiles.add(entryName);
+                        skippedReasons.add("解压失败: " + (e.getMessage() != null ? e.getMessage() : e.getClass().getSimpleName()));
+                    }
+                    continue;
+                }
+
+                // 分卷文件（.z01, .z02等）现在在ZIP解压完成后统一处理，这里只解压保存
+                
+                // 过滤以点开头的文件（如 .DS_Store, .__MACOSX, ._xxx 等临时/隐藏文件）
+                String fileName = Paths.get(normalizedEntryName).getFileName().toString();
+                if (fileName.startsWith(".")) {
+                    String reason = "跳过临时/隐藏文件: " + fileName;
+                    logger.debug("跳过临时/隐藏文件: {} (原始条目: {})", fileName, entryName);
+                    skippedEntries++;
+                    skippedFiles.add(entryName);
+                    skippedReasons.add(reason);
+                    continue;
+                }
+                
+                // 防止路径遍历攻击
+                Path targetPath = extractPath.resolve(normalizedEntryName).normalize();
+                Path normalizedExtractPath = extractPath.normalize();
+                
+                if (!targetPath.startsWith(normalizedExtractPath)) {
+                    String reason = "路径遍历攻击防护";
+                    logger.warn("检测到路径遍历攻击，跳过文件: {} (目标路径: {})", entryName, targetPath);
+                    skippedEntries++;
+                    skippedFiles.add(entryName);
+                    skippedReasons.add(reason);
+                    continue;
+                }
+
+                // 创建父目录
+                Path parentDir = targetPath.getParent();
+                if (parentDir != null) {
+                    Files.createDirectories(parentDir);
+                    logger.debug("创建父目录: {}", parentDir);
+                }
+
+                // 写入文件
+                long totalBytes = 0;
+                BufferedOutputStream bos = null;
+                try (InputStream entryInputStream = zipFile.getInputStream(entry);
+                     BufferedInputStream bis = new BufferedInputStream(entryInputStream)) {
+                    
+                    // 确保文件目录存在
+                    Files.createDirectories(targetPath.getParent());
+                    
+                    bos = new BufferedOutputStream(Files.newOutputStream(targetPath), 8192);
+                    byte[] buffer = new byte[8192];
+                    int len;
+                    while ((len = bis.read(buffer)) != -1) {  // 使用 != -1 而不是 > 0，确保0KB文件也能正确处理
+                        if (len > 0) {
+                            bos.write(buffer, 0, len);
+                            totalBytes += len;
+                        }
+                    }
+                    // 确保所有数据都被写入磁盘
+                    bos.flush();
+                } catch (IOException e) {
+                    logger.error("解压文件时发生IO错误: {}, 错误: {}", targetPath, e.getMessage(), e);
+                    throw e;
+                } finally {
+                    // 确保输出流被正确关闭
+                    if (bos != null) {
+                        try {
+                            bos.close();
+                        } catch (IOException e) {
+                            logger.warn("关闭输出流失败: {}", e.getMessage());
+                        }
+                    }
+                }
+                
+                // 验证文件已创建并验证实际大小（在流关闭后验证）
+                if (!Files.exists(targetPath)) {
+                    logger.warn("文件解压后不存在，可能创建失败: {}", targetPath);
+                    skippedEntries++;
+                    skippedFiles.add(entryName);
+                    skippedReasons.add("文件创建失败");
+                    continue;
+                }
+                
+                // 验证文件实际大小（在流关闭后验证）
+                long actualFileSize = Files.size(targetPath);
+                
+                // 检查ZIP条目中记录的大小（如果可用）与实际解压大小是否匹配
+                if (entrySize != -1 && entrySize > 0 && totalBytes == 0) {
+                    logger.error("严重错误：ZIP条目记录大小为 {} 字节，但实际读取为 0 字节！文件: {}", 
+                            entrySize, targetPath);
+                    // 这种情况下，文件可能没有正确解压，跳过
+                    skippedEntries++;
+                    skippedFiles.add(entryName);
+                    skippedReasons.add("读取失败：ZIP记录大小不为0但读取为0");
+                    try {
+                        Files.deleteIfExists(targetPath); // 删除可能损坏的空文件
+                    } catch (IOException e) {
+                        logger.warn("删除损坏文件失败: {}", e.getMessage());
+                    }
+                    continue;
+                }
+                
+                if (totalBytes != actualFileSize) {
+                    logger.warn("文件大小不匹配！读取: {} 字节，磁盘: {} 字节，文件: {}", 
+                            totalBytes, actualFileSize, targetPath);
+                }
+                
+                // 记录解压信息（包括ZIP中的预期大小）
+                if (entrySize != -1 && entrySize != totalBytes && entrySize > 0) {
+                    logger.warn("ZIP条目大小与解压大小不一致：ZIP记录 {} 字节，实际解压 {} 字节，文件: {}", 
+                            entrySize, totalBytes, targetPath);
+                }
+                
+                logger.info("文件已解压: {} (ZIP记录: {} 字节, 实际解压: {} 字节, 磁盘: {} 字节)", 
+                        targetPath, entrySize != -1 ? entrySize : "未知", totalBytes, actualFileSize);
+                
+                // 记录解压的文件路径（使用绝对路径）- 包括0KB文件
+                String absolutePath = targetPath.toAbsolutePath().toString();
+                
+                // 如果解压出来的文件还是ZIP文件，继续递归解压
+                // 注意：这里处理的ZIP文件是普通文件解压后发现的，不是ZIP条目中的ZIP
+                // ZIP条目中的ZIP已经在上面处理过了
+                if ("zip".equalsIgnoreCase(extension)) {
+                    try {
+                        // 创建嵌套ZIP的解压目录（使用文件名去掉.zip后缀）
+                        String zipFileName = targetPath.getFileName().toString();
+                        String dirName = zipFileName.substring(0, zipFileName.lastIndexOf('.'));
+                        Path nestedExtractPath = targetPath.getParent().resolve(dirName);
+                        
+                        // 递归解压，depth+1 表示进入下一层
+                        List<String> nestedFiles = extractZipFile(targetPath, nestedExtractPath, depth + 1);
+                        extractedFiles.addAll(nestedFiles);
+                        
+                        logger.info("检测到解压后的ZIP文件，继续递归解压 [深度: {}]: {} -> {} (解压 {} 个文件)", 
+                                depth + 1, targetPath, nestedExtractPath, nestedFiles.size());
+                        
+                        // 解压完成后，删除解压目录中的ZIP文件（保留上传目录中的原始ZIP）
+                        try {
+                            Files.deleteIfExists(targetPath);
+                            logger.info("已删除解压目录中的ZIP文件: {}", targetPath);
+                        } catch (IOException deleteEx) {
+                            logger.warn("删除ZIP文件失败: {}, 错误: {}", targetPath, deleteEx.getMessage());
+                        }
+                        
+                    } catch (Exception e) {
+                        logger.error("递归解压解压后的ZIP文件失败 [深度: {}]: {}, 错误: {}", 
+                                depth + 1, targetPath, e.getMessage(), e);
+                        // 递归解压失败时不添加ZIP文件到返回列表
+                        skippedEntries++;
+                        skippedFiles.add(entryName);
+                        skippedReasons.add("递归解压失败: " + e.getMessage());
+                    }
+                } else {
+                    // 只添加内容文件（PDF, Word, PPT, TXT, JSON）到返回列表
+                    if (isContentFileType(extension)) {
+                    extractedFiles.add(absolutePath);
+                    } else {
+                        logger.debug("跳过非内容文件: {} (扩展名: {})", entryName, extension);
+                    }
+                }
+            }
+            
+        } catch (Exception e) {
+            logger.error("使用ZipFile方式解压失败: {}", e.getMessage(), e);
+            useZipInputStream = true;
+        } finally {
+            if (zipFile != null) {
+                try {
+                    zipFile.close();
+                } catch (IOException closeEx) {
+                    logger.warn("关闭ZipFile失败: {}", closeEx.getMessage());
+                }
+            }
+        }
+        
+        // 如果上面的方法失败，尝试使用ZipInputStream
+        if (useZipInputStream || extractedFiles.isEmpty()) {
+            logger.info("尝试使用ZipInputStream方式解压...");
+            // 这里可以添加ZipInputStream的备用逻辑
+        }
+        
+        logger.info("ZIP文件解压完成 [深度: {}]: {} (共处理 {} 个条目，跳过 {} 个，成功解压 {} 个文件)", 
+                depth, zipFilePath, totalEntries, skippedEntries, extractedFiles.size());
+        
+        return extractedFiles;
+    }
+
+    /**
+     * 从PDF中提取可能的标题
+     */
+    private java.util.List<String> extractPotentialTitlesFromPDF(String filePath) {
+        java.util.List<String> titles = new java.util.ArrayList<>();
+
+        try {
+            // 使用PDFBox正确读取PDF文件的文本内容
+            java.nio.file.Path path = java.nio.file.Paths.get(filePath);
+            if (!java.nio.file.Files.exists(path)) {
+                logger.warn("PDF文件不存在: {}", filePath);
+                return titles;
+            }
+
+            // 使用PDFBox提取PDF文本内容
+            org.apache.pdfbox.pdmodel.PDDocument document = org.apache.pdfbox.Loader.loadPDF(new java.io.File(filePath));
+            org.apache.pdfbox.text.PDFTextStripper stripper = new org.apache.pdfbox.text.PDFTextStripper();
+
+            // 只读取前几页以提高性能，并设置文本提取选项
+            stripper.setStartPage(1);
+            stripper.setEndPage(Math.min(document.getNumberOfPages(), 5)); // 最多读取前5页
+            stripper.setSortByPosition(true); // 按位置排序
+            stripper.setAddMoreFormatting(false); // 简化格式
+
+            String content = stripper.getText(document);
+            document.close();
+
+            // 过滤和清理文本内容
+            content = content.replaceAll("\\n\\s*\\n\\s*\\n", "\n\n"); // 合并多余的空行
+            content = content.replaceAll("[\\r\\n]+", "\n"); // 统一换行符
+
+            // 查找可能的标题模式
+            // 只匹配真正的章节标题，避免匹配表格数据、百分比等
+            java.util.regex.Pattern titlePattern = java.util.regex.Pattern.compile(
+                "(?m)^\\s*(?:第[一二三四五六七八九十\\d]+章|[\\d]+\\.[\\d]+(?:\\.[\\d]+)*)[^\\n]*(?:概述|要求|规范|说明|内容|规定|标准|方法|程序|测试|报告|证书|证明|文件|资料|参数|指标|要求|条件|条款|偏离|承诺|情况|信息|表|图|附录|附件)[^\\n]{0,50}"
+            );
+
+            java.util.regex.Matcher matcher = titlePattern.matcher(content);
+            java.util.Set<String> uniqueTitles = new java.util.HashSet<>();
+
+            while (matcher.find() && uniqueTitles.size() < 20) { // 限制最多20个标题
+                String title = matcher.group().trim();
+                // 清理标题
+                title = title.replaceAll("^[\\s\\*]+", "").replaceAll("[\\s]+$", "");
+
+                // 过滤掉纯数字、百分比或不合理的标题
+                if (title.length() > 3 && title.length() < 100 &&
+                    !uniqueTitles.contains(title) &&
+                    !title.matches("^\\d+(\\.\\d+)*%?$") && // 排除纯数字和百分比
+                    !title.matches("^\\d{1,2}\\.\\d{1,2}%$") && // 排除如"2.60%"这样的百分比
+                    title.matches(".*[一二三四五六七八九十百千万章节条款要求规范].*")) { // 必须包含章节相关关键词
+
+                    uniqueTitles.add(title);
+                    titles.add(title);
+                    logger.debug("提取到有效章节标题: {}", title);
+                    logger.debug("提取到标题: {}", title);
+                }
+            }
+
+            // 如果没有找到标题模式，尝试查找其他可能的标题
+            if (titles.isEmpty()) {
+                // 查找包含关键词的行
+                String[] keywords = {"概述", "介绍", "背景", "目的", "范围", "定义", "要求", "规范", "标准", "技术"};
+                java.util.regex.Pattern keywordPattern = java.util.regex.Pattern.compile(
+                    "(?m)^\\s*([^\\n]{10,60}(?:" + String.join("|", keywords) + ")[^\\n]{0,40})"
+                );
+
+                matcher = keywordPattern.matcher(content);
+                uniqueTitles.clear();
+
+                while (matcher.find() && uniqueTitles.size() < 10) {
+                    String title = matcher.group(1).trim();
+                    if (title.length() > 5 && title.length() < 80 && !uniqueTitles.contains(title)) {
+                        uniqueTitles.add(title);
+                        titles.add(title);
+                        logger.debug("通过关键词提取到标题: {}", title);
+                    }
+                }
+            }
+
+            logger.info("从PDF中提取到 {} 个潜在标题", titles.size());
+
+        } catch (Exception e) {
+            logger.warn("提取PDF标题失败: {}", e.getMessage());
+        }
+
+        return titles;
+    }
+
+    /**
+     * 递归提取PDF目录项
+     */
+    private void extractOutlineItems(PDOutlineItem item, String indent, java.util.List<com.alibaba.fastjson2.JSONObject> chapters, int level) {
+        while (item != null) {
+            try {
+                String title = item.getTitle();
+                if (title != null && !title.trim().isEmpty()) {
+                    com.alibaba.fastjson2.JSONObject chapter = new com.alibaba.fastjson2.JSONObject();
+                    chapter.put("chapterId", "chap-" + chapters.size());
+                    chapter.put("title", title.trim());
+                    chapter.put("level", level);
+
+                    // 尝试获取页码信息
+                    int startPage = -1;
+                    try {
+                        if (item.getDestination() != null) {
+                            // 如果有直接的目的地，尝试获取页码
+                            Object dest = item.getDestination();
+                            if (dest instanceof org.apache.pdfbox.pdmodel.interactive.documentnavigation.destination.PDPageDestination) {
+                                org.apache.pdfbox.pdmodel.interactive.documentnavigation.destination.PDPageDestination pageDest =
+                                    (org.apache.pdfbox.pdmodel.interactive.documentnavigation.destination.PDPageDestination) dest;
+                                startPage = pageDest.retrievePageNumber() + 1; // PDFBox页码从0开始
+                            }
+                        } else if (item.getAction() != null) {
+                            // 如果有动作，尝试从中提取页码
+                            Object action = item.getAction();
+                            if (action instanceof org.apache.pdfbox.pdmodel.interactive.action.PDActionGoTo) {
+                                org.apache.pdfbox.pdmodel.interactive.action.PDActionGoTo goToAction =
+                                    (org.apache.pdfbox.pdmodel.interactive.action.PDActionGoTo) action;
+                                if (goToAction.getDestination() instanceof org.apache.pdfbox.pdmodel.interactive.documentnavigation.destination.PDPageDestination) {
+                                    org.apache.pdfbox.pdmodel.interactive.documentnavigation.destination.PDPageDestination pageDest =
+                                        (org.apache.pdfbox.pdmodel.interactive.documentnavigation.destination.PDPageDestination) goToAction.getDestination();
+                                    startPage = pageDest.retrievePageNumber() + 1;
+                                }
+                            }
+                        }
+                    } catch (Exception e) {
+                        logger.debug("无法获取章节页码信息: {}", title, e);
+                    }
+
+                    // 如果获取到了起始页码
+                    if (startPage != -1) {
+                        chapter.put("page", String.valueOf(startPage));
+                        chapter.put("startPage", startPage);
+                        
+                        // 尝试计算结束页码
+                        // 逻辑：结束页码 = 下一个章节的起始页码 - 1
+                        // 如果没有下一个章节，则暂时设为起始页码，后续统一处理
+                        int endPage = startPage;
+                        
+                        // 尝试获取下一个兄弟节点的页码
+                        PDOutlineItem nextSibling = item.getNextSibling();
+                        if (nextSibling != null) {
+                            int nextPage = getPageNumberFromOutlineItem(nextSibling);
+                            if (nextPage != -1 && nextPage > startPage) {
+                                endPage = nextPage - 1;
+                            }
+                        } 
+                        // 如果没有兄弟节点，但有父节点，尝试获取父节点的下一个兄弟节点的页码
+                        // 这部分逻辑比较复杂，且PDFBox的结构可能不支持直接向上查找
+                        // 因此，我们采取一种简化的策略：
+                        // 在所有章节提取完成后，统一遍历一次列表，修正endPage
+                        
+                        chapter.put("endPage", endPage);
+                    } else {
+                        // 如果没有页码信息，设置默认值
+                        chapter.put("page", "1");
+                        chapter.put("startPage", 1);
+                        chapter.put("endPage", 1);
+                    }
+
+                    chapter.put("children", new com.alibaba.fastjson2.JSONArray());
+
+                    chapters.add(chapter);
+                    logger.debug("{}• {}", indent, title);
+                }
+
+                // 递归处理子项
+                if (item.hasChildren()) {
+                    extractOutlineItems(item.getFirstChild(), indent + "  ", chapters, level + 1);
+                }
+
+            } catch (Exception e) {
+                logger.warn("处理目录项失败: {}", e.getMessage());
+            }
+
+            item = item.getNextSibling();
+        }
+    }
+
+    /**
+     * 辅助方法：从目录项获取页码
+     */
+    private int getPageNumberFromOutlineItem(PDOutlineItem item) {
+        try {
+            if (item.getDestination() != null) {
+                Object dest = item.getDestination();
+                if (dest instanceof org.apache.pdfbox.pdmodel.interactive.documentnavigation.destination.PDPageDestination) {
+                    return ((org.apache.pdfbox.pdmodel.interactive.documentnavigation.destination.PDPageDestination) dest).retrievePageNumber() + 1;
+                }
+            } else if (item.getAction() != null) {
+                Object action = item.getAction();
+                if (action instanceof org.apache.pdfbox.pdmodel.interactive.action.PDActionGoTo) {
+                    org.apache.pdfbox.pdmodel.interactive.action.PDActionGoTo goToAction =
+                        (org.apache.pdfbox.pdmodel.interactive.action.PDActionGoTo) action;
+                    if (goToAction.getDestination() instanceof org.apache.pdfbox.pdmodel.interactive.documentnavigation.destination.PDPageDestination) {
+                        return ((org.apache.pdfbox.pdmodel.interactive.documentnavigation.destination.PDPageDestination) goToAction.getDestination()).retrievePageNumber() + 1;
+                    }
+                }
+            }
+        } catch (Exception e) {
+            // ignore
+        }
+        return -1;
+    }
+
+    /**
+     * 构建层级目录结构
+     */
+    private com.alibaba.fastjson2.JSONArray buildHierarchicalStructure(java.util.List<com.alibaba.fastjson2.JSONObject> chapters) {
+        com.alibaba.fastjson2.JSONArray rootArray = new com.alibaba.fastjson2.JSONArray();
+        if (chapters.isEmpty()) {
+            return rootArray;
+        }
+
+        java.util.List<com.alibaba.fastjson2.JSONObject> levelParents = new java.util.ArrayList<>();
+        for (com.alibaba.fastjson2.JSONObject chapter : chapters) {
+            int level = Math.max(0, chapter.getIntValue("level"));
+            if (chapter.getJSONArray("children") == null) {
+                chapter.put("children", new com.alibaba.fastjson2.JSONArray());
+            }
+
+            if (level == 0 || levelParents.isEmpty() || levelParents.get(0) == null) {
+                rootArray.add(chapter);
+                levelParents.clear();
+                levelParents.add(chapter);
+                continue;
+        }
+
+            int parentIndex = Math.min(level - 1, levelParents.size() - 1);
+            com.alibaba.fastjson2.JSONObject parent = parentIndex >= 0 ? levelParents.get(parentIndex) : null;
+            if (parent == null) {
+                // 找不到父节点时降级为根节点
+                rootArray.add(chapter);
+                levelParents.clear();
+                levelParents.add(chapter);
+                continue;
+            }
+
+            com.alibaba.fastjson2.JSONArray parentChildren = parent.getJSONArray("children");
+            if (parentChildren == null) {
+                parentChildren = new com.alibaba.fastjson2.JSONArray();
+                parent.put("children", parentChildren);
+            }
+            parentChildren.add(chapter);
+
+            if (level >= levelParents.size()) {
+                while (levelParents.size() < level) {
+                    levelParents.add(null);
+                }
+                levelParents.add(chapter);
+            } else {
+                levelParents.set(level, chapter);
+            }
+            for (int idx = level + 1; idx < levelParents.size(); idx++) {
+                levelParents.set(idx, null);
+            }
+        }
+
+        limitDirectoryDepth(rootArray, 2); // 只保留最多两层目录
+        return rootArray;
+    }
+
+    /**
+     * 控制目录层级深度
+     */
+    private void limitDirectoryDepth(com.alibaba.fastjson2.JSONArray nodes, int depthRemaining) {
+        if (nodes == null || depthRemaining <= 0) {
+            return;
+        }
+
+        for (int i = 0; i < nodes.size(); i++) {
+            com.alibaba.fastjson2.JSONObject node = nodes.getJSONObject(i);
+            if (depthRemaining == 1) {
+                node.put("children", new com.alibaba.fastjson2.JSONArray());
+            } else {
+                limitDirectoryDepth(node.getJSONArray("children"), depthRemaining - 1);
+            }
+        }
+    }
+
+    /**
+     * 生成基本的目录结构（当无法从PDF中准确抽取时使用）
+     */
+    private com.alibaba.fastjson2.JSONArray generateBasicDirectoryStructure(String filePath, String fileId, String fileName) {
+        try {
+            com.alibaba.fastjson2.JSONArray structured = buildStructuredDirectoryFallback(filePath);
+            if (structured != null && !structured.isEmpty()) {
+                logger.info("使用文本解析生成基础目录结构，文件ID: {}", fileId);
+                return structured;
+            }
+
+            com.alibaba.fastjson2.JSONArray directoryArray = new com.alibaba.fastjson2.JSONArray();
+            com.alibaba.fastjson2.JSONObject fallbackChapter = createChapterNode("chap-auto-0",
+                    fileName != null ? fileName.replaceAll("\\.pdf$", "") : "文档内容");
+            directoryArray.add(fallbackChapter);
+            logger.info("生成默认基础目录结构，文件ID: {}", fileId);
+            return directoryArray;
+
+        } catch (Exception e) {
+            logger.error("生成基本目录结构失败，文件ID: {}", fileId, e);
+            return null;
+        }
+    }
+
+    private com.alibaba.fastjson2.JSONArray buildStructuredDirectoryFallback(String filePath) {
+        String text = readPdfTextForOutline(filePath);
+        if (text == null || text.isBlank()) {
+            return new com.alibaba.fastjson2.JSONArray();
+        }
+        java.util.List<TitleNode> nodes = parseOutlineTitles(text);
+        if (nodes.isEmpty()) {
+            return new com.alibaba.fastjson2.JSONArray();
+        }
+        com.alibaba.fastjson2.JSONArray result = new com.alibaba.fastjson2.JSONArray();
+        com.alibaba.fastjson2.JSONObject currentParent = null;
+        int parentIndex = 0;
+        int childIndex = 0;
+        for (TitleNode node : nodes) {
+            if (node.depth <= 0 || currentParent == null) {
+                com.alibaba.fastjson2.JSONObject parent = createChapterNode("chap-auto-" + parentIndex++, node.title);
+                result.add(parent);
+                currentParent = parent;
+                childIndex = 0;
+            } else {
+                com.alibaba.fastjson2.JSONArray children = currentParent.getJSONArray("children");
+                if (children == null) {
+                    children = new com.alibaba.fastjson2.JSONArray();
+                    currentParent.put("children", children);
+                }
+                if (children.size() >= 20) {
+                    continue;
+                }
+                com.alibaba.fastjson2.JSONObject child = createChapterNode(
+                        currentParent.getString("chapterId") + "-" + childIndex++, node.title);
+                children.add(child);
+            }
+        }
+        return result;
+    }
+
+    private com.alibaba.fastjson2.JSONObject createChapterNode(String chapterId, String title) {
+        com.alibaba.fastjson2.JSONObject node = new com.alibaba.fastjson2.JSONObject();
+        node.put("chapterId", chapterId);
+        node.put("title", title);
+        node.put("page", "1");
+        node.put("startPage", 1);
+        node.put("endPage", 1);
+        node.put("children", new com.alibaba.fastjson2.JSONArray());
+        return node;
+    }
+
+    private String readPdfTextForOutline(String filePath) {
+        try (org.apache.pdfbox.pdmodel.PDDocument document = org.apache.pdfbox.Loader.loadPDF(new java.io.File(filePath))) {
+            org.apache.pdfbox.text.PDFTextStripper stripper = new org.apache.pdfbox.text.PDFTextStripper();
+            stripper.setStartPage(1);
+            stripper.setEndPage(Math.min(document.getNumberOfPages(), 15));
+            stripper.setSortByPosition(true);
+            return stripper.getText(document);
+        } catch (Exception e) {
+            logger.warn("读取PDF文本以生成目录失败: {}", e.getMessage());
+            return "";
+        }
+    }
+
+    private java.util.List<TitleNode> parseOutlineTitles(String text) {
+        java.util.List<TitleNode> nodes = new java.util.ArrayList<>();
+        java.util.Set<String> uniqueTitles = new java.util.LinkedHashSet<>();
+        String[] lines = text.split("\\r?\\n");
+        for (String rawLine : lines) {
+            String line = rawLine == null ? "" : rawLine.trim();
+            if (line.length() < 2) {
+                continue;
+            }
+            TitleNode node = toTitleNode(line);
+            if (node == null || node.title.length() < 2) {
+                continue;
+            }
+            if (uniqueTitles.add(node.title)) {
+                nodes.add(node);
+            }
+            if (nodes.size() >= 50) {
+                break;
+            }
+        }
+        return nodes;
+    }
+
+    private TitleNode toTitleNode(String line) {
+        for (PatternMapping mapping : PATTERN_MAPPINGS) {
+            Matcher matcher = mapping.pattern.matcher(line);
+            if (matcher.find()) {
+                String cleaned = line.substring(matcher.end()).trim();
+                if (cleaned.isEmpty()) {
+                    cleaned = line;
+                }
+                return new TitleNode(cleaned, mapping.depth);
+            }
+        }
+            return null;
+        }
+
+    private static class TitleNode {
+        final String title;
+        final int depth;
+
+        TitleNode(String title, int depth) {
+            this.title = title;
+            this.depth = depth;
+        }
+    }
+
+    private static class PatternMapping {
+        final Pattern pattern;
+        final int depth;
+
+        PatternMapping(String regex, int depth) {
+            this.pattern = Pattern.compile(regex);
+            this.depth = depth;
+        }
+    }
+
+    private static final java.util.List<PatternMapping> PATTERN_MAPPINGS = java.util.List.of(
+        new PatternMapping("^第[一二三四五六七八九十百千0-9]+章[\\s、\\.．-]*", 0),
+        new PatternMapping("^[一二三四五六七八九十]+、", 0),
+        new PatternMapping("^\\d{1,2}[\\.．、\\s]+", 0),
+        new PatternMapping("^\\d{1,2}\\.\\d{1,2}[\\.．、\\s]+", 1),
+        new PatternMapping("^\\d{1,2}\\.\\d{1,2}\\.\\d{1,2}[\\.．、\\s]+", 2),
+        new PatternMapping("^[（(][一二三四五六七八九十]+[）)]", 1),
+        new PatternMapping("^[（(]\\d+[）)]", 1)
+    );
+
+    /**
+     * 从文件路径中提取文件ID
+     */
+    private String extractFileIdFromPath(String filePath) {
+        try {
+            // 从文件路径中查找UUID格式的文件ID
+            // 例如: .../extracts/project-001_20251103_112040_39ad4981/eb949f7d-ba5f-4112-9daa-37d4ab7282e1.pdf
+            Path path = Paths.get(filePath);
+            String fileName = path.getFileName().toString();
+
+            // 移除扩展名
+            int lastDotIndex = fileName.lastIndexOf('.');
+            if (lastDotIndex > 0) {
+                String nameWithoutExt = fileName.substring(0, lastDotIndex);
+
+                // 检查是否是UUID格式
+                if (nameWithoutExt.matches("^[0-9a-f]{8}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{12}$")) {
+                    return nameWithoutExt;
+                }
+            }
+        } catch (Exception e) {
+            logger.warn("提取文件ID失败: {}", filePath, e);
+        }
+        return null;
+    }
+
+    /**
+     * 根据文件路径和项目ID从数据库中查找文件ID
+     */
+    private String findFileIdByPath(String filePath, String projectId) {
+        try {
+            Path path = Paths.get(filePath);
+            String fileName = path.getFileName().toString();
+
+            logger.debug("从数据库中查找文件ID: projectId={}, fileName={}", projectId, fileName);
+
+            // 从数据库中查找匹配的文件记录
+            List<File> files = fileRepository.findByProjectIdAndFileName(projectId, fileName);
+            if (!files.isEmpty()) {
+                String fileId = files.get(0).getFileId();
+                logger.debug("找到文件ID: {} for fileName: {}", fileId, fileName);
+                return fileId;
+            } else {
+                logger.warn("数据库中未找到匹配的文件记录: projectId={}, fileName={}", projectId, fileName);
+            }
+        } catch (Exception e) {
+            logger.error("从数据库查找文件ID失败: filePath={}, projectId={}", filePath, projectId, e);
+        }
+        return null;
+    }
+
+    /**
+     * 处理所有章节的总结（并行处理优化，修复乐观锁冲突）
+     */
+    @Transactional
+    private void processChapterSummaries(String fileId) {
+        try {
+            logger.info("开始并行处理章节总结，文件ID: {}", fileId);
+
+            // 获取所有章节
+            List<FileDirectory> directories = fileDirectoryRepository.findByFileIdOrderBySortOrderAsc(fileId);
+            
+            if (directories.isEmpty()) {
+                logger.info("没有需要处理的章节，文件ID: {}", fileId);
+                return;
+            }
+            
+            logger.info("共有 {} 个章节需要处理，文件ID: {}", directories.size(), fileId);
+            
+            // 用于收集每个章节的总结结果 (chapterId -> summary)
+            java.util.concurrent.ConcurrentHashMap<String, String> summaryResults = 
+                new java.util.concurrent.ConcurrentHashMap<>();
+            
+            // 使用并行处理，控制并发数为3避免数据库连接池耗尽
+            int batchSize = 3;
+            List<List<FileDirectory>> batches = new ArrayList<>();
+            for (int i = 0; i < directories.size(); i += batchSize) {
+                batches.add(directories.subList(i, Math.min(i + batchSize, directories.size())));
+            }
+            
+            for (List<FileDirectory> batch : batches) {
+                // 每批章节并行生成总结（只生成，不保存）
+                List<CompletableFuture<Void>> futures = batch.stream()
+                    .map(directory -> CompletableFuture.runAsync(() -> {
+                        try {
+                            String summary = generateChapterSummary(directory);
+                            if (summary != null && !summary.isEmpty()) {
+                                // 将结果存入并发Map，而不是直接更新实体
+                                summaryResults.put(directory.getChapterId(), summary);
+                                logger.debug("章节总结生成完成: fileId={}, chapterId={}", 
+                                    fileId, directory.getChapterId());
+                            }
+                        } catch (Exception e) {
+                            logger.error("生成章节总结失败: fileId={}, chapterId={}", 
+                                fileId, directory.getChapterId(), e);
+                        }
+                    }))
+                    .collect(Collectors.toList());
+                
+                // 等待当前批次完成
+                CompletableFuture.allOf(futures.toArray(new CompletableFuture[0])).join();
+            }
+            
+            // 统一保存所有结果（在主线程中按顺序保存，避免乐观锁冲突）
+            if (!summaryResults.isEmpty()) {
+                logger.info("开始保存 {} 个章节的总结结果", summaryResults.size());
+                for (FileDirectory directory : directories) {
+                    String summary = summaryResults.get(directory.getChapterId());
+                    if (summary != null) {
+                        // 重新从数据库加载最新的实体，避免过期实体问题
+                        FileDirectory freshDirectory = fileDirectoryRepository.findByFileIdAndChapterId(
+                            fileId, directory.getChapterId());
+                        if (freshDirectory != null) {
+                            freshDirectory.setSummary(summary);
+                            fileDirectoryRepository.save(freshDirectory);
+                        }
+                    }
+                }
+            }
+
+            logger.info("章节总结并行处理完成，文件ID: {}, 章节数: {}, 成功总结: {}", 
+                fileId, directories.size(), summaryResults.size());
+
+        } catch (Exception e) {
+            logger.error("处理章节总结出错，文件ID: {}", fileId, e);
+        }
+    }
+
+    /**
+     * 使用AI模型生成章节总结
+     */
+    private String generateChapterSummary(FileDirectory directory) {
+        try {
+            // 获取文档摘要提示词模板
+            com.zhpb.gdyd_zhpb_zgf.entity.mysql.PromptTemplate summaryTemplate = promptTemplateService.getActiveTemplate("summary");
+            if (summaryTemplate == null) {
+                throw new RuntimeException("未找到激活的文档摘要提示词模板，请在提示词配置页面启用或创建summary类型的提示词模板");
+            }
+
+            String promptTemplate = summaryTemplate.getPromptContent();
+            logger.info("使用数据库提示词模板: {} (ID: {}), 内容长度: {}", summaryTemplate.getTemplateName(), summaryTemplate.getTemplateId(), promptTemplate.length());
+            logger.debug("提示词内容: {}", promptTemplate);
+
+            // 使用模板替换变量
+            String prompt = promptTemplate.replace("{title}", directory.getTitle());
+            logger.debug("最终使用的提示词: {}", prompt);
+
+            logger.info("调用AI模型生成章节总结: chapterId={}", directory.getChapterId());
+
+            String userMessage = chapterContentResolver.buildUserMessage(directory);
+
+            ModelRequestContextHolder.setRequestSource("DOC_SUMMARY_AUTO");
+            String summary;
+            try {
+                summary = aiModelService.generateResponse(prompt, userMessage);
+            } finally {
+                ModelRequestContextHolder.clear();
+            }
+
+            if (summary != null && !summary.trim().isEmpty()) {
+                logger.info("AI模型返回章节总结，长度: {}", summary.length());
+                return summary.trim();
+            }
+
+            return null;
+
+        } catch (Exception e) {
+            logger.error("调用AI模型生成章节总结失败: chapterId={}", directory.getChapterId(), e);
+            return null;
+        }
+    }
+
+    @Override
+    public void processProcurementFilesAsync(String projectId, String packageId, List<String> fileUrls, String traceId) {
+        // 异步处理采购文件下载
+        CompletableFuture.runAsync(() -> {
+            try {
+                logger.info("开始异步处理采购文件下载: traceId={}, projectId={}, fileCount={}",
+                           traceId, projectId, fileUrls.size());
+
+                // 确保项目存在
+                createOrUpdateProject(projectId, null);
+
+                // 替换下载地址
+                List<String> processedFileUrls = new ArrayList<>();
+                if (fileUrls != null) {
+                    for (String url : fileUrls) {
+                        processedFileUrls.add(replaceDownloadUrl(url));
+                    }
+                }
+
+                // 下载文件并处理
+                List<String> downloadedFiles = downloadFilesFromUrls(processedFileUrls, projectId, null, traceId);
+
+                if (!downloadedFiles.isEmpty()) {
+                    // 处理下载的文件（解压、保存到数据库、向量化等）
+                    for (String zipFilePath : downloadedFiles) {
+                        try {
+                            // 处理ZIP文件
+                            processDownloadedZipFile(projectId, zipFilePath, null, traceId, null);
+                        } catch (Exception e) {
+                            logger.error("处理下载的ZIP文件失败: file={}, traceId={}", zipFilePath, traceId, e);
+                        }
+                    }
+                }
+
+                logger.info("采购文件异步处理完成: traceId={}", traceId);
+
+            } catch (Exception e) {
+                logger.error("异步处理采购文件下载失败: traceId={}", traceId, e);
+            }
+        });
+    }
+
+    @Override
+    public void processResponseFilesAsync(String projectId, String packageId, String supplierId,
+                                        List<String> fileUrls, String fileType, String jsonFileId,
+                                        String unitId, String traceId) {
+        // 异步处理应答文件下载
+        CompletableFuture.runAsync(() -> {
+            try {
+                logger.info("开始异步处理应答文件下载: traceId={}, projectId={}, supplierId={}, fileCount={}",
+                           traceId, projectId, supplierId, fileUrls.size());
+
+                // 确保项目存在
+                createOrUpdateProject(projectId, null);
+                // 确保供应商存在，名称暂用supplierId
+                if (supplierId != null && !supplierId.isEmpty()) {
+                    ensureSupplierExists(supplierId, supplierId);
+                }
+
+                // 替换下载地址
+                List<String> processedFileUrls = new ArrayList<>();
+                if (fileUrls != null) {
+                    for (String url : fileUrls) {
+                        processedFileUrls.add(replaceDownloadUrl(url));
+                    }
+                }
+                
+                // 映射fileType到数据库ID
+                Long dbFileTypeId = null;
+                if ("1".equals(fileType)) {
+                    dbFileTypeId = 6L; // 商务相关的文件
+                } else if ("2".equals(fileType)) {
+                    dbFileTypeId = 5L; // 技术方案、技术规范等相关文件
+                }
+
+                // 下载文件并处理
+                List<String> downloadedFiles = downloadFilesFromUrls(processedFileUrls, projectId, supplierId, traceId);
+
+                if (!downloadedFiles.isEmpty()) {
+                    // 处理下载的文件（解压、保存到数据库、向量化等）
+                    for (String zipFilePath : downloadedFiles) {
+                        try {
+                            // 处理ZIP文件
+                            processDownloadedZipFile(projectId, zipFilePath, supplierId, traceId, dbFileTypeId);
+                        } catch (Exception e) {
+                            logger.error("处理下载的ZIP文件失败: file={}, traceId={}", zipFilePath, traceId, e);
+                        }
+                    }
+                }
+
+                logger.info("应答文件异步处理完成: traceId={}", traceId);
+
+            } catch (Exception e) {
+                logger.error("异步处理应答文件下载失败: traceId={}", traceId, e);
+            }
+        });
+    }
+
+    /**
+     * 从URL列表下载文件
+     */
+    private List<String> downloadFilesFromUrls(List<String> fileUrls, String projectId, String supplierId, String traceId) {
+        List<String> downloadedFiles = new ArrayList<>();
+
+        for (String url : fileUrls) {
+            try {
+                logger.info("开始下载文件: url={}, traceId={}", url, traceId);
+
+                // 生成下载文件名
+                String fileName = generateFileNameFromUrl(url);
+                Path baseUploadPath = getUploadPath(projectId);
+                if (supplierId != null && !supplierId.isEmpty()) {
+                    baseUploadPath = baseUploadPath.resolve("supplier").resolve(supplierId);
+                }
+                Files.createDirectories(baseUploadPath);
+                Path downloadPath = baseUploadPath.resolve(fileName);
+
+                // 下载文件
+                downloadFileFromUrl(url, downloadPath.toString());
+
+                downloadedFiles.add(downloadPath.toString());
+                logger.info("文件下载完成: url={}, localPath={}, traceId={}", url, downloadPath, traceId);
+
+            } catch (Exception e) {
+                logger.error("下载文件失败: url={}, traceId={}", url, traceId, e);
+            }
+        }
+
+        return downloadedFiles;
+    }
+
+    /**
+     * 确保供应商存在；若不存在则以supplierId作为名称创建
+     */
+    private void ensureSupplierExists(String supplierId, String supplierName) {
+        if (supplierId == null || supplierId.isEmpty()) {
+            return;
+        }
+        try {
+            Supplier existing = supplierRepository.findBySupplierId(supplierId);
+            if (existing == null) {
+                Supplier newSupplier = new Supplier();
+                newSupplier.setSupplierId(supplierId);
+                newSupplier.setSupplierName(
+                        supplierName != null && !supplierName.isEmpty() ? supplierName : supplierId);
+                newSupplier.setStatus("active");
+                supplierRepository.save(newSupplier);
+                logger.info("自动创建供应商: {}", supplierId);
+            }
+        } catch (Exception e) {
+            logger.warn("自动创建供应商失败: supplierId={}, error={}", supplierId, e.getMessage());
+        }
+    }
+
+    /**
+     * 处理下载的ZIP文件
+     */
+    private void processDownloadedZipFile(String projectId, String zipFilePath, String supplierId, String traceId, Long explicitFileTypeId) throws IOException {
+        logger.info("开始处理下载的ZIP文件: projectId={}, supplierId={}, zipFile={}, traceId={}, fileTypeId={}", 
+                projectId, supplierId, zipFilePath, traceId, explicitFileTypeId);
+
+        Path zipPath = Paths.get(zipFilePath);
+        if (!Files.exists(zipPath)) {
+            throw new FileNotFoundException("ZIP文件不存在: " + zipFilePath);
+        }
+
+        // 获取解压路径
+        Path extractPath = getExtractPath(projectId);
+        if (supplierId != null && !supplierId.isEmpty()) {
+            extractPath = extractPath.resolve("supplier").resolve(supplierId);
+        }
+        Files.createDirectories(extractPath);
+
+        // 解压ZIP文件
+        List<String> extractedFilePaths = extractZipFile(zipPath, extractPath);
+
+        if (extractedFilePaths.isEmpty()) {
+            logger.warn("ZIP文件解压后没有找到有效文件: {}", zipFilePath);
+            return;
+        }
+
+        // 保存ZIP文件路径信息（可选）
+        String savedZipPath = zipPath.toString();
+
+        // 异步处理解压后的文件
+        asyncProcessUploadedFiles(projectId, savedZipPath, extractedFilePaths, extractedFilePaths, supplierId, explicitFileTypeId);
+
+        logger.info("ZIP文件处理完成: projectId={}, zipFile={}, extractedFiles={}, traceId={}",
+                   projectId, zipFilePath, extractedFilePaths.size(), traceId);
+    }
+
+    /**
+     * 从URL生成文件名
+     */
+    private String generateFileNameFromUrl(String url) {
+        try {
+            String fileName = url.substring(url.lastIndexOf('/') + 1);
+            if (fileName.contains("?")) {
+                fileName = fileName.substring(0, fileName.indexOf("?"));
+            }
+            if (fileName.isEmpty()) {
+                fileName = "downloaded_file_" + System.currentTimeMillis() + ".zip";
+            }
+            return fileName;
+        } catch (Exception e) {
+            return "downloaded_file_" + System.currentTimeMillis() + ".zip";
+        }
+    }
+
+    /**
+     * 替换下载URL中的Host
+     */
+    private String replaceDownloadUrl(String url) {
+        if (url == null || downloadReplaceSource == null || downloadReplaceSource.isEmpty() ||
+            downloadReplaceTarget == null || downloadReplaceTarget.isEmpty()) {
+            return url;
+        }
+        
+        if (url.contains(downloadReplaceSource)) {
+            String newUrl = url.replace(downloadReplaceSource, downloadReplaceTarget);
+            logger.debug("替换下载地址: {} -> {}", url, newUrl);
+            return newUrl;
+        }
+        
+        return url;
+    }
+
+    /**
+     * 从URL下载文件
+     */
+    private void downloadFileFromUrl(String url, String localPath) throws IOException {
+        logger.info("正在下载文件: {} -> {}", url, localPath);
+
+        try {
+            // 使用HttpClientUtils下载文件
+            byte[] fileData = HttpClientUtils.downloadFile(url);
+
+            // 写入本地文件
+            java.io.File file = new java.io.File(localPath);
+            file.getParentFile().mkdirs();
+
+            try (FileOutputStream fos = new FileOutputStream(file)) {
+                fos.write(fileData);
+            }
+
+            logger.info("文件下载成功: url={}, localPath={}, size={}", url, localPath, fileData.length);
+
+        } catch (Exception e) {
+            logger.error("文件下载失败: url={}", url, e);
+            throw new IOException("文件下载失败: " + url, e);
+        }
+    }
+
+    /**
+     * 检查文件扩展名是否为内容文件类型（PDF、Word、PPT、TXT、JSON）
+     */
+    private boolean isContentFileType(String extension) {
+        if (extension == null || extension.isEmpty()) {
+            return false;
+        }
+        String ext = extension.toLowerCase();
+        return ext.equals("pdf") || ext.equals("doc") || ext.equals("docx") 
+                || ext.equals("ppt") || ext.equals("pptx") 
+                || ext.equals("txt") || ext.equals("json");
+    }
+
+    /**
+     * 获取项目的上传路径
+     */
+    @Override
+    public Path getUploadPath(String projectId) {
+        return Paths.get(absoluteUploadPath, projectId);
+    }
+
+    /**
+     * 获取项目的解压路径
+     */
+    @Override
+    public Path getExtractPath(String projectId) {
+        return Paths.get(absoluteExtractPath, projectId);
+    }
+
+    /**
+     * 获取文件扩展名
+     */
+    private String getFileExtension(String fileName) {
+        if (fileName == null || fileName.isEmpty()) {
+            return "";
+        }
+        int lastDot = fileName.lastIndexOf('.');
+        return lastDot > 0 ? fileName.substring(lastDot + 1).toLowerCase() : "";
+    }
+
+    /**
+     * 列出ZIP文件内容（用于调试）
+     */
+    private void listZipContents(Path zipFilePath) {
+        try (java.util.zip.ZipFile zipFile = new java.util.zip.ZipFile(zipFilePath.toFile())) {
+            java.util.Enumeration<? extends java.util.zip.ZipEntry> entries = zipFile.entries();
+            logger.info("ZIP文件内容列表:");
+            while (entries.hasMoreElements()) {
+                java.util.zip.ZipEntry entry = entries.nextElement();
+                logger.info("  - {} (大小: {} 字节)", entry.getName(), entry.getSize());
+            }
+        } catch (Exception e) {
+            logger.warn("列出ZIP内容失败: {}", e.getMessage());
+        }
+    }
+
+    /**
+     * 异步处理文件目录抽取
+     */
+    @Async
+    public void processFileDirectoriesAsync(List<String> filePaths, String projectId) {
+        long startTime = System.currentTimeMillis();
+        logger.info("[目录摘取] ========== 开始异步处理文件目录抽取 ==========");
+        logger.info("[目录摘取] 项目ID: {}, 文件数量: {}", projectId, filePaths != null ? filePaths.size() : 0);
+        
+        if (filePaths == null || filePaths.isEmpty()) {
+            logger.warn("[目录摘取] 文件路径列表为空，跳过处理");
+            return;
+        }
+        
+        List<CompletableFuture<Void>> futures = new ArrayList<>();
+        for (String filePath : filePaths) {
+            futures.add(CompletableFuture.runAsync(() -> processDirectoryTask(filePath, projectId), directoryExecutor));
+        }
+
+        for (CompletableFuture<Void> future : futures) {
+            try {
+                future.join();
+            } catch (Exception e) {
+                logger.error("[目录摘取] 并发处理文件目录失败: {}", e.getMessage());
+            }
+        }
+
+        long duration = System.currentTimeMillis() - startTime;
+        logger.info("[目录摘取] ========== 完成! 耗时: {}ms，项目ID: {} ==========", duration, projectId);
+    }
+
+    /**
+     * 处理单个文件的目录抽取任务
+     */
+    private void processDirectoryTask(String filePath, String projectId) {
+        logger.info("[目录摘取] 开始处理文件: {}", filePath);
+        try {
+            Path path = Paths.get(filePath);
+            String fileName = path.getFileName().toString();
+            if (fileName.startsWith(".")) {
+                logger.info("[目录摘取] 跳过临时/隐藏文件: {}", fileName);
+                return;
+            }
+            
+            transactionTemplate.execute(status -> {
+                try {
+                    Path absolutePath = path.toAbsolutePath().normalize();
+                    String storedPath = filePathResolver.normalizeForStorage(absolutePath);
+                    java.util.Optional<File> fileOpt = fileRepository.findByFilePath(storedPath);
+                    
+                    if (fileOpt.isPresent()) {
+                        File file = fileOpt.get();
+                        file.setCatalogExtractionStatus("processing");
+                        file.setCatalogExtractionRetryCount(file.getCatalogExtractionRetryCount() != null ? file.getCatalogExtractionRetryCount() + 1 : 1);
+                        fileRepository.save(file);
+
+                        // 执行实际的目录抽取
+                        com.alibaba.fastjson2.JSONArray directoryStructure = null;
+                        String fileExtension = file.getFileExtension();
+                        if ("pdf".equalsIgnoreCase(fileExtension)) {
+                            directoryStructure = extractDirectoryFromPDF(filePath, file.getFileId(), file.getFileName());
+                        }
+
+                        if (directoryStructure != null && !directoryStructure.isEmpty()) {
+                            // 先删除旧的目录数据
+                            fileDirectoryRepository.deleteByFileId(file.getFileId());
+                            // 保存新的目录数据
+                            saveDirectoryItems(file.getFileId(), directoryStructure, null, 1);
+                            file.setCatalogExtractionStatus("completed");
+                            file.setCatalogExtractionTime(LocalDateTime.now());
+                            fileRepository.save(file);
+                            logger.info("[目录摘取] 文件目录抽取成功: fileId={}, fileName={}, 目录项数={}", 
+                                    file.getFileId(), file.getFileName(), directoryStructure.size());
+                        } else {
+                            // 尝试生成基本目录结构
+                            com.alibaba.fastjson2.JSONArray basicDirectory = generateBasicDirectoryStructure(filePath, file.getFileId(), file.getFileName());
+                            if (basicDirectory != null && !basicDirectory.isEmpty()) {
+                                fileDirectoryRepository.deleteByFileId(file.getFileId());
+                                saveDirectoryItems(file.getFileId(), basicDirectory, null, 1);
+                                file.setCatalogExtractionStatus("completed");
+                                file.setCatalogExtractionTime(LocalDateTime.now());
+                                fileRepository.save(file);
+                                logger.info("[目录摘取] 使用基本目录结构: fileId={}, fileName={}", file.getFileId(), file.getFileName());
+                            } else {
+                                file.setCatalogExtractionStatus("failed");
+                                file.setDescription("目录抽取失败：未提取到有效目录");
+                                fileRepository.save(file);
+                                logger.warn("[目录摘取] 文件目录抽取失败: fileId={}, fileName={}", file.getFileId(), file.getFileName());
+                            }
+                        }
+                    } else {
+                        logger.warn("[目录摘取] 未找到文件记录，跳过目录抽取: {}", filePath);
+                    }
+                    return null;
+                } catch (Exception e) {
+                    logger.error("[目录摘取] 事务内处理失败: {}", filePath, e);
+                    status.setRollbackOnly();
+                    throw e;
+                }
+            });
+        } catch (Exception e) {
+            logger.error("[目录摘取] 处理文件目录失败: {}", filePath, e);
+        }
+    }
+
+    /**
+     * 从文件路径提取文件类型ID
+     */
+    private Long extractFileTypeIdFromPath(String filePath) {
+        if (filePath == null) {
+            return null;
+        }
+        String lowerPath = filePath.toLowerCase();
+        FileType fileType = null;
+        if (lowerPath.contains("技术文件") || lowerPath.contains("技术标") || lowerPath.contains("技术方案")) {
+            fileType = fileTypeService.createOrGetFileType("技术文件", "技术类投标文件");
+        } else if (lowerPath.contains("商务文件") || lowerPath.contains("商务标")) {
+            fileType = fileTypeService.createOrGetFileType("商务文件", "商务类投标文件");
+        } else if (lowerPath.contains("报价文件") || lowerPath.contains("报价标") || lowerPath.contains("价格")) {
+            fileType = fileTypeService.createOrGetFileType("报价文件", "报价类投标文件");
+        }
+        return fileType != null ? fileType.getId() : null;
+    }
+
+    /**
+     * 从文件路径提取供应商ID
+     */
+    private String extractSupplierIdFromPath(String filePath) {
+        if (filePath == null) {
+            return null;
+        }
+        // 尝试从路径中提取供应商名称
+        // 典型路径格式: /uploads/projectId/供应商名称/技术文件/xxx.pdf
+        try {
+            Path path = Paths.get(filePath);
+            int count = path.getNameCount();
+            // 查找可能的供应商目录
+            for (int i = 0; i < count - 1; i++) {
+                String segment = path.getName(i).toString();
+                // 排除常见的非供应商目录名
+                if (!segment.equals("uploads") && !segment.equals("extracts") && 
+                    !segment.contains("技术文件") && !segment.contains("商务文件") && 
+                    !segment.contains("报价文件") && !segment.endsWith(".pdf") &&
+                    !segment.endsWith(".doc") && !segment.endsWith(".docx") &&
+                    segment.length() > 4 && !segment.matches("^\\d+_.*")) {
+                    // 可能是供应商名称
+                    Supplier existing = supplierRepository.findBySupplierId(segment);
+                    if (existing != null) {
+                        return segment;
+                    }
+                    // 尝试按名称查找
+                    existing = supplierRepository.findBySupplierName(segment);
+                    if (existing != null) {
+                        return existing.getSupplierId();
+                    }
+                }
+            }
+        } catch (Exception e) {
+            logger.debug("从路径提取供应商ID失败: {}", e.getMessage());
+        }
+        return null;
+    }
+
+    /**
+     * 根据项目ID获取项目文件信息
+     */
+    @Override
+    public com.zhpb.gdyd_zhpb_zgf.dto.ProjectFilesResponse.ProjectFilesData getProjectFiles(String projectId) {
+        com.zhpb.gdyd_zhpb_zgf.dto.ProjectFilesResponse.ProjectFilesData data = 
+            new com.zhpb.gdyd_zhpb_zgf.dto.ProjectFilesResponse.ProjectFilesData();
+        data.setProjectId(projectId);
+        
+        List<File> files = fileRepository.findByProjectId(projectId);
+        data.setTotalFiles(files.size());
+        
+        // 转换为DTO
+        List<com.zhpb.gdyd_zhpb_zgf.dto.ProjectFilesResponse.FileInfo> fileInfoList = new ArrayList<>();
+        for (File file : files) {
+            com.zhpb.gdyd_zhpb_zgf.dto.ProjectFilesResponse.FileInfo info = 
+                new com.zhpb.gdyd_zhpb_zgf.dto.ProjectFilesResponse.FileInfo();
+            info.setFileId(file.getFileId());
+            info.setFileName(file.getFileName());
+            info.setFileSize(file.getFileSize());
+            info.setFileType(file.getFileExtension());
+            info.setCompanyName(file.getSupplierId());
+            fileInfoList.add(info);
+        }
+        data.setFiles(fileInfoList);
+        
+        return data;
+    }
+
+    /**
+     * 从PDF中提取目录结构
+     */
+    private com.alibaba.fastjson2.JSONArray extractDirectoryFromPDF(String filePath, String fileId, String fileName) {
+        try {
+            logger.info("[目录摘取] 开始处理PDF文件: {}", fileName);
+            
+            int totalPages = 1;
+            java.util.List<com.alibaba.fastjson2.JSONObject> chapters = new java.util.ArrayList<>();
+            
+            try (org.apache.pdfbox.pdmodel.PDDocument document = org.apache.pdfbox.Loader.loadPDF(new java.io.File(filePath))) {
+                totalPages = document.getNumberOfPages();
+                logger.info("[目录摘取] PDF总页数: {}", totalPages);
+                
+                // 尝试从PDF大纲中提取目录
+                org.apache.pdfbox.pdmodel.interactive.documentnavigation.outline.PDDocumentOutline outline = 
+                    document.getDocumentCatalog().getDocumentOutline();
+                if (outline != null && outline.getFirstChild() != null) {
+                    logger.info("[目录摘取] 发现PDF大纲，开始提取...");
+                    extractOutlineItems(outline.getFirstChild(), "", chapters, 1);
+                    logger.info("[目录摘取] 从大纲提取到 {} 个章节", chapters.size());
+                }
+            }
+            
+            if (!chapters.isEmpty()) {
+                // 构建层级结构
+                com.alibaba.fastjson2.JSONArray hierarchicalArray = buildHierarchicalStructure(chapters);
+                logger.info("[目录摘取] 成功构建层级目录结构，根节点数: {}", hierarchicalArray.size());
+                return hierarchicalArray;
+            }
+            
+            // 回退方案：使用文件名作为单个章节
+            String title = fileName;
+            if (title != null && title.toLowerCase().endsWith(".pdf")) {
+                title = title.substring(0, title.length() - 4);
+            }
+            if (title == null || title.isBlank()) {
+                title = "文档内容";
+            }
+            
+            com.alibaba.fastjson2.JSONArray directoryArray = new com.alibaba.fastjson2.JSONArray();
+            com.alibaba.fastjson2.JSONObject chapter = new com.alibaba.fastjson2.JSONObject();
+            chapter.put("chapterId", "chap-0");
+            chapter.put("title", title);
+            chapter.put("level", 1);
+            chapter.put("page", "1");
+            chapter.put("startPage", 1);
+            chapter.put("endPage", totalPages);
+            chapter.put("children", new com.alibaba.fastjson2.JSONArray());
+            directoryArray.add(chapter);
+            
+            logger.info("[目录摘取] 使用文件名作为默认章节: '{}', 页数范围: 1-{}", title, totalPages);
+            return directoryArray;
+            
+        } catch (Exception e) {
+            logger.error("[目录摘取] 处理PDF失败: {}", e.getMessage(), e);
+            return null;
+        }
+    }
+
+    /**
+     * 保存目录项到数据库
+     */
+    private void saveDirectoryItems(String fileId, com.alibaba.fastjson2.JSONArray items, String parentChapterId, int sortStart) {
+        if (items == null || items.isEmpty()) {
+            return;
+        }
+        
+        int sortOrder = sortStart;
+        for (int i = 0; i < items.size(); i++) {
+            com.alibaba.fastjson2.JSONObject item = items.getJSONObject(i);
+            try {
+                FileDirectory directory = new FileDirectory();
+                directory.setFileId(fileId);
+                directory.setChapterId(item.getString("chapterId"));
+                directory.setTitle(item.getString("title"));
+                directory.setPage(item.getString("page"));
+                directory.setStartPage(item.getInteger("startPage"));
+                directory.setEndPage(item.getInteger("endPage"));
+                directory.setParentChapterId(parentChapterId);
+                directory.setLevel(item.getIntValue("level", 1));
+                directory.setSortOrder(sortOrder++);
+                directory.setSummary(item.getString("summary"));
+                
+                fileDirectoryRepository.save(directory);
+                
+                // 递归保存子章节
+                com.alibaba.fastjson2.JSONArray children = item.getJSONArray("children");
+                if (children != null && !children.isEmpty()) {
+                    saveDirectoryItems(fileId, children, item.getString("chapterId"), 1);
+                }
+                
+            } catch (Exception e) {
+                logger.error("[目录摘取] 保存目录项失败: {}", e.getMessage(), e);
+            }
+        }
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/FileTypeServiceImpl.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/FileTypeServiceImpl.java
new file mode 100644
index 0000000..db047b4
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/FileTypeServiceImpl.java
@@ -0,0 +1,198 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.service.impl
+ * @ClassName: FileTypeServiceImpl
+ * @Description: 文件类型服务实现类
+ * @Author: 张志锋
+ * @Date: 2025-11-01
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.service.impl;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.FileType;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.FileTypeRepository;
+import com.zhpb.gdyd_zhpb_zgf.service.FileTypeService;
+import com.alibaba.fastjson2.JSON;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.stereotype.Service;
+import org.springframework.transaction.annotation.Transactional;
+
+import jakarta.annotation.PostConstruct;
+import java.util.ArrayList;
+import java.util.List;
+import java.util.Optional;
+
+/**
+ * 文件类型服务实现类
+ */
+@Service
+public class FileTypeServiceImpl implements FileTypeService {
+
+    private static final Logger logger = LoggerFactory.getLogger(FileTypeServiceImpl.class);
+
+    @Autowired
+    private FileTypeRepository fileTypeRepository;
+
+    /**
+     * 初始化默认文件类型
+     */
+    @PostConstruct
+    public void initDefaultFileTypes() {
+        try {
+            // 投标文件
+            createOrGetFileType("投标文件", "投标相关的文件，包括投标书、报价单等", 
+                    new String[]{"投标", "标书", "报价", "tender", "bid"}, 100);
+            
+            // 采购文件
+            createOrGetFileType("采购文件", "采购相关的文件，包括采购计划、采购需求等", 
+                    new String[]{"采购", "purchase", "procurement"}, 90);
+            
+            // 结果文件
+            createOrGetFileType("结果文件", "评标结果、评审结果等相关文件", 
+                    new String[]{"结果", "评审", "评标", "result", "review", "evaluation"}, 80);
+            
+            // 合同文件
+            createOrGetFileType("合同文件", "合同相关的文件", 
+                    new String[]{"合同", "contract", "agreement"}, 70);
+            
+            // 技术文件
+            createOrGetFileType("技术文件", "技术方案、技术规范等相关文件", 
+                    new String[]{"技术", "technical", "specification", "规范"}, 60);
+            
+            // 商务文件
+            createOrGetFileType("商务文件", "商务相关的文件", 
+                    new String[]{"商务", "business", "commercial"}, 50);
+            
+            logger.info("默认文件类型初始化完成");
+        } catch (Exception e) {
+            logger.error("初始化默认文件类型失败: {}", e.getMessage(), e);
+        }
+    }
+
+    /**
+     * 创建或获取文件类型（带关键词）
+     */
+    private FileType createOrGetFileType(String typeName, String description, String[] keywords, Integer priority) {
+        String typeCode = generateTypeCode(typeName);
+        
+        Optional<FileType> existing = fileTypeRepository.findByTypeCode(typeCode);
+        if (existing.isPresent()) {
+            return existing.get();
+        }
+
+        FileType fileType = new FileType();
+        fileType.setTypeCode(typeCode);
+        fileType.setTypeName(typeName);
+        fileType.setDescription(description);
+        fileType.setKeywords(JSON.toJSONString(keywords != null ? List.of(keywords) : new ArrayList<>()));
+        fileType.setPriority(priority != null ? priority : 0);
+        fileType.setEnabled(true);
+
+        fileTypeRepository.save(fileType);
+        logger.info("创建文件类型: {} (code: {})", typeName, typeCode);
+        return fileType;
+    }
+
+    @Override
+    @Transactional
+    public FileType identifyFileType(String fileName, String filePath) {
+        if (fileName == null) {
+            fileName = "";
+        }
+        if (filePath == null) {
+            filePath = "";
+        }
+
+        // 合并文件名和路径用于匹配
+        String searchText = (fileName + " " + filePath).toLowerCase();
+
+        // 获取所有启用的文件类型，按优先级排序
+        List<FileType> enabledTypes = fileTypeRepository.findByEnabledOrderByPriorityDesc(true);
+
+        // 遍历匹配关键词
+        for (FileType fileType : enabledTypes) {
+            if (fileType.getKeywords() != null && !fileType.getKeywords().isEmpty()) {
+                try {
+                    List<String> keywords = JSON.parseArray(fileType.getKeywords(), String.class);
+                    for (String keyword : keywords) {
+                        if (keyword != null && !keyword.isEmpty() && searchText.contains(keyword.toLowerCase())) {
+                            logger.debug("文件类型识别成功: {} -> {} (匹配关键词: {})", fileName, fileType.getTypeName(), keyword);
+                            return fileType;
+                        }
+                    }
+                } catch (Exception e) {
+                    logger.warn("解析文件类型关键词失败: {}", e.getMessage());
+                }
+            }
+        }
+
+        // 如果无法识别，返回null或创建一个"未知类型"
+        logger.debug("无法识别文件类型: {}", fileName);
+        return null;
+    }
+
+    @Override
+    @Transactional
+    public FileType createOrGetFileType(String typeName, String description) {
+        // 先尝试通过名称查找，避免重复创建
+        Optional<FileType> existingByName = fileTypeRepository.findByTypeName(typeName);
+        if (existingByName.isPresent()) {
+            logger.debug("通过名称找到已存在的文件类型: {} -> ID={}", typeName, existingByName.get().getId());
+            return existingByName.get();
+        }
+        
+        // 如果不存在，创建新文件类型
+        return createOrGetFileType(typeName, description, null, 0);
+    }
+
+    @Override
+    public FileType getFileTypeByCode(String typeCode) {
+        return fileTypeRepository.findByTypeCode(typeCode).orElse(null);
+    }
+
+    /**
+     * 从类型名称生成类型代码
+     * 对于中文名称，使用拼音首字母或直接使用名称的哈希值作为后缀
+     */
+    private String generateTypeCode(String typeName) {
+        if (typeName == null || typeName.isEmpty()) {
+            return "unknown";
+        }
+        
+        // 先转换为小写并替换空格
+        String code = typeName.toLowerCase().replaceAll("\\s+", "_");
+        
+        // 检查是否包含中文字符
+        boolean hasChinese = code.matches(".*[\\u4e00-\\u9fa5].*");
+        
+        if (hasChinese) {
+            // 如果包含中文，使用名称的哈希值作为代码（确保唯一性）
+            // 使用typeName的哈希值，避免不同中文名称生成相同代码
+            int hash = Math.abs(typeName.hashCode());
+            // 对于常见的中文文件类型，使用固定映射
+            if (typeName.contains("技术")) {
+                return "technical_file";
+            } else if (typeName.contains("商务")) {
+                return "business_file";
+            } else if (typeName.contains("报价")) {
+                return "quotation_file";
+            } else if (typeName.contains("投标")) {
+                return "tender_file";
+            } else if (typeName.contains("采购")) {
+                return "purchase_file";
+            } else if (typeName.contains("结果")) {
+                return "result_file";
+            } else {
+                // 其他中文名称使用哈希值
+                return "file_type_" + hash;
+            }
+        } else {
+            // 不包含中文，直接移除非字母数字下划线字符
+            return code.replaceAll("[^a-z0-9_]", "");
+        }
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/ModelRequestLogServiceImpl.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/ModelRequestLogServiceImpl.java
new file mode 100644
index 0000000..6860dff
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/ModelRequestLogServiceImpl.java
@@ -0,0 +1,37 @@
+package com.zhpb.gdyd_zhpb_zgf.service.impl;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.ModelRequestLog;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.ModelRequestLogRepository;
+import com.zhpb.gdyd_zhpb_zgf.service.ModelRequestLogService;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.stereotype.Service;
+
+@Service
+public class ModelRequestLogServiceImpl implements ModelRequestLogService {
+
+    private static final Logger logger = LoggerFactory.getLogger(ModelRequestLogServiceImpl.class);
+
+    private final ModelRequestLogRepository modelRequestLogRepository;
+
+    public ModelRequestLogServiceImpl(ModelRequestLogRepository modelRequestLogRepository) {
+        this.modelRequestLogRepository = modelRequestLogRepository;
+    }
+
+    @Override
+    public void record(ModelRequestLog log) {
+        if (log == null) {
+            return;
+        }
+        try {
+            if (log.getRequestTime() == null) {
+                log.setRequestTime(java.time.LocalDateTime.now());
+            }
+            modelRequestLogRepository.save(log);
+        } catch (Exception e) {
+            logger.warn("记录模型请求日志失败: {}", e.getMessage());
+        }
+    }
+}
+
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/OfficeGatewayServiceImpl.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/OfficeGatewayServiceImpl.java
new file mode 100644
index 0000000..03ff8e4
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/OfficeGatewayServiceImpl.java
@@ -0,0 +1,365 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.service.impl
+ * @ClassName: OfficeGatewayServiceImpl
+ * @Description: 办公网关服务实现（磐智宁波池）
+ * @Author: 张志锋
+ * @Date: 2025-11-05
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.service.impl;
+
+import com.alibaba.fastjson2.JSON;
+import com.alibaba.fastjson2.JSONArray;
+import com.alibaba.fastjson2.JSONObject;
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.ModelRequestLog;
+import com.zhpb.gdyd_zhpb_zgf.service.ModelRequestLogService;
+import com.zhpb.gdyd_zhpb_zgf.utils.HttpClientUtils;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.beans.factory.annotation.Value;
+import org.springframework.stereotype.Service;
+
+import java.nio.charset.StandardCharsets;
+import java.security.MessageDigest;
+import java.security.NoSuchAlgorithmException;
+import java.util.*;
+import java.util.concurrent.CompletableFuture;
+import java.util.Base64;
+import java.util.UUID;
+
+/**
+ * 办公网关服务实现
+ * 实现磐智宁波池的模型网关通用协议
+ */
+@Service
+public class OfficeGatewayServiceImpl {
+
+    private static final Logger logger = LoggerFactory.getLogger(OfficeGatewayServiceImpl.class);
+
+    @Value("${office-gateway.uat-url}")
+    private String uatUrl;
+
+    @Value("${office-gateway.prod-url:http://prod-gateway-url}")
+    private String prodUrl;
+
+    @Value("${office-gateway.appid}")
+    private String appId;
+
+    @Value("${office-gateway.appkey}")
+    private String appKey;
+
+    @Value("${office-gateway.capabilityname:semantic0000000000000000}")
+    private String capabilityName;
+
+    @Value("${office-gateway.embedding-model}")
+    private String embeddingModel;
+
+    @Value("${office-gateway.timeout:30000}")
+    private int timeout;
+
+    @Value("${vectorization.mode:local}")
+    private String vectorizationMode;
+
+    @Autowired(required = false)
+    private ModelRequestLogService modelRequestLogService;
+
+    /**
+     * 对二进制数据进行base64编码
+     */
+    private String encodeBase64(byte[] rawBytes) {
+        return Base64.getEncoder().encodeToString(rawBytes);
+    }
+
+    /**
+     * 计算MD5哈希值
+     */
+    private String calculateMd5(String rawText) {
+        try {
+            MessageDigest md = MessageDigest.getInstance("MD5");
+            byte[] hash = md.digest(rawText.getBytes(StandardCharsets.UTF_8));
+            StringBuilder sb = new StringBuilder();
+            for (byte b : hash) {
+                sb.append(String.format("%02x", b));
+            }
+            return sb.toString();
+        } catch (NoSuchAlgorithmException e) {
+            throw new RuntimeException("MD5算法不可用", e);
+        }
+    }
+
+    /**
+     * 确保capabilityname为24位，不足补0
+     */
+    private String getCapabilityName24(String rawCapabilityName) {
+        if (rawCapabilityName == null) {
+            rawCapabilityName = "semantic0000000000000000";
+        }
+        List<Character> capabilityNameList = new ArrayList<>();
+        for (char c : rawCapabilityName.toCharArray()) {
+            capabilityNameList.add(c);
+        }
+        while (capabilityNameList.size() < 24) {
+            capabilityNameList.add('0');
+        }
+        StringBuilder sb = new StringBuilder();
+        for (Character c : capabilityNameList) {
+            sb.append(c);
+        }
+        return sb.toString();
+    }
+
+    /**
+     * 生成包含授权信息的请求头
+     */
+    private Map<String, String> generateAuthHeaders() {
+        // 生成唯一标识符
+        String uuid = UUID.randomUUID().toString().replace("-", "");
+
+        // 确保capabilityName为24位
+        String capabilityName24 = getCapabilityName24(capabilityName);
+
+        // 构建并编码X-Server-Param
+        Map<String, String> xServerParam = new HashMap<>();
+        xServerParam.put("appid", appId);
+        xServerParam.put("csid", appId + capabilityName24 + uuid);
+
+        String xServerParamJson = JSON.toJSONString(xServerParam);
+        String xServerParamEncoded = encodeBase64(xServerParamJson.getBytes(StandardCharsets.UTF_8));
+
+        // 生成时间戳
+        String xCurTime = String.valueOf(System.currentTimeMillis() / 1000);
+
+        // 计算校验和
+        String xCheckSum = calculateMd5(appKey + xCurTime + xServerParamEncoded);
+
+        // 构建请求头
+        Map<String, String> headers = new HashMap<>();
+        headers.put("Content-Type", "application/json");
+        headers.put("Accept", "application/json");
+        headers.put("X-Server-Param", xServerParamEncoded);
+        headers.put("X-CurTime", xCurTime);
+        headers.put("X-CheckSum", xCheckSum);
+        headers.put("traceId", "smart-qa-" + System.currentTimeMillis());
+
+        return headers;
+    }
+
+    /**
+     * 获取当前使用的URL（UAT或生产）
+     */
+    private String getCurrentUrl() {
+        // 可以根据环境变量或其他配置来决定使用UAT还是生产环境
+        String env = System.getProperty("spring.profiles.active", "default");
+        if ("prod".equals(env) || "production".equals(env)) {
+            return prodUrl;
+        }
+        return uatUrl;
+    }
+
+    /**
+     * 调用Embedding模型进行文本向量化
+     *
+     * @param texts 待向量化的文本列表
+     * @return 向量结果列表
+     */
+    public List<float[]> embedTexts(List<String> texts) {
+        if (!"office".equals(vectorizationMode)) {
+            throw new IllegalStateException("当前向量化模式不是办公网关模式");
+        }
+
+        try {
+            logger.info("调用办公网关Embedding模型，文本数量: {}", texts.size());
+
+            // 构建请求体
+            Map<String, Object> requestBody = new HashMap<>();
+            requestBody.put("model_type", "embedding");
+            requestBody.put("model_name", embeddingModel);
+
+            Map<String, Object> params = new HashMap<>();
+            params.put("texts", texts);
+            requestBody.put("params", params);
+
+            // 发送请求
+            String responseStr = callGatewayApi(JSON.toJSONString(requestBody),
+                    "VECTORIZE_OFFICE", embeddingModel);
+            JSONObject response = JSON.parseObject(responseStr);
+
+            // 检查响应
+            String code = response.getString("code");
+            if (!"000000".equals(code)) {
+                String desc = response.getString("desc");
+                throw new RuntimeException("办公网关API调用失败: " + code + " - " + desc);
+            }
+
+            // 解析结果
+            JSONObject result = response.getJSONObject("result");
+            JSONArray embeddings = result.getJSONArray("embeddings");
+
+            List<float[]> vectors = new ArrayList<>();
+            for (int i = 0; i < embeddings.size(); i++) {
+                JSONObject embedding = embeddings.getJSONObject(i);
+                JSONArray embeddingArray = embedding.getJSONArray("embedding");
+
+                float[] vector = new float[embeddingArray.size()];
+                for (int j = 0; j < embeddingArray.size(); j++) {
+                    vector[j] = embeddingArray.getFloat(j);
+                }
+                vectors.add(vector);
+            }
+
+            logger.info("办公网关Embedding成功，返回向量数量: {}", vectors.size());
+            return vectors;
+
+        } catch (Exception e) {
+            logger.error("调用办公网关Embedding失败", e);
+            throw new RuntimeException("办公网关Embedding调用失败: " + e.getMessage());
+        }
+    }
+
+    /**
+     * 调用Reranker模型进行文档重排序
+     *
+     * @param query 查询文本
+     * @param documents 候选文档列表，每个文档包含id和content
+     * @param topK 返回前K个结果，可选
+     * @return 重排序后的文档ID和分数
+     */
+    public List<Map<String, Object>> rerankDocuments(String query, List<Map<String, Object>> documents, Integer topK) {
+        try {
+            logger.info("调用办公网关Reranker模型，文档数量: {}", documents.size());
+
+            // 构建请求体
+            Map<String, Object> requestBody = new HashMap<>();
+            requestBody.put("model_type", "reranker");
+            requestBody.put("model_name", "rerank-bge-v2-m3"); // 使用默认的reranker模型
+
+            Map<String, Object> params = new HashMap<>();
+            params.put("query", query);
+            params.put("documents", documents);
+            if (topK != null && topK > 0) {
+                params.put("top_k", topK);
+            }
+            requestBody.put("params", params);
+
+            // 发送请求
+            String responseStr = callGatewayApi(JSON.toJSONString(requestBody),
+                    "OFFICE_RERANK", "rerank-bge-v2-m3");
+            JSONObject response = JSON.parseObject(responseStr);
+
+            // 检查响应
+            String code = response.getString("code");
+            if (!"000000".equals(code)) {
+                String desc = response.getString("desc");
+                throw new RuntimeException("办公网关Reranker API调用失败: " + code + " - " + desc);
+            }
+
+            // 解析结果
+            JSONObject result = response.getJSONObject("result");
+            JSONArray results = result.getJSONArray("results");
+
+            List<Map<String, Object>> rankedResults = new ArrayList<>();
+            for (int i = 0; i < results.size(); i++) {
+                JSONObject item = results.getJSONObject(i);
+                Map<String, Object> rankedItem = new HashMap<>();
+                rankedItem.put("id", item.getString("id"));
+                rankedItem.put("score", item.getDouble("score"));
+                rankedResults.add(rankedItem);
+            }
+
+            logger.info("办公网关Reranker成功，返回结果数量: {}", rankedResults.size());
+            return rankedResults;
+
+        } catch (Exception e) {
+            logger.error("调用办公网关Reranker失败", e);
+            throw new RuntimeException("办公网关Reranker调用失败: " + e.getMessage());
+        }
+    }
+
+    /**
+     * 调用办公网关API
+     */
+    private String callGatewayApi(String requestBody, String requestSource, String modelName) throws Exception {
+        String url = getCurrentUrl();
+        logger.debug("调用办公网关API: {} , 请求体长度: {}", url, requestBody.length());
+
+        Map<String, String> headers = generateAuthHeaders();
+        long startTime = System.currentTimeMillis();
+        String response = null;
+        boolean success = false;
+        String errorMsg = null;
+        try {
+            response = HttpClientUtils.postJson(url, requestBody, headers);
+            if (response == null || response.trim().isEmpty()) {
+                throw new RuntimeException("办公网关API无响应");
+            }
+            logger.debug("办公网关API响应长度: {}", response.length());
+            success = true;
+            return response;
+        } catch (Exception e) {
+            errorMsg = e.getMessage();
+            throw e;
+        } finally {
+            recordGatewayModelLog(
+                requestSource,
+                modelName,
+                url,
+                requestBody,
+                response,
+                System.currentTimeMillis() - startTime,
+                success,
+                errorMsg,
+                headers
+            );
+        }
+    }
+
+    private void recordGatewayModelLog(String requestSource,
+                                       String modelName,
+                                       String requestUrl,
+                                       String requestBody,
+                                       String responseBody,
+                                       Long durationMs,
+                                       boolean success,
+                                       String errorMessage,
+                                       Map<String, String> headers) {
+        if (modelRequestLogService == null) {
+            return;
+        }
+        ModelRequestLog logEntry = new ModelRequestLog();
+        logEntry.setRequestSource(requestSource);
+        logEntry.setRequestUrl(requestUrl);
+        logEntry.setRequestMethod("POST");
+        logEntry.setModelName(modelName);
+        logEntry.setRequestPayload(requestBody);
+        logEntry.setResponsePayload(responseBody);
+        logEntry.setStatus(success ? "SUCCESS" : "FAILED");
+        logEntry.setDurationMs(durationMs);
+        logEntry.setAppKey(appKey);
+        logEntry.setAppSecret(capabilityName);
+        logEntry.setErrorMessage(errorMessage);
+        if (headers != null) {
+            logEntry.setExtraInfo(String.format("X-Server-Param=%s;X-CurTime=%s;X-CheckSum=%s",
+                headers.get("X-Server-Param"),
+                headers.get("X-CurTime"),
+                headers.get("X-CheckSum")));
+        }
+        modelRequestLogService.record(logEntry);
+    }
+
+    /**
+     * 批量向量化文本（支持异步）
+     */
+    public CompletableFuture<List<float[]>> embedTextsAsync(List<String> texts) {
+        return CompletableFuture.supplyAsync(() -> embedTexts(texts));
+    }
+
+    /**
+     * 获取当前配置的嵌入模型名称
+     */
+    public String getEmbeddingModel() {
+        return embeddingModel;
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/PdfReadServiceImpl.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/PdfReadServiceImpl.java
new file mode 100644
index 0000000..1711c22
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/PdfReadServiceImpl.java
@@ -0,0 +1,577 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.service.impl
+ * @ClassName: PdfReadServiceImpl
+ * @Description: PDF文件读取服务实现类，使用Apache PDFBox异步读取PDF文件内容并输出到控制台
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.service.impl;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.FileType;
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.Supplier;
+import com.zhpb.gdyd_zhpb_zgf.entity.pgsql.PdfVector;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.FileRepository;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.FileTypeRepository;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.SupplierRepository;
+import com.zhpb.gdyd_zhpb_zgf.repository.pgsql.PdfVectorRepository;
+import com.zhpb.gdyd_zhpb_zgf.service.FileTypeService;
+import com.zhpb.gdyd_zhpb_zgf.service.FilePathResolver;
+import com.zhpb.gdyd_zhpb_zgf.service.PdfReadService;
+import com.zhpb.gdyd_zhpb_zgf.service.VectorizationService;
+import com.zhpb.gdyd_zhpb_zgf.utils.FilePathOrganizer;
+import jakarta.annotation.PostConstruct;
+import jakarta.annotation.PreDestroy;
+import jakarta.persistence.EntityManager;
+import jakarta.persistence.EntityManagerFactory;
+import org.apache.pdfbox.Loader;
+import org.apache.pdfbox.pdmodel.PDDocument;
+import org.apache.pdfbox.text.PDFTextStripper;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.beans.factory.annotation.Value;
+import org.springframework.scheduling.annotation.Async;
+import org.springframework.stereotype.Service;
+import org.springframework.transaction.annotation.Transactional;
+
+import java.io.IOException;
+import java.nio.file.Files;
+import java.nio.file.Path;
+import java.nio.file.Paths;
+import java.util.List;
+
+/**
+ * PDF文件读取服务实现类
+ */
+@Service
+public class PdfReadServiceImpl implements PdfReadService {
+
+    private static final Logger logger = LoggerFactory.getLogger(PdfReadServiceImpl.class);
+
+    @Autowired
+    private PdfVectorRepository pdfVectorRepository;
+
+    @Autowired
+    private FileRepository fileRepository;
+
+    @Autowired
+    private SupplierRepository supplierRepository;
+
+    @Autowired
+    private FileTypeRepository fileTypeRepository;
+
+    @Autowired
+    private FileTypeService fileTypeService;
+
+    @Autowired
+    private FilePathResolver filePathResolver;
+
+    @Autowired
+    private VectorizationService vectorizationService;
+
+    @Autowired
+    private EntityManagerFactory entityManagerFactory;
+    
+    private EntityManager getEntityManager() {
+        return entityManagerFactory.createEntityManager();
+    }
+
+    @Value("${vectorize.base-url:http://localhost:8001}")
+    private String vectorizeBaseUrl;
+
+    @Value("${vectorize.api-path:/vectorize}")
+    private String vectorizeApiPath;
+
+    @Value("${vectorize.dimension:896}")
+    private Integer defaultDimension;
+
+    @Value("${vectorize.concurrency:2}")
+    private int concurrency;
+
+    private java.util.concurrent.ExecutorService executorService;
+
+    @PostConstruct
+    public void init() {
+        // 确保并发数至少为1
+        int nThreads = Math.max(1, concurrency);
+        logger.info("初始化向量化线程池，并发数: {}", nThreads);
+        this.executorService = java.util.concurrent.Executors.newFixedThreadPool(nThreads);
+    }
+
+    @PreDestroy
+    public void destroy() {
+        if (executorService != null) {
+            executorService.shutdown();
+        }
+    }
+
+    @Override
+    @Async
+    public void readPdfFilesAsync(List<String> pdfFilePaths, String projectId) {
+        long startTime = System.currentTimeMillis();
+        logger.info("[文本解析] ========== 开始处理PDF文件，共 {} 个，项目ID: {} ==========", pdfFilePaths.size(), projectId);
+        
+        // 更新文件向量化状态为处理中
+        updateFilesVectorizationStatus(pdfFilePaths, "processing");
+        
+        // 使用原子计数器确保线程安全
+        java.util.concurrent.atomic.AtomicInteger successCount = new java.util.concurrent.atomic.AtomicInteger(0);
+        java.util.concurrent.atomic.AtomicInteger failCount = new java.util.concurrent.atomic.AtomicInteger(0);
+        java.util.concurrent.atomic.AtomicInteger vectorizeSuccessCount = new java.util.concurrent.atomic.AtomicInteger(0);
+        java.util.concurrent.atomic.AtomicInteger vectorizeFailCount = new java.util.concurrent.atomic.AtomicInteger(0);
+        
+        List<java.util.concurrent.CompletableFuture<Void>> futures = new java.util.ArrayList<>();
+
+        for (String pdfFilePath : pdfFilePaths) {
+            java.util.concurrent.CompletableFuture<Void> future = java.util.concurrent.CompletableFuture.runAsync(() -> {
+                try {
+                    Path path = Paths.get(pdfFilePath);
+                    
+                    // 检查文件是否存在
+                    if (!Files.exists(path)) {
+                        logger.warn("PDF文件不存在: {}", pdfFilePath);
+                        failCount.incrementAndGet();
+                        return;
+                    }
+                    
+                    // 检查文件扩展名
+                    String fileName = path.getFileName().toString();
+                    if (!fileName.toLowerCase().endsWith(".pdf")) {
+                        logger.debug("跳过非PDF文件: {}", pdfFilePath);
+                        return;
+                    }
+                    
+                    // 读取PDF内容
+                    String content = readPdfContent(path);
+                    
+                    // 表格提取功能已禁用（消耗大量时间）
+                    // 如需启用，取消以下注释：
+                    // try {
+                    //     content = com.zhpb.gdyd_zhpb_zgf.utils.PdfTableExtractor.enhanceTextWithTables(pdfFilePath, content);
+                    //     logger.debug("PDF表格增强完成: {}", fileName);
+                    // } catch (Exception e) {
+                    //     logger.warn("PDF表格提取失败，使用原始文本: {}", e.getMessage());
+                    // }
+                    
+                    // 输出到控制台（使用System.out和logger双重输出）
+                    // 为了避免多线程控制台输出混乱，使用同步块
+                    synchronized (System.out) {
+                        System.out.println("\n" + "=".repeat(80));
+                        System.out.println("PDF文件内容 [" + fileName + "]");
+                        System.out.println("=".repeat(80));
+                        System.out.println("文件路径: " + pdfFilePath);
+                        System.out.println("内容长度: " + content.length() + " 字符");
+                        System.out.println("-".repeat(80));
+                        System.out.println("内容预览（前500字符）:");
+                        if (content.length() > 500) {
+                            System.out.println(content.substring(0, 500) + "...");
+                        } else {
+                            System.out.println(content);
+                        }
+                        System.out.println("-".repeat(80));
+                        System.out.println("=".repeat(80));
+                        System.out.println("PDF文件内容结束 [" + fileName + "]");
+                        System.out.println("=".repeat(80) + "\n");
+                    }
+                    
+                    // 同时记录到日志
+                    logger.info("========== PDF文件内容 [{}] ==========", fileName);
+                    logger.info("文件路径: {}", pdfFilePath);
+                    logger.info("内容长度: {} 字符", content.length());
+                    
+                    // 进行向量化并保存到PostgreSQL
+                    try {
+                        vectorizeAndSave(pdfFilePath, fileName, content, projectId);
+                        vectorizeSuccessCount.incrementAndGet();
+                        logger.info("PDF文件向量化并保存成功: {}", fileName);
+                    } catch (Exception e) {
+                        vectorizeFailCount.incrementAndGet();
+                        String errorType = "未知错误";
+                        String errorDetail = e.getMessage();
+                        
+                        // 识别错误类型
+                        if (e.getMessage() != null) {
+                            if (e.getMessage().contains("无法连接到向量化服务") || e.getMessage().contains("Connection refused")) {
+                                errorType = "向量化服务连接失败";
+                            } else if (e.getMessage().contains("超时") || e.getMessage().contains("timeout")) {
+                                errorType = "向量化接口超时";
+                            } else if (e.getMessage().contains("状态码")) {
+                                errorType = "向量化接口返回错误";
+                            } else if (e.getMessage().contains("返回格式错误")) {
+                                errorType = "向量化接口返回格式错误";
+                            } else if (e.getMessage().contains("网络IO异常")) {
+                                errorType = "网络IO异常";
+                            }
+                        }
+                        
+                        logger.error("PDF文件向量化失败 [{}]: 文件={}, 错误类型={}, 错误详情={}", 
+                                errorType, fileName, errorType, errorDetail, e);
+                        
+                        // 更新文件状态为失败
+                        updateFileVectorizationStatusFailed(pdfFilePath, errorType, errorDetail);
+                    }
+                    
+                    successCount.incrementAndGet();
+                    
+                } catch (Exception e) {
+                    logger.error("读取PDF文件失败: {}, 错误: {}", pdfFilePath, e.getMessage(), e);
+                    failCount.incrementAndGet();
+                }
+            }, executorService);
+            
+            futures.add(future);
+        }
+        
+        // 等待所有任务完成
+        try {
+            java.util.concurrent.CompletableFuture.allOf(futures.toArray(new java.util.concurrent.CompletableFuture[0])).join();
+        } catch (Exception e) {
+            logger.error("并行处理PDF文件时发生异常", e);
+        }
+        
+        long duration = System.currentTimeMillis() - startTime;
+        logger.info("[文本解析] ========== 完成! 耗时: {}ms，成功: {}个，失败: {}个，向量化成功: {}个，向量化失败: {}个 ==========", 
+                duration, successCount.get(), failCount.get(), vectorizeSuccessCount.get(), vectorizeFailCount.get());
+    }
+
+    /**
+     * 向量化PDF内容并保存到PostgreSQL
+     *
+     * @param filePath 文件路径
+     * @param fileName 文件名
+     * @param content PDF文本内容
+     * @param projectId 项目ID
+     */
+    // 移除 @Transactional，因为我们手动管理 EntityManager 事务，且在多线程环境下私有方法的事务注解无效
+    private void vectorizeAndSave(String filePath, String fileName, String content, String projectId) {
+        try {
+            // 检查是否已经向量化过
+            String storedPath = normalizeStoredPath(filePath);
+            PdfVector existingVector = pdfVectorRepository.findByFilePath(storedPath);
+            if (existingVector != null) {
+                logger.debug("PDF文件已存在向量化记录，跳过: {}", storedPath);
+                return;
+            }
+
+            PdfVector legacyVector = pdfVectorRepository.findByFilePath(filePath);
+            if (legacyVector != null) {
+                legacyVector.setFilePath(storedPath);
+                pdfVectorRepository.save(legacyVector);
+                logger.debug("更新历史向量记录路径: {}", storedPath);
+                return;
+            }
+
+            // 调用向量化服务
+            logger.info("开始向量化PDF文件: {}", fileName);
+            float[] vector = vectorizationService.vectorizeText(content);
+            Integer dimension = vectorizationService.getDimension();
+            
+            // 构建向量字符串（pgvector格式：[0.1,0.2,0.3,...]）
+            StringBuilder vectorStr = new StringBuilder("[");
+            for (int i = 0; i < vector.length; i++) {
+                if (i > 0) {
+                    vectorStr.append(",");
+                }
+                vectorStr.append(vector[i]);
+            }
+            vectorStr.append("]");
+
+            // 保存部分内容作为摘要（如果内容太长，只保存前10000字符）
+            String contentSummary = content.length() > 10000 ? content.substring(0, 10000) + "..." : content;
+
+            // 首先检查表是否存在，如果不存在则创建（包含vector类型的列）
+            EntityManager em = getEntityManager();
+            try {
+                em.getTransaction().begin();
+                
+                // 检查表是否存在
+                String checkTableSql = "SELECT EXISTS (SELECT FROM information_schema.tables WHERE table_name = 't_pdf_vector')";
+                Boolean tableExists = (Boolean) em.createNativeQuery(checkTableSql).getSingleResult();
+                
+                if (!tableExists) {
+                    // 创建表，vector列使用vector类型，包含file_id字段
+                    String createTableSql = "CREATE TABLE t_pdf_vector (" +
+                            "id BIGSERIAL PRIMARY KEY, " +
+                            "file_path VARCHAR(1000) NOT NULL, " +
+                            "file_name VARCHAR(500) NOT NULL, " +
+                            "content TEXT, " +
+                            "vector vector, " +
+                            "dimension INTEGER NOT NULL, " +
+                            "file_id VARCHAR(64), " +
+                            "project_id VARCHAR(100), " +
+                            "supplier_id VARCHAR(64), " +
+                            "file_type_id BIGINT, " +
+                            "create_time TIMESTAMP(6) NOT NULL DEFAULT NOW(), " +
+                            "update_time TIMESTAMP(6) DEFAULT NOW()" +
+                            ")";
+                    em.createNativeQuery(createTableSql).executeUpdate();
+                    logger.info("创建t_pdf_vector表成功（包含vector类型列和file_id字段）");
+                }
+                
+                // 查找文件ID（关联文件表）
+                String fileId = null;
+                String supplierId = null;
+                Long fileTypeId = null;
+
+                try {
+                    java.util.Optional<com.zhpb.gdyd_zhpb_zgf.entity.mysql.File> fileOpt = fileRepository.findByFilePath(storedPath);
+                    if (fileOpt.isPresent()) {
+                        com.zhpb.gdyd_zhpb_zgf.entity.mysql.File file = fileOpt.get();
+                        fileId = file.getFileId();
+                        supplierId = file.getSupplierId();
+                        fileTypeId = file.getFileTypeId();
+
+                        logger.debug("从数据库获取文件信息: fileId={}, supplierId={}, fileTypeId={}",
+                                   fileId, supplierId, fileTypeId);
+                    }
+                } catch (Exception e) {
+                    logger.warn("查找文件ID失败: {}", e.getMessage());
+                }
+
+                // 使用原生SQL插入向量数据（包含file_id、supplier_id、file_type_id）
+                String insertSql = "INSERT INTO t_pdf_vector (file_path, file_name, content, vector, dimension, file_id, project_id, supplier_id, file_type_id, create_time, update_time) " +
+                        "VALUES (:filePath, :fileName, :content, CAST(:vectorStr AS vector), :dimension, :fileId, :projectId, :supplierId, :fileTypeId, NOW(), NOW())";
+                
+                em.createNativeQuery(insertSql)
+                        .setParameter("filePath", storedPath)
+                        .setParameter("fileName", fileName)
+                        .setParameter("content", contentSummary)
+                        .setParameter("vectorStr", vectorStr.toString())
+                        .setParameter("dimension", dimension)
+                        .setParameter("fileId", fileId)
+                        .setParameter("projectId", projectId)
+                        .setParameter("supplierId", supplierId)
+                        .setParameter("fileTypeId", fileTypeId)
+                        .executeUpdate();
+                        
+                em.getTransaction().commit();
+                
+                // 更新文件的向量化状态和数量
+                if (fileId != null) {
+                    updateFileVectorizationStatus(fileId);
+                }
+            } catch (Exception e) {
+                em.getTransaction().rollback();
+                throw e;
+            } finally {
+                em.close();
+            }
+            
+            logger.info("PDF向量化并保存成功，文件: {}，维度: {}", fileName, dimension);
+            
+        } catch (Exception e) {
+            logger.error("向量化并保存PDF失败: {}, 错误: {}", fileName, e.getMessage(), e);
+            throw new RuntimeException("向量化并保存PDF失败: " + e.getMessage(), e);
+        }
+    }
+
+    /**
+     * 读取PDF文件内容
+     *
+     * @param pdfPath PDF文件路径
+     * @return PDF文本内容
+     * @throws IOException 读取异常
+     */
+    private String readPdfContent(Path pdfPath) throws IOException {
+        java.io.File file = pdfPath.toFile();
+
+        // 检查文件是否存在
+        if (!file.exists()) {
+            throw new IOException("PDF文件不存在: " + pdfPath);
+        }
+
+        // 检查文件大小
+        if (file.length() == 0) {
+            throw new IOException("PDF文件为空文件: " + pdfPath);
+        }
+
+        // 检查文件是否过小（PDF文件最小应该有几个KB）
+        if (file.length() < 1024) {
+            throw new IOException("PDF文件过小，可能已损坏，文件大小: " + file.length() + " bytes, 路径: " + pdfPath);
+        }
+
+        // 检查PDF文件头是否正确
+        try (java.io.RandomAccessFile raf = new java.io.RandomAccessFile(file, "r")) {
+            byte[] header = new byte[8];
+            int bytesRead = raf.read(header);
+            if (bytesRead < 8) {
+                throw new IOException("PDF文件头部不完整，读取到 " + bytesRead + " 字节，路径: " + pdfPath);
+            }
+
+            String headerStr = new String(header, "ISO-8859-1");
+            if (!headerStr.startsWith("%PDF-")) {
+                throw new IOException("无效的PDF文件头部: " + headerStr + ", 路径: " + pdfPath);
+            }
+
+            logger.debug("PDF文件头部验证通过: {}", headerStr);
+        }
+
+        // 设置PDFBox配置以更好地处理损坏的文件
+        java.io.FileInputStream fis = null;
+        PDDocument document = null;
+
+        try {
+            fis = new java.io.FileInputStream(file);
+
+            // 使用更宽松的加载方式
+            org.apache.pdfbox.io.RandomAccessReadBuffer buffer = new org.apache.pdfbox.io.RandomAccessReadBuffer(fis);
+            document = Loader.loadPDF(buffer);
+
+            // 检查文档是否为空
+            if (document.getNumberOfPages() == 0) {
+                throw new IOException("PDF文档页数为0，可能已损坏: " + pdfPath);
+            }
+
+            PDFTextStripper stripper = new PDFTextStripper();
+
+            // 设置文本提取参数
+            stripper.setStartPage(1);
+            stripper.setEndPage(Math.min(document.getNumberOfPages(), 100)); // 限制最大页数，避免处理过大的文档
+            stripper.setSortByPosition(true); // 按位置排序文本
+
+            String text = stripper.getText(document);
+
+            // 检查提取的文本是否为空或过短
+            if (text == null || text.trim().isEmpty()) {
+                logger.warn("PDF文件未提取到有效文本内容: {}", pdfPath);
+                text = "[PDF文件内容为空或无法提取文本]";
+            } else if (text.trim().length() < 10) {
+                logger.warn("PDF文件提取的文本过短，可能有问题: {}，内容: {}", pdfPath, text.trim());
+            }
+
+            logger.info("成功读取PDF文件: {}，共 {} 页，内容长度: {} 字符",
+                    pdfPath.getFileName(), document.getNumberOfPages(), text.length());
+
+            return text;
+
+        } catch (RuntimeException e) {
+            // PDFBox特定的运行时异常，通常表示文件损坏
+            // 检查是否为PDFBox相关的异常
+            if (e.getClass().getName().contains("pdfbox") ||
+                e.getMessage() != null && e.getMessage().contains("COS")) {
+                logger.error("PDF文件结构损坏，无法读取: {}，错误: {}", pdfPath, e.getMessage());
+                throw new IOException("PDF文件结构损坏: " + e.getMessage() + ", 路径: " + pdfPath, e);
+            } else {
+                throw e; // 重新抛出非PDFBox异常
+            }
+        } catch (Exception e) {
+            // 其他异常
+            logger.error("读取PDF文件失败: {}，错误: {}", pdfPath, e.getMessage());
+            if (e instanceof IOException) {
+                throw e;
+            } else {
+                throw new IOException("读取PDF文件失败: " + e.getMessage() + ", 路径: " + pdfPath, e);
+            }
+        } finally {
+            // 确保资源被正确释放
+            if (document != null) {
+                try {
+                    document.close();
+                } catch (Exception e) {
+                    logger.warn("关闭PDF文档失败: {}", e.getMessage());
+                }
+            }
+            if (fis != null) {
+                try {
+                    fis.close();
+                } catch (Exception e) {
+                    logger.warn("关闭文件输入流失败: {}", e.getMessage());
+                }
+            }
+        }
+    }
+
+
+    /**
+     * 批量更新文件向量化状态为处理中
+     */
+    private void updateFilesVectorizationStatus(List<String> filePaths, String status) {
+        try {
+            for (String filePath : filePaths) {
+                String storedPath = normalizeStoredPath(filePath);
+                java.util.Optional<com.zhpb.gdyd_zhpb_zgf.entity.mysql.File> fileOpt = fileRepository.findByFilePath(storedPath);
+                if (fileOpt.isPresent()) {
+                    com.zhpb.gdyd_zhpb_zgf.entity.mysql.File file = fileOpt.get();
+                    file.setVectorizationStatus(status);
+                    fileRepository.save(file);
+                }
+            }
+        } catch (Exception e) {
+            logger.warn("批量更新文件向量化状态失败: {}", e.getMessage());
+        }
+    }
+
+    private void updateFileVectorizationStatusFailed(String filePath, String errorType, String errorDetail) {
+        try {
+            String storedPath = normalizeStoredPath(filePath);
+            java.util.Optional<com.zhpb.gdyd_zhpb_zgf.entity.mysql.File> fileOpt = fileRepository.findByFilePath(storedPath);
+            if (fileOpt.isPresent()) {
+                com.zhpb.gdyd_zhpb_zgf.entity.mysql.File file = fileOpt.get();
+                file.setVectorizationStatus("failed");
+                file.setStatus("vectorization_failed");
+                
+                // 如果提供了错误信息，更新描述字段
+                if (errorType != null && errorDetail != null) {
+                    String description = String.format("向量化失败: %s - %s", errorType, 
+                            errorDetail.length() > 200 ? errorDetail.substring(0, 200) + "..." : errorDetail);
+                    file.setDescription(description);
+                }
+                
+                fileRepository.save(file);
+                logger.debug("文件向量化状态已更新为失败: {}, 错误类型: {}", storedPath, errorType);
+            }
+        } catch (Exception e) {
+            logger.warn("更新文件向量化失败状态失败: {}", e.getMessage());
+        }
+    }
+
+    /**
+     * 更新文件的向量化状态和数量
+     */
+    private void updateFileVectorizationStatus(String fileId) {
+        try {
+            java.util.Optional<com.zhpb.gdyd_zhpb_zgf.entity.mysql.File> fileOpt = fileRepository.findByFileId(fileId);
+            if (fileOpt.isPresent()) {
+                com.zhpb.gdyd_zhpb_zgf.entity.mysql.File file = fileOpt.get();
+                
+                // 统计该文件对应的向量化记录数量
+                long vectorCount = pdfVectorRepository.countByFileId(fileId);
+                
+                // 更新向量化状态和数量
+                file.setVectorizationCount((int) vectorCount);
+                if (vectorCount > 0) {
+                    file.setIsVectorized(true);
+                    file.setVectorizationStatus("completed");
+                    file.setVectorizedTime(java.time.LocalDateTime.now());
+                } else {
+                    file.setIsVectorized(false);
+                    file.setVectorizationStatus("not_started");
+                }
+                
+                file.setStatus("vectorized");
+                fileRepository.save(file);
+                
+                logger.debug("更新文件向量化状态: fileId={}, vectorizationCount={}, status={}", 
+                        fileId, vectorCount, file.getVectorizationStatus());
+            }
+        } catch (Exception e) {
+            logger.warn("更新文件向量化状态失败: {}", e.getMessage());
+        }
+    }
+
+    private String normalizeStoredPath(String filePath) {
+        if (filePath == null || filePath.isBlank()) {
+            return filePath;
+        }
+        try {
+            Path path = Paths.get(filePath).toAbsolutePath().normalize();
+            return filePathResolver.normalizeForStorage(path);
+        } catch (Exception e) {
+            return filePath;
+        }
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/PrivateModelServiceImpl.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/PrivateModelServiceImpl.java
new file mode 100644
index 0000000..46a610c
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/PrivateModelServiceImpl.java
@@ -0,0 +1,242 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.service.impl
+ * @ClassName: PrivateModelServiceImpl
+ * @Description: 私有大模型服务实现
+ * @Author: 张志锋
+ * @Date: 2025-11-05
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.service.impl;
+
+import com.alibaba.fastjson2.JSON;
+import com.alibaba.fastjson2.JSONObject;
+import com.zhpb.gdyd_zhpb_zgf.context.ModelRequestContextHolder;
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.ModelRequestLog;
+import com.zhpb.gdyd_zhpb_zgf.service.AIModelService;
+import com.zhpb.gdyd_zhpb_zgf.service.ModelRequestLogService;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.beans.factory.annotation.Value;
+import org.springframework.http.*;
+import org.springframework.stereotype.Service;
+import org.springframework.web.client.RestTemplate;
+import org.springframework.web.client.HttpStatusCodeException;
+
+import java.nio.charset.StandardCharsets;
+import java.security.MessageDigest;
+import java.security.NoSuchAlgorithmException;
+import java.util.*;
+
+/**
+ * 私有大模型服务实现
+ */
+@Service
+public class PrivateModelServiceImpl implements AIModelService {
+
+    private static final Logger logger = LoggerFactory.getLogger(PrivateModelServiceImpl.class);
+
+    @Value("${private-model.url}")
+    private String apiUrl;
+
+    @Value("${private-model.appid}")
+    private String appId;
+
+    @Value("${private-model.appkey}")
+    private String appKey;
+
+    @Value("${private-model.capabilityname:semantic0000000000000000}")
+    private String capabilityName;
+
+    @Value("${private-model.model:JiuTian-75B-8K}")
+    private String model;
+
+    @Value("${private-model.temperature:0.1}")
+    private double temperature;
+
+    // max_tokens不再限制，让模型自由决定输出长度
+
+    private final RestTemplate restTemplate;
+    private final ModelRequestLogService modelRequestLogService;
+
+    /**
+     * 创建带超时配置的RestTemplate
+     */
+    private static RestTemplate createRestTemplateWithTimeout() {
+        org.springframework.http.client.SimpleClientHttpRequestFactory factory = 
+            new org.springframework.http.client.SimpleClientHttpRequestFactory();
+        factory.setConnectTimeout(30000);  // 连接超时30秒
+        factory.setReadTimeout(120000);    // 读取超时120秒（AI模型响应可能较慢）
+        return new RestTemplate(factory);
+    }
+
+    public PrivateModelServiceImpl(ModelRequestLogService modelRequestLogService) {
+        this.restTemplate = createRestTemplateWithTimeout();
+        this.modelRequestLogService = modelRequestLogService;
+    }
+
+    @Override
+    public String generateResponse(String systemPrompt, String userMessage) {
+        return generateResponse(systemPrompt, userMessage, false);
+    }
+
+    @Override
+    public String generateResponse(String systemPrompt, String userMessage, boolean stream) {
+        long startTime = System.currentTimeMillis();
+        String requestJson = null;
+        String responseBody = null;
+        boolean success = false;
+        String errorMessage = null;
+        Integer promptTokens = null;
+        Integer completionTokens = null;
+        Integer totalTokens = null;
+        String extraInfo = null;
+        try {
+            logger.info("调用私有模型API - 模型: {}, 系统提示长度: {}, 用户消息长度: {}",
+                       model, systemPrompt.length(), userMessage.length());
+
+            // 生成请求参数
+            String uuid = UUID.randomUUID().toString().replace("-", "");
+            String capabilityName24 = getCapabilityName24(capabilityName);
+            String csid = appId + capabilityName24 + uuid;
+            String curTime = String.valueOf(System.currentTimeMillis() / 1000);
+
+            // 构建X-Server-Param
+            Map<String, String> serverParam = new HashMap<>();
+            serverParam.put("appid", appId);
+            serverParam.put("csid", csid);
+
+            String serverParamJson = JSON.toJSONString(serverParam);
+            String serverParamBase64 = Base64.getEncoder().encodeToString(
+                serverParamJson.getBytes(StandardCharsets.UTF_8));
+
+            // 计算校验和
+            String checkSum = calculateMD5(appKey + curTime + serverParamBase64);
+
+            // 设置请求头
+            HttpHeaders headers = new HttpHeaders();
+            headers.setContentType(MediaType.APPLICATION_JSON);
+            headers.set("X-Server-Param", serverParamBase64);
+            headers.set("X-CurTime", curTime);
+            headers.set("X-CheckSum", checkSum);
+            extraInfo = String.format("X-Server-Param=%s;X-CurTime=%s;X-CheckSum=%s", serverParamBase64, curTime, checkSum);
+
+            // 构建消息
+            Map<String, String> systemMessage = new HashMap<>();
+            systemMessage.put("role", "system");
+            systemMessage.put("content", systemPrompt);
+
+            Map<String, String> userMessageMap = new HashMap<>();
+            userMessageMap.put("role", "user");
+            userMessageMap.put("content", userMessage);
+
+            // 构建请求体
+            Map<String, Object> requestBody = new HashMap<>();
+            requestBody.put("messages", Arrays.asList(systemMessage, userMessageMap));
+            requestBody.put("model", model);
+            // 不设置max_tokens限制，让模型自由决定输出长度
+            requestBody.put("stream", stream);
+            requestBody.put("temperature", temperature);
+
+            requestJson = JSON.toJSONString(requestBody);
+            HttpEntity<Map<String, Object>> requestEntity = new HttpEntity<>(requestBody, headers);
+
+            // 发送请求
+            ResponseEntity<String> responseEntity = restTemplate.exchange(
+                apiUrl, HttpMethod.POST, requestEntity, String.class);
+
+            if (responseEntity.getStatusCode() == HttpStatus.OK) {
+                responseBody = responseEntity.getBody();
+                JSONObject jsonResponse = JSON.parseObject(responseBody);
+
+                // 检查是否符合安全要求
+                String finishReason = jsonResponse.getJSONArray("choices")
+                    .getJSONObject(0)
+                    .getString("finish_reason");
+
+                if ("content_filter".equals(finishReason)) {
+                    logger.warn("私有模型响应被内容过滤器拦截");
+                    return "抱歉，您的请求包含不适当的内容，请重新表述。【大模型返回拦截内容·】";
+                } else {
+                    // 提取回复内容
+                    String content = jsonResponse.getJSONArray("choices")
+                        .getJSONObject(0)
+                        .getJSONObject("message")
+                        .getString("content");
+
+                    JSONObject usage = jsonResponse.getJSONObject("usage");
+                    if (usage != null) {
+                        promptTokens = usage.getInteger("prompt_tokens");
+                        completionTokens = usage.getInteger("completion_tokens");
+                        totalTokens = usage.getInteger("total_tokens");
+                    }
+
+                    success = true;
+                    logger.info("私有模型API调用成功，响应长度: {}", content.length());
+                    return content;
+                }
+            } else {
+                logger.error("私有模型API调用失败，状态码: {}", responseEntity.getStatusCode());
+                throw new RuntimeException("私有模型API调用失败: " + responseEntity.getStatusCode());
+            }
+
+        } catch (HttpStatusCodeException e) {
+            responseBody = e.getResponseBodyAsString();
+            errorMessage = e.getMessage();
+            logger.error("调用私有模型API异常: {}", e.getMessage());
+            throw new RuntimeException("调用AI模型失败: " + e.getMessage(), e);
+        } catch (Exception e) {
+            errorMessage = e.getMessage();
+            logger.error("调用私有模型API异常", e);
+            throw new RuntimeException("调用AI模型失败: " + e.getMessage(), e);
+        } finally {
+            ModelRequestLog logEntry = new ModelRequestLog();
+            logEntry.setRequestSource(ModelRequestContextHolder.getRequestSourceOrDefault("AI_PRIVATE"));
+            logEntry.setRequestUrl(apiUrl);
+            logEntry.setRequestMethod("POST");
+            logEntry.setModelName(model);
+            logEntry.setAppKey(appKey);
+            logEntry.setAppSecret(capabilityName);
+            logEntry.setRequestPayload(requestJson);
+            logEntry.setResponsePayload(responseBody);
+            logEntry.setStatus(success ? "SUCCESS" : "FAILED");
+            logEntry.setDurationMs(System.currentTimeMillis() - startTime);
+            logEntry.setPromptTokens(promptTokens);
+            logEntry.setCompletionTokens(completionTokens);
+            logEntry.setTotalTokens(totalTokens);
+            logEntry.setErrorMessage(errorMessage);
+            logEntry.setExtraInfo(extraInfo);
+            modelRequestLogService.record(logEntry);
+            ModelRequestContextHolder.clear();
+        }
+    }
+
+    /**
+     * 将capabilityname补齐到24位
+     */
+    private String getCapabilityName24(String rawCapabilityName) {
+        StringBuilder sb = new StringBuilder(rawCapabilityName);
+        while (sb.length() < 24) {
+            sb.append("0");
+        }
+        return sb.toString();
+    }
+
+    /**
+     * 计算MD5值
+     */
+    private String calculateMD5(String input) {
+        try {
+            MessageDigest md = MessageDigest.getInstance("MD5");
+            byte[] hashBytes = md.digest(input.getBytes(StandardCharsets.UTF_8));
+            StringBuilder sb = new StringBuilder();
+            for (byte b : hashBytes) {
+                sb.append(String.format("%02x", b));
+            }
+            return sb.toString();
+        } catch (NoSuchAlgorithmException e) {
+            throw new RuntimeException("MD5计算失败", e);
+        }
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/PromptTemplateServiceImpl.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/PromptTemplateServiceImpl.java
new file mode 100644
index 0000000..5435fb0
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/PromptTemplateServiceImpl.java
@@ -0,0 +1,162 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.service.impl
+ * @ClassName: PromptTemplateServiceImpl
+ * @Description: 提示词模板服务实现
+ * @Author: 张志锋
+ * @Date: 2025-11-03
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.service.impl;
+
+import com.zhpb.gdyd_zhpb_zgf.dto.PromptTemplateDto;
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.PromptTemplate;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.PromptTemplateRepository;
+import com.zhpb.gdyd_zhpb_zgf.service.PromptTemplateService;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.stereotype.Service;
+import org.springframework.transaction.annotation.Transactional;
+
+import java.util.List;
+import java.util.stream.Collectors;
+
+/**
+ * 提示词模板服务实现
+ */
+@Service
+public class PromptTemplateServiceImpl implements PromptTemplateService {
+
+    private static final Logger logger = LoggerFactory.getLogger(PromptTemplateServiceImpl.class);
+
+    @Autowired
+    private PromptTemplateRepository promptTemplateRepository;
+
+    @Override
+    public List<PromptTemplateDto> getAllTemplates(String templateType) {
+        List<PromptTemplate> templates;
+        if (templateType != null && !templateType.trim().isEmpty()) {
+            templates = promptTemplateRepository.findAllByTemplateType(templateType);
+        } else {
+            templates = promptTemplateRepository.findAll();
+        }
+
+        return templates.stream()
+                .map(this::convertToDto)
+                .collect(Collectors.toList());
+    }
+
+    @Override
+    public PromptTemplateDto getTemplateById(String templateId) {
+        PromptTemplate template = promptTemplateRepository.findByTemplateId(templateId);
+        return template != null ? convertToDto(template) : null;
+    }
+
+    @Override
+    @Transactional
+    public PromptTemplateDto createTemplate(PromptTemplateDto dto, String userId) {
+        // 检查模板名称是否已存在
+        if (promptTemplateRepository.findByTemplateName(dto.getTemplateName()) != null) {
+            throw new IllegalArgumentException("模板名称已存在: " + dto.getTemplateName());
+        }
+
+        PromptTemplate template = new PromptTemplate();
+        template.setTemplateName(dto.getTemplateName());
+        template.setTemplateType(dto.getTemplateType() != null ? dto.getTemplateType() : "qa");
+        template.setPromptContent(dto.getPromptContent());
+        template.setDescription(dto.getDescription());
+        template.setIsActive(dto.getIsActive() != null ? dto.getIsActive() : true);
+        template.setSortOrder(dto.getSortOrder() != null ? dto.getSortOrder() : 0);
+        template.setCreatedBy(userId);
+        template.setUpdatedBy(userId);
+
+        template = promptTemplateRepository.save(template);
+
+        logger.info("创建提示词模板: {} (ID: {})", template.getTemplateName(), template.getTemplateId());
+        return convertToDto(template);
+    }
+
+    @Override
+    @Transactional
+    public PromptTemplateDto updateTemplate(String templateId, PromptTemplateDto dto, String userId) {
+        PromptTemplate template = promptTemplateRepository.findByTemplateId(templateId);
+        if (template == null) {
+            throw new IllegalArgumentException("模板不存在: " + templateId);
+        }
+
+        // 检查模板名称是否已被其他模板使用
+        PromptTemplate existingTemplate = promptTemplateRepository.findByTemplateName(dto.getTemplateName());
+        if (existingTemplate != null && !existingTemplate.getTemplateId().equals(templateId)) {
+            throw new IllegalArgumentException("模板名称已存在: " + dto.getTemplateName());
+        }
+
+        template.setTemplateName(dto.getTemplateName());
+        template.setTemplateType(dto.getTemplateType());
+        template.setPromptContent(dto.getPromptContent());
+        template.setDescription(dto.getDescription());
+        template.setIsActive(dto.getIsActive());
+        template.setSortOrder(dto.getSortOrder());
+        template.setUpdatedBy(userId);
+
+        template = promptTemplateRepository.save(template);
+
+        logger.info("更新提示词模板: {} (ID: {})", template.getTemplateName(), template.getTemplateId());
+        return convertToDto(template);
+    }
+
+    @Override
+    @Transactional
+    public void deleteTemplate(String templateId) {
+        PromptTemplate template = promptTemplateRepository.findByTemplateId(templateId);
+        if (template == null) {
+            throw new IllegalArgumentException("模板不存在: " + templateId);
+        }
+
+        promptTemplateRepository.delete(template);
+        logger.info("删除提示词模板: {} (ID: {})", template.getTemplateName(), template.getTemplateId());
+    }
+
+    @Override
+    @Transactional
+    public void toggleTemplateStatus(String templateId, Boolean isActive, String userId) {
+        PromptTemplate template = promptTemplateRepository.findByTemplateId(templateId);
+        if (template == null) {
+            throw new IllegalArgumentException("模板不存在: " + templateId);
+        }
+
+        template.setIsActive(isActive);
+        template.setUpdatedBy(userId);
+
+        promptTemplateRepository.save(template);
+
+        logger.info("修改提示词模板状态: {} (ID: {}, 状态: {})",
+                template.getTemplateName(), template.getTemplateId(), isActive ? "启用" : "禁用");
+    }
+
+    @Override
+    public PromptTemplate getActiveTemplate(String templateType) {
+        List<PromptTemplate> templates = promptTemplateRepository.findActiveByTemplateType(templateType);
+        return templates.isEmpty() ? null : templates.get(0);
+    }
+
+    /**
+     * 将实体转换为DTO
+     */
+    private PromptTemplateDto convertToDto(PromptTemplate template) {
+        PromptTemplateDto dto = new PromptTemplateDto();
+        dto.setTemplateId(template.getTemplateId());
+        dto.setTemplateName(template.getTemplateName());
+        dto.setTemplateType(template.getTemplateType());
+        dto.setPromptContent(template.getPromptContent());
+        dto.setDescription(template.getDescription());
+        dto.setIsActive(template.getIsActive());
+        dto.setSortOrder(template.getSortOrder());
+        dto.setCreatedBy(template.getCreatedBy());
+        dto.setUpdatedBy(template.getUpdatedBy());
+        dto.setCreateTime(template.getCreateTime());
+        dto.setUpdateTime(template.getUpdateTime());
+        return dto;
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/QaServiceImpl.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/QaServiceImpl.java
new file mode 100644
index 0000000..9c4a5eb
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/QaServiceImpl.java
@@ -0,0 +1,519 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.service.impl
+ * @ClassName: QaServiceImpl
+ * @Description: 智能问答服务实现
+ * @Author: 张志锋
+ * @Date: 2025-11-03
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.service.impl;
+
+import com.alibaba.fastjson2.JSON;
+import com.zhpb.gdyd_zhpb_zgf.context.ModelRequestContextHolder;
+import com.zhpb.gdyd_zhpb_zgf.dto.QaHistoryResponse;
+import com.zhpb.gdyd_zhpb_zgf.dto.QaRequest;
+import com.zhpb.gdyd_zhpb_zgf.dto.QaResponse;
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.File;
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.PromptTemplate;
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.QaMessage;
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.QaSession;
+import com.zhpb.gdyd_zhpb_zgf.entity.pgsql.PdfVector;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.FileRepository;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.PromptTemplateRepository;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.QaMessageRepository;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.QaSessionRepository;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.SupplierRepository;
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.Supplier;
+import com.zhpb.gdyd_zhpb_zgf.repository.pgsql.PdfVectorRepository;
+import com.zhpb.gdyd_zhpb_zgf.service.AIModelService;
+import com.zhpb.gdyd_zhpb_zgf.service.QaService;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.beans.factory.annotation.Value;
+import org.springframework.stereotype.Service;
+
+import java.util.*;
+
+/**
+ * 智能问答服务实现
+ */
+@Service
+public class QaServiceImpl implements QaService {
+
+    private static final Logger logger = LoggerFactory.getLogger(QaServiceImpl.class);
+
+    @Autowired
+    private FileRepository fileRepository;
+
+    @Autowired
+    private SupplierRepository supplierRepository;
+
+    @Autowired
+    private PromptTemplateRepository promptTemplateRepository;
+
+    @Autowired
+    private QaSessionRepository qaSessionRepository;
+
+    @Autowired
+    private QaMessageRepository qaMessageRepository;
+
+    @Autowired
+    private AIModelService aiModelService;
+
+    @Autowired
+    private PdfVectorRepository pdfVectorRepository;
+
+    @Value("${vectorize.base-url:http://localhost:8001}")
+    private String vectorizeBaseUrl;
+
+    @Value("${deepseek.api-key:}")
+    private String deepseekApiKey;
+
+    @Value("${deepseek.api-url:https://api.deepseek.com/v1/chat/completions}")
+    private String deepseekApiUrl;
+
+    @Override
+    public QaHistoryResponse.SessionsData getSessionHistory(String projectId, String userId) {
+        logger.info("查询历史会话: projectId={}, userId={}", projectId, userId);
+
+        List<QaSession> sessions = qaSessionRepository.findByProjectIdAndUserIdOrderByStartTimeDesc(projectId, userId);
+
+        List<QaHistoryResponse.SessionInfo> sessionInfos = sessions.stream()
+            .map(session -> {
+                QaHistoryResponse.SessionInfo info = new QaHistoryResponse.SessionInfo();
+                info.setSessionId(session.getSessionId());
+                info.setStartTime(session.getStartTime().toString());
+                info.setFirstQuery(session.getFirstQuery());
+
+                // 获取该会话最后一次引用的文件
+                try {
+                    QaMessage lastFileMsg = qaMessageRepository.findLastMessageWithFiles(session.getSessionId());
+                    if (lastFileMsg != null && lastFileMsg.getFileIds() != null) {
+                        List<String> fileIds = JSON.parseArray(lastFileMsg.getFileIds(), String.class);
+                        if (fileIds != null && !fileIds.isEmpty()) {
+                            info.setFileIds(fileIds);
+                            
+                            // 获取文件名和供应商名
+                            List<String> fileNames = new ArrayList<>();
+                            List<String> supplierNames = new ArrayList<>();
+                            List<String> supplierIds = new ArrayList<>();
+                            
+                            for (String rawFileId : fileIds) {
+                                String fileId = rawFileId != null ? rawFileId.trim() : "";
+                                if (fileId.isEmpty()) continue;
+
+                                Optional<File> fileOpt = fileRepository.findByFileId(fileId);
+                                if (fileOpt.isPresent()) {
+                                    File file = fileOpt.get();
+                                    fileNames.add(file.getFileName());
+                                    
+                                    // 获取供应商名称
+                                    String supplierName = "未知供应商";
+                                    String supplierId = file.getSupplierId();
+                                    
+                                    if (supplierId != null) {
+                                        Supplier supplier = supplierRepository.findBySupplierId(supplierId);
+                                        if (supplier != null && supplier.getSupplierName() != null) {
+                                            supplierName = supplier.getSupplierName();
+                                        } else {
+                                            supplierName = supplierId;
+                                        }
+                                    }
+                                    supplierNames.add(supplierName);
+                                    supplierIds.add(supplierId);
+                                } else {
+                                    logger.warn("未找到文件记录: fileId={}", fileId);
+                                    fileNames.add("未知文件");
+                                    supplierNames.add("未知供应商");
+                                    supplierIds.add("");
+                                }
+                            }
+                            
+                            info.setFileNames(fileNames);
+                            info.setSupplierNames(supplierNames);
+                            info.setSupplierIds(supplierIds);
+                        }
+                    }
+                } catch (Exception e) {
+                    logger.warn("获取会话文件信息失败: sessionId={}, error={}", session.getSessionId(), e.getMessage());
+                }
+
+                return info;
+            })
+            .toList();
+
+        QaHistoryResponse.SessionsData data = new QaHistoryResponse.SessionsData();
+        data.setSessions(sessionInfos);
+
+        return data;
+    }
+
+    @Override
+    public QaHistoryResponse.MessagesData getMessageHistory(String sessionId) {
+        logger.info("查询会话消息历史: sessionId={}", sessionId);
+
+        List<QaMessage> messages = qaMessageRepository.findBySessionIdOrderByTimestampAsc(sessionId);
+
+        List<QaHistoryResponse.MessageInfo> messageInfos = messages.stream()
+            .map(message -> {
+                QaHistoryResponse.MessageInfo info = new QaHistoryResponse.MessageInfo();
+                info.setMessageId(message.getMessageId());
+                info.setSender(message.getSender());
+                info.setContent(message.getContent());
+                info.setTimestamp(message.getTimestamp().toString());
+
+                // 解析sources JSON
+                if (message.getSources() != null && !message.getSources().isEmpty()) {
+                    try {
+                        info.setSources(JSON.parse(message.getSources()));
+                    } catch (Exception e) {
+                        logger.warn("解析消息sources失败: messageId={}", message.getMessageId());
+                        info.setSources(null);
+                    }
+                }
+
+                // 解析fileIds JSON
+                if (message.getFileIds() != null && !message.getFileIds().isEmpty()) {
+                    try {
+                        info.setFileIds(JSON.parseArray(message.getFileIds(), String.class));
+                    } catch (Exception e) {
+                        logger.warn("解析消息fileIds失败: messageId={}", message.getMessageId());
+                        info.setFileIds(null);
+                    }
+                }
+
+                return info;
+            })
+            .toList();
+
+        QaHistoryResponse.MessagesData data = new QaHistoryResponse.MessagesData();
+        data.setSessionId(sessionId);
+        data.setMessages(messageInfos);
+
+        return data;
+    }
+
+    @Override
+    public QaResponse.QaData sendMessage(QaRequest request) {
+        logger.info("开始处理智能问答请求: projectId={}, userId={}, query={}",
+                request.getProjectId(), request.getUserId(), request.getQuery());
+
+        QaResponse.QaData response = new QaResponse.QaData();
+
+        String userQuery = request.getQuery() != null ? request.getQuery().trim() : "";
+        if (userQuery.isEmpty()) {
+            logger.warn("收到空查询，请求被拒绝");
+            response.setAnswer("请先输入要咨询的问题，再提交请求。");
+            response.setSources(new ArrayList<>());
+            response.setSessionId(null);
+            response.setMessageId(null);
+            return response;
+        }
+
+        try {
+            // 会话管理：如果 sessionId 为空则创建新会话，如果有值则使用该值
+            String sessionId;
+            
+            if (request.getSessionId() != null && !request.getSessionId().trim().isEmpty()) {
+                // 如果请求中提供了 sessionId，直接使用（不创建新会话）
+                sessionId = request.getSessionId().trim();
+                logger.info("使用已有会话ID: {}", sessionId);
+            } else {
+                // 如果 sessionId 为空，创建新会话
+                sessionId = "session-" + UUID.randomUUID().toString().replace("-", "").substring(0, 12);
+                logger.info("创建新会话: {}", sessionId);
+            }
+            
+            String userMessageId = "user-msg-" + UUID.randomUUID().toString().replace("-", "").substring(0, 12);
+            String aiMessageId = "msg-" + UUID.randomUUID().toString().replace("-", "").substring(0, 12);
+
+            response.setSessionId(sessionId);
+            response.setMessageId(aiMessageId);
+
+            // 获取相关文件信息
+            List<String> fileIds = request.getFileIds();
+            
+            // 如果请求未指定文件，尝试从会话历史中获取最后一次使用的文件（上下文继承）
+            if ((fileIds == null || fileIds.isEmpty()) && sessionId != null) {
+                try {
+                    QaMessage lastFileMsg = qaMessageRepository.findLastMessageWithFiles(sessionId);
+                    if (lastFileMsg != null && lastFileMsg.getFileIds() != null) {
+                        List<String> historyFileIds = JSON.parseArray(lastFileMsg.getFileIds(), String.class);
+                        if (historyFileIds != null && !historyFileIds.isEmpty()) {
+                            fileIds = historyFileIds;
+                            logger.info("使用会话历史文件上下文: sessionId={}, fileIds={}", sessionId, fileIds);
+                        }
+                    }
+                } catch (Exception e) {
+                    logger.warn("获取会话历史文件上下文失败: sessionId={}", sessionId);
+                }
+            }
+
+            if (fileIds == null) {
+                fileIds = Collections.emptyList();
+            }
+
+            List<File> files = fileIds.stream()
+                .map(fileId -> fileRepository.findByFileId(fileId))
+                .filter(Optional::isPresent)
+                .map(Optional::get)
+                .toList();
+
+            // 获取提示词模板
+            PromptTemplate promptTemplate = getActivePromptTemplate("qa");
+            if (promptTemplate == null) {
+                logger.warn("未找到启用的智能问答提示词模板");
+                response.setAnswer("抱歉，系统配置不完整，无法处理您的请求。");
+                response.setSources(new ArrayList<>());
+                return response;
+            }
+
+            StringBuilder documentContent = new StringBuilder();
+            List<QaResponse.SourceInfo> sources = new ArrayList<>();
+
+            if (!files.isEmpty()) {
+                // 如果有文件引用，构建文档内容
+                for (File file : files) {
+                    documentContent.append("文件: ").append(file.getFileName()).append("\n");
+                    documentContent.append("供应商: ").append(getSupplierName(file)).append("\n");
+
+                    String fileContent = loadVectorContentByFileId(file.getFileId());
+                    if (fileContent == null || fileContent.isBlank()) {
+                        fileContent = "未能从向量数据库读取该文件的正文内容，请确认该文件已成功向量化。";
+                    }
+                    documentContent.append("内容: ").append(fileContent).append("\n\n");
+
+                    // 添加来源信息
+                    QaResponse.SourceInfo source = new QaResponse.SourceInfo();
+                    source.setFileId(file.getFileId());
+                    source.setFileName(file.getFileName());
+                    source.setVendorName(getSupplierName(file));
+                    source.setSimilarity(0.8); // 简化处理
+                    sources.add(source);
+                }
+            } else {
+                // 如果没有文件引用，直接使用大模型回答
+                logger.info("没有文件引用，直接使用大模型回答问题");
+                documentContent.append("请直接回答用户的问题，无需引用任何文档内容。\n\n用户问题：").append(userQuery);
+            }
+
+            // 调用DeepSeek API
+            String answer = callDeepSeekApi(userQuery, documentContent.toString(), promptTemplate.getPromptContent());
+
+            response.setAnswer(answer);
+            response.setSources(sources);
+
+            // 持久化会话和消息
+            saveQaHistory(sessionId, request.getProjectId(), request.getUserId(), userQuery,
+                         fileIds, answer, sources, userMessageId, aiMessageId);
+
+            logger.info("智能问答处理完成: sessionId={}, aiMessageId={}", sessionId, aiMessageId);
+
+        } catch (Exception e) {
+            logger.error("智能问答处理失败", e);
+            response.setAnswer("抱歉，处理您的问题时出现系统错误，请稍后重试。");
+            response.setSources(new ArrayList<>());
+        }
+
+        return response;
+    }
+
+    /**
+     * 获取启用的提示词模板
+     */
+    private PromptTemplate getActivePromptTemplate(String templateType) {
+        List<PromptTemplate> templates = promptTemplateRepository.findActiveByTemplateType(templateType);
+        return templates.isEmpty() ? null : templates.get(0);
+    }
+
+    /**
+     * 获取供应商名称
+     */
+    private String getSupplierName(File file) {
+        // 简化处理，实际应该从Supplier表中查询
+        return file.getSupplierId() != null ? "供应商-" + file.getSupplierId() : "未知供应商";
+    }
+
+    /**
+     * 持久化QA历史记录
+     */
+    private void saveQaHistory(String sessionId, String projectId, String userId, String query,
+                              List<String> fileIds, String answer, List<QaResponse.SourceInfo> sources,
+                              String userMessageId, String aiMessageId) {
+        try {
+            logger.info("开始保存QA历史记录: sessionId={}, projectId={}, userId={}", 
+                sessionId, projectId, userId);
+            
+            // 检查会话是否存在
+            QaSession session = qaSessionRepository.findBySessionId(sessionId).orElse(null);
+
+            if (session == null) {
+                // 创建新会话
+                logger.info("会话不存在，创建新会话: sessionId={}", sessionId);
+                session = new QaSession();
+                session.setSessionId(sessionId);
+                session.setProjectId(projectId);
+                session.setUserId(userId);
+                session.setFirstQuery(query);
+                session.setStartTime(java.time.LocalDateTime.now());
+                session.setMessageCount(0);
+                session.setStatus("active");
+                
+                try {
+                    session = qaSessionRepository.save(session);
+                    logger.info("新QA会话创建成功: sessionId={}, id={}", sessionId, session.getId());
+                } catch (Exception e) {
+                    logger.error("保存QA会话失败: sessionId={}, error={}", sessionId, e.getMessage(), e);
+                    throw e;
+                }
+            } else {
+                logger.info("找到已存在的会话: sessionId={}, messageCount={}", 
+                    sessionId, session.getMessageCount());
+            }
+
+            // 保存用户消息
+            logger.info("开始保存用户消息: messageId={}", userMessageId);
+            QaMessage userMessage = new QaMessage();
+            userMessage.setMessageId(userMessageId);
+            userMessage.setSessionId(sessionId);
+            userMessage.setSender("user");
+            userMessage.setContent(query);
+            userMessage.setFileIds(JSON.toJSONString(fileIds));
+            userMessage.setTimestamp(java.time.LocalDateTime.now());
+            
+            try {
+                QaMessage savedUserMessage = qaMessageRepository.save(userMessage);
+                logger.info("用户消息保存成功: messageId={}, id={}", userMessageId, savedUserMessage.getId());
+            } catch (Exception e) {
+                logger.error("保存用户消息失败: messageId={}, error={}", userMessageId, e.getMessage(), e);
+                throw e;
+            }
+
+            // 保存AI回复消息
+            logger.info("开始保存AI回复消息: messageId={}", aiMessageId);
+            QaMessage aiMessage = new QaMessage();
+            aiMessage.setMessageId(aiMessageId);
+            aiMessage.setSessionId(sessionId);
+            aiMessage.setSender("ai");
+            aiMessage.setContent(answer);
+            aiMessage.setSources(JSON.toJSONString(sources));
+            aiMessage.setFileIds(JSON.toJSONString(fileIds));
+            aiMessage.setTimestamp(java.time.LocalDateTime.now());
+            
+            try {
+                QaMessage savedAiMessage = qaMessageRepository.save(aiMessage);
+                logger.info("AI回复消息保存成功: messageId={}, id={}", aiMessageId, savedAiMessage.getId());
+            } catch (Exception e) {
+                logger.error("保存AI回复消息失败: messageId={}, error={}", aiMessageId, e.getMessage(), e);
+                throw e;
+            }
+
+            // 更新会话统计信息
+            logger.info("开始更新会话统计信息: sessionId={}", sessionId);
+            session.setLastMessageTime(java.time.LocalDateTime.now());
+            session.setMessageCount(session.getMessageCount() + 2);
+            
+            try {
+                qaSessionRepository.save(session);
+                logger.info("会话统计信息更新成功: sessionId={}, 新消息数={}", 
+                    sessionId, session.getMessageCount());
+            } catch (Exception e) {
+                logger.error("更新会话统计信息失败: sessionId={}, error={}", sessionId, e.getMessage(), e);
+                throw e;
+            }
+            
+            logger.info("QA历史记录保存完成: sessionId={}, userMessageId={}, aiMessageId={}", 
+                sessionId, userMessageId, aiMessageId);
+
+        } catch (Exception e) {
+            logger.error("保存QA历史记录失败: sessionId={}, projectId={}, userId={}, error={}", 
+                sessionId, projectId, userId, e.getMessage(), e);
+            logger.error("异常堆栈信息:", e);
+            // 不影响主要业务流程，但记录详细错误
+        }
+    }
+
+    /**
+     * 从向量数据库中加载文件内容
+     */
+    private String loadVectorContentByFileId(String fileId) {
+        try {
+            List<PdfVector> vectors = pdfVectorRepository.findByFileId(fileId);
+            if (vectors != null && !vectors.isEmpty()) {
+                String content = vectors.get(0).getContent();
+                if (content != null && !content.isBlank()) {
+                    return truncateContent(content);
+                }
+            }
+        } catch (Exception e) {
+            logger.warn("读取向量数据库内容失败: fileId={}, error={}", fileId, e.getMessage());
+        }
+        return null;
+    }
+
+    private String truncateContent(String content) {
+        String normalized = content.trim();
+        int maxLength = 2000;
+        if (normalized.length() > maxLength) {
+            return normalized.substring(0, maxLength) + "...（内容已截断）";
+        }
+        return normalized;
+    }
+
+    /**
+     * 调用AI模型API
+     */
+    private String callDeepSeekApi(String query, String documentContent, String promptTemplate) {
+        try {
+            logger.info("开始调用AI模型API");
+
+            // 构建提示词
+            String systemPrompt = promptTemplate
+                    .replace("{query}", query)
+                    .replace("{document_content}", documentContent);
+
+            // 强制要求纯文本格式
+            systemPrompt += "\n\n请注意：请仅返回纯文本格式，不要使用Markdown、HTML或其他标记语言（如 **加粗**、# 标题等）。";
+
+            // 构造用户消息，确保大模型收到明确的提问上下文
+            String userMessage = buildUserMessage(query, documentContent);
+
+            ModelRequestContextHolder.setRequestSource("QA_CHAT");
+            String response;
+            try {
+                response = aiModelService.generateResponse(systemPrompt, userMessage);
+            } finally {
+                ModelRequestContextHolder.clear();
+            }
+
+            if (response != null && !response.trim().isEmpty()) {
+                logger.info("AI模型返回回答，长度: {}", response.length());
+                return response.trim();
+            }
+
+            logger.warn("AI模型返回空响应");
+            return "抱歉，AI服务响应为空，请稍后重试。";
+
+        } catch (Exception e) {
+            logger.error("调用AI模型API失败", e);
+            return "抱歉，AI服务暂时不可用，请稍后重试。";
+        }
+    }
+
+    private String buildUserMessage(String query, String documentContent) {
+        StringBuilder builder = new StringBuilder();
+        if (documentContent != null && !documentContent.isBlank()) {
+            builder.append(documentContent.trim());
+        }
+        if (query != null && !query.isBlank()) {
+            if (builder.length() > 0) {
+                builder.append("\n\n");
+            }
+            builder.append("用户问题：").append(query.trim());
+        }
+        return builder.toString();
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/SysLogServiceImpl.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/SysLogServiceImpl.java
new file mode 100644
index 0000000..7f405e4
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/SysLogServiceImpl.java
@@ -0,0 +1,63 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.service.impl
+ * @ClassName: SysLogServiceImpl
+ * @Description: 系统日志服务实现类，异步保存日志到数据库
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.service.impl;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.SysLog;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.SysLogRepository;
+import com.zhpb.gdyd_zhpb_zgf.service.SysLogService;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.beans.factory.annotation.Value;
+import org.springframework.scheduling.annotation.Async;
+import org.springframework.stereotype.Service;
+import org.springframework.transaction.annotation.Transactional;
+
+import java.time.LocalDateTime;
+
+/**
+ * 系统日志服务实现类
+ */
+@Service
+public class SysLogServiceImpl implements SysLogService {
+
+    private static final Logger logger = LoggerFactory.getLogger(SysLogServiceImpl.class);
+
+    @Autowired
+    private SysLogRepository sysLogRepository;
+
+    @Value("${spring.application.name:gdyd_zhpb_zgf}")
+    private String applicationName;
+
+    @Override
+    @Async
+    @Transactional
+    public void log(String logLevel, String loggerName, String logContent, String fullMessage, String exceptionStack) {
+        try {
+            SysLog sysLog = new SysLog();
+            sysLog.setLogLevel(logLevel != null ? logLevel.toUpperCase() : "INFO");
+            sysLog.setLoggerName(loggerName);
+            sysLog.setLogContent(logContent != null ? logContent : "");
+            sysLog.setFullMessage(fullMessage);
+            sysLog.setExceptionStack(exceptionStack);
+            sysLog.setThreadName(Thread.currentThread().getName());
+            sysLog.setCreateTime(LocalDateTime.now());
+            sysLog.setApplicationName(applicationName);
+
+            sysLogRepository.save(sysLog);
+        } catch (Exception e) {
+            // 记录日志失败时，避免无限循环，只输出到标准错误
+            System.err.println("保存系统日志失败: " + e.getMessage());
+            logger.error("保存系统日志失败", e);
+        }
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/UserLogServiceImpl.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/UserLogServiceImpl.java
new file mode 100644
index 0000000..df6241d
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/UserLogServiceImpl.java
@@ -0,0 +1,115 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.service.impl
+ * @ClassName: UserLogServiceImpl
+ * @Description: 用户操作日志服务实现类
+ * @Author: 
+ * @Date: 2025-11-28
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.service.impl;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.UserLog;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.UserLogRepository;
+import com.zhpb.gdyd_zhpb_zgf.service.UserLogService;
+import jakarta.servlet.http.HttpServletRequest;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.stereotype.Service;
+
+import java.util.List;
+
+/**
+ * 用户操作日志服务实现类
+ */
+@Service
+public class UserLogServiceImpl implements UserLogService {
+
+    private static final Logger logger = LoggerFactory.getLogger(UserLogServiceImpl.class);
+
+    @Autowired
+    private UserLogRepository userLogRepository;
+
+    @Override
+    public void logUserOperation(String userId, String userName, String projectId,
+                                String operationType, String operationDetail,
+                                String relatedId, HttpServletRequest request) {
+        try {
+            UserLog userLog = new UserLog();
+            // 如果 userId 为 null，设置默认值
+            userLog.setUserId(userId != null ? userId : "anonymous");
+            userLog.setUserName(userName != null ? userName : "未知用户");
+            userLog.setProjectId(projectId);
+            userLog.setOperationType(operationType);
+            userLog.setOperationDetail(operationDetail);
+            userLog.setRelatedId(relatedId);
+            
+            // 获取客户端IP地址
+            if (request != null) {
+                userLog.setIpAddress(getClientIp(request));
+                userLog.setUserAgent(request.getHeader("User-Agent"));
+            }
+
+            userLogRepository.save(userLog);
+            logger.debug("记录用户操作日志成功: userId={}, operationType={}", userId, operationType);
+        } catch (Exception e) {
+            logger.warn("记录用户操作日志失败: {}", e.getMessage());
+        }
+    }
+
+    @Override
+    public void logUserOperation(String userId, String userName,
+                                String operationType, String operationDetail,
+                                HttpServletRequest request) {
+        logUserOperation(userId, userName, null, operationType, operationDetail, null, request);
+    }
+
+    @Override
+    public List<UserLog> getUserLogsByUserId(String userId) {
+        return userLogRepository.findByUserIdOrderByCreateTimeDesc(userId);
+    }
+
+    @Override
+    public List<UserLog> getUserLogsByProjectId(String projectId) {
+        return userLogRepository.findByProjectIdOrderByCreateTimeDesc(projectId);
+    }
+
+    @Override
+    public List<UserLog> getUserLogsByOperationType(String operationType) {
+        return userLogRepository.findByOperationTypeOrderByCreateTimeDesc(operationType);
+    }
+
+    @Override
+    public List<UserLog> getUserLogsByUserIdAndOperationType(String userId, String operationType) {
+        return userLogRepository.findByUserIdAndOperationTypeOrderByCreateTimeDesc(userId, operationType);
+    }
+
+    @Override
+    public Long countUserOperations(String userId) {
+        return userLogRepository.countByUserId(userId);
+    }
+
+    @Override
+    public Long countOperationsByType(String operationType) {
+        return userLogRepository.countByOperationType(operationType);
+    }
+
+    /**
+     * 获取客户端IP地址
+     */
+    private String getClientIp(HttpServletRequest request) {
+        String ip = request.getHeader("X-Forwarded-For");
+        if (ip == null || ip.isEmpty() || "unknown".equalsIgnoreCase(ip)) {
+            ip = request.getHeader("Proxy-Client-IP");
+        }
+        if (ip == null || ip.isEmpty() || "unknown".equalsIgnoreCase(ip)) {
+            ip = request.getHeader("WL-Proxy-Client-IP");
+        }
+        if (ip == null || ip.isEmpty() || "unknown".equalsIgnoreCase(ip)) {
+            ip = request.getRemoteAddr();
+        }
+        return ip;
+    }
+}
\ No newline at end of file
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/UserServiceImpl.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/UserServiceImpl.java
new file mode 100644
index 0000000..db3dc45
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/UserServiceImpl.java
@@ -0,0 +1,63 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.service.impl
+ * @ClassName: UserServiceImpl
+ * @Description: 用户服务实现类，实现用户查询、创建、更新等业务逻辑
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.service.impl;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.User;
+import com.zhpb.gdyd_zhpb_zgf.repository.mysql.UserRepository;
+import com.zhpb.gdyd_zhpb_zgf.service.UserService;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.stereotype.Service;
+import org.springframework.transaction.annotation.Transactional;
+
+import java.time.LocalDateTime;
+
+/**
+ * 用户服务实现类
+ */
+@Service
+public class UserServiceImpl implements UserService {
+    
+    @Autowired
+    private UserRepository userRepository;
+    
+    @Override
+    public User findByUserId(String userId) {
+        return userRepository.findByUserId(userId);
+    }
+    
+    @Override
+    @Transactional
+    public User createUser(String userId, String name) {
+        return createUser(userId, name, null);
+    }
+    
+    @Override
+    @Transactional
+    public User createUser(String userId, String name, String projectId) {
+        User user = new User();
+        user.setUserId(userId);
+        user.setName(name);
+        user.setProjectId(projectId);
+        user.setStatus("ACTIVE");
+        user.setCreateTime(LocalDateTime.now());
+        user.setUpdateTime(LocalDateTime.now());
+        
+        return userRepository.save(user);
+    }
+    
+    @Override
+    @Transactional
+    public User updateUser(User user) {
+        user.setUpdateTime(LocalDateTime.now());
+        return userRepository.save(user);
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/VectorizationServiceImpl.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/VectorizationServiceImpl.java
new file mode 100644
index 0000000..6415905
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/service/impl/VectorizationServiceImpl.java
@@ -0,0 +1,430 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.service.impl
+ * @ClassName: VectorizationServiceImpl
+ * @Description: 向量化服务实现
+ * @Author: 张志锋
+ * @Date: 2025-11-05
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.service.impl;
+
+import com.zhpb.gdyd_zhpb_zgf.entity.mysql.ModelRequestLog;
+import com.zhpb.gdyd_zhpb_zgf.service.ModelRequestLogService;
+import com.zhpb.gdyd_zhpb_zgf.service.VectorizationService;
+import com.zhpb.gdyd_zhpb_zgf.utils.VectorizeUtils;
+import com.zhpb.gdyd_zhpb_zgf.utils.VectorizeUtils.VectorizeRequestException;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.beans.factory.annotation.Value;
+import org.springframework.core.env.Environment;
+import org.springframework.stereotype.Service;
+
+import java.util.ArrayList;
+import java.util.List;
+
+/**
+ * 向量化服务实现
+ * 支持三种向量化模式：local（本地）、online（在线API）、office（办公网关）
+ */
+@Service
+public class VectorizationServiceImpl implements VectorizationService {
+
+    private static final Logger logger = LoggerFactory.getLogger(VectorizationServiceImpl.class);
+
+    @Value("${vectorization.mode:local}")
+    private String vectorizationMode;
+
+    @Value("${vectorize.base-url:http://localhost:8001}")
+    private String vectorizeBaseUrl;
+
+    @Value("${vectorize.api-path:/vectorize}")
+    private String vectorizeApiPath;
+
+    @Value("${vectorize.api-key:}")
+    private String vectorizeApiKey;
+
+    @Value("${vectorize.model:}")
+    private String vectorizeModel;
+
+    @Value("${vector-limits.default.max-chars:0}")
+    private int defaultMaxChars;
+
+    @Value("${vector-limits.default.max-tokens:0}")
+    private int defaultMaxTokens;
+
+    @Value("${vectorization.batch-size:10}")
+    private int batchSize;
+
+    @Autowired
+    private OfficeGatewayServiceImpl officeGatewayService;
+
+    @Autowired
+    private Environment environment;
+
+    @Autowired
+    private ModelRequestLogService modelRequestLogService;
+
+    @Override
+    public float[] vectorizeText(String text) {
+        try {
+            List<float[]> results = vectorizeTexts(List.of(text));
+            return results.isEmpty() ? new float[0] : results.get(0);
+        } catch (Exception e) {
+            logger.error("单文本向量化失败", e);
+            throw new RuntimeException("向量化失败: " + e.getMessage());
+        }
+    }
+
+    @Override
+    public List<float[]> vectorizeTexts(List<String> texts) {
+        logger.info("开始向量化文本，模式: {}, 文本数量: {}", vectorizationMode, texts.size());
+
+        LimitSettings limitSettings = resolveLimitSettings();
+        List<String> normalizedTexts = normalizeTexts(texts, limitSettings);
+        String modelKey = limitSettings.getModelKey();
+
+        try {
+            switch (vectorizationMode.toLowerCase()) {
+                case "office":
+                    return vectorizeWithOfficeGateway(normalizedTexts);
+                case "online":
+                    return vectorizeWithOnlineApi(normalizedTexts, modelKey);
+                case "local":
+                default:
+                    return vectorizeWithLocalService(normalizedTexts, modelKey);
+            }
+        } catch (Exception e) {
+            logger.error("向量化失败，模式: {}", vectorizationMode, e);
+            throw new RuntimeException("向量化失败: " + e.getMessage());
+        }
+    }
+
+    @Override
+    public int getDimension() {
+        // 根据不同模式返回不同的维度
+        switch (vectorizationMode.toLowerCase()) {
+            case "office":
+                // 办公网关embed-qwen3-8b模型的维度
+                return 1024; // 根据embed-qwen3-8b的实际维度调整
+            case "online":
+            case "local":
+            default:
+                // 如果配置了 OpenAI 模型，返回 1536（text-embedding-ada-002 的维度）
+                if (vectorizeModel != null && !vectorizeModel.trim().isEmpty()) {
+                    // OpenAI text-embedding-ada-002 的维度是 1536
+                    if (vectorizeModel.contains("ada-002") || vectorizeModel.contains("text-embedding")) {
+                        return 1536;
+                    }
+                }
+                return 896; // 默认维度
+        }
+    }
+
+    /**
+     * 使用办公网关进行向量化
+     */
+    private List<float[]> vectorizeWithOfficeGateway(List<String> texts) {
+        logger.info("使用办公网关进行向量化，批次大小: {}", batchSize);
+
+        List<float[]> allVectors = new ArrayList<>();
+
+        // 分批处理
+        for (int i = 0; i < texts.size(); i += batchSize) {
+            int endIndex = Math.min(i + batchSize, texts.size());
+            List<String> batch = texts.subList(i, endIndex);
+
+            try {
+                List<float[]> batchVectors = officeGatewayService.embedTexts(batch);
+                allVectors.addAll(batchVectors);
+
+                logger.debug("办公网关向量化批次完成: {}-{}", i, endIndex);
+            } catch (Exception e) {
+                logger.error("办公网关向量化批次失败: {}-{}", i, endIndex, e);
+                // 为失败的批次添加空向量
+                for (int j = 0; j < batch.size(); j++) {
+                    allVectors.add(new float[getDimension()]);
+                }
+            }
+        }
+
+        return allVectors;
+    }
+
+    /**
+     * 使用在线API进行向量化
+     */
+    private List<float[]> vectorizeWithOnlineApi(List<String> texts, String modelKey) {
+        logger.info("使用在线API进行向量化");
+
+        List<float[]> vectors = new ArrayList<>();
+
+        for (String text : texts) {
+            try {
+                VectorizeUtils.VectorizeResult result = VectorizeUtils.vectorize(
+                    text, vectorizeBaseUrl, vectorizeApiPath, vectorizeApiKey, vectorizeModel);
+                vectors.add(result.getVector());
+
+                recordVectorizeLog(
+                    "VECTORIZE_ONLINE",
+                    result.getModel() != null ? result.getModel() : modelKey,
+                    result.getRequestUrl(),
+                    result.getRequestPayload(),
+                    result.getResponsePayload(),
+                    result.getDurationMs(),
+                    true,
+                    null,
+                    result.getApiKey(),
+                    result.getPromptTokens(),
+                    result.getCompletionTokens(),
+                    result.getTotalTokens()
+                );
+            } catch (Exception e) {
+                logger.error("在线API向量化单文本失败: {}", text.substring(0, Math.min(100, text.length())), e);
+                // 添加空向量
+                vectors.add(new float[getDimension()]);
+
+                if (e instanceof VectorizeRequestException vex) {
+                    recordVectorizeLog(
+                        "VECTORIZE_ONLINE",
+                        vex.getModel() != null ? vex.getModel() : modelKey,
+                        vex.getRequestUrl(),
+                        vex.getRequestPayload(),
+                        vex.getResponsePayload(),
+                        vex.getDurationMs(),
+                        false,
+                        vex.getMessage(),
+                        vex.getApiKey(),
+                        null,
+                        null,
+                        null
+                    );
+                } else {
+                    recordVectorizeLog(
+                        "VECTORIZE_ONLINE",
+                        modelKey,
+                        null,
+                        null,
+                        null,
+                        null,
+                        false,
+                        e.getMessage(),
+                        null,
+                        null,
+                        null,
+                        null
+                    );
+                }
+            }
+        }
+
+        return vectors;
+    }
+
+    /**
+     * 使用本地服务进行向量化
+     */
+    private List<float[]> vectorizeWithLocalService(List<String> texts, String modelKey) {
+        logger.info("使用本地服务进行向量化");
+
+        List<float[]> vectors = new ArrayList<>();
+
+        for (String text : texts) {
+            try {
+                VectorizeUtils.VectorizeResult result = VectorizeUtils.vectorize(
+                    text, vectorizeBaseUrl, vectorizeApiPath, vectorizeApiKey, vectorizeModel);
+                vectors.add(result.getVector());
+
+                recordVectorizeLog(
+                    "VECTORIZE_LOCAL",
+                    result.getModel() != null ? result.getModel() : modelKey,
+                    result.getRequestUrl(),
+                    result.getRequestPayload(),
+                    result.getResponsePayload(),
+                    result.getDurationMs(),
+                    true,
+                    null,
+                    result.getApiKey(),
+                    result.getPromptTokens(),
+                    result.getCompletionTokens(),
+                    result.getTotalTokens()
+                );
+            } catch (Exception e) {
+                logger.error("本地服务向量化单文本失败: {}", text.substring(0, Math.min(100, text.length())), e);
+                // 添加空向量
+                vectors.add(new float[getDimension()]);
+
+                if (e instanceof VectorizeRequestException vex) {
+                    recordVectorizeLog(
+                        "VECTORIZE_LOCAL",
+                        vex.getModel() != null ? vex.getModel() : modelKey,
+                        vex.getRequestUrl(),
+                        vex.getRequestPayload(),
+                        vex.getResponsePayload(),
+                        vex.getDurationMs(),
+                        false,
+                        vex.getMessage(),
+                        vex.getApiKey(),
+                        null,
+                        null,
+                        null
+                    );
+                } else {
+                    recordVectorizeLog(
+                        "VECTORIZE_LOCAL",
+                        modelKey,
+                        null,
+                        null,
+                        null,
+                        null,
+                        false,
+                        e.getMessage(),
+                        null,
+                        null,
+                        null,
+                        null
+                    );
+                }
+            }
+        }
+
+        return vectors;
+    }
+
+    private List<String> normalizeTexts(List<String> texts, LimitSettings limits) {
+        List<String> normalized = new ArrayList<>(texts == null ? 0 : texts.size());
+        if (texts == null) {
+            return normalized;
+        }
+        for (String text : texts) {
+            normalized.add(limitTextLength(text, limits));
+        }
+        return normalized;
+    }
+
+    private String limitTextLength(String rawText, LimitSettings limits) {
+        if (rawText == null) {
+            return "";
+        }
+        String trimmed = rawText.trim();
+        if (trimmed.isEmpty()) {
+            return trimmed;
+        }
+
+        if (limits.getMaxChars() > 0) {
+            int codePoints = trimmed.codePointCount(0, trimmed.length());
+            if (codePoints > limits.getMaxChars()) {
+                int endIndex = trimmed.offsetByCodePoints(0, limits.getMaxChars());
+                logger.debug("模型 {} 文本超出单次{}字限制，已截断。原始长度: {}", limits.getModelKey(), limits.getMaxChars(), codePoints);
+                trimmed = trimmed.substring(0, endIndex);
+            }
+        }
+
+        if (limits.getMaxTokens() > 0) {
+            int tokenCount = estimateTokenCount(trimmed);
+            if (tokenCount > limits.getMaxTokens()) {
+                int endIndex = trimmed.offsetByCodePoints(0, limits.getMaxTokens());
+                logger.debug("模型 {} 文本超出{} token限制，已截断。原始token数: {}", limits.getModelKey(), limits.getMaxTokens(), tokenCount);
+                trimmed = trimmed.substring(0, endIndex);
+            }
+        }
+
+        return trimmed;
+    }
+
+    private int estimateTokenCount(String text) {
+        if (text == null || text.isEmpty()) {
+            return 0;
+        }
+        return text.codePointCount(0, text.length());
+    }
+
+    private LimitSettings resolveLimitSettings() {
+        String modelKey = determineModelKey();
+        int charLimit = resolveLimitValue(modelKey, "max-chars", defaultMaxChars);
+        int tokenLimit = resolveLimitValue(modelKey, "max-tokens", defaultMaxTokens);
+        logger.debug("向量化模型 {} 使用阈值: {} 字 / {} token", modelKey, charLimit, tokenLimit);
+        return new LimitSettings(modelKey, charLimit, tokenLimit);
+    }
+
+    private String determineModelKey() {
+        if ("office".equalsIgnoreCase(vectorizationMode)) {
+            String officeModel = officeGatewayService.getEmbeddingModel();
+            if (officeModel != null && !officeModel.isBlank()) {
+                return officeModel;
+            }
+            return "office";
+        }
+        if (vectorizeModel != null && !vectorizeModel.isBlank()) {
+            return vectorizeModel;
+        }
+        return "default";
+    }
+
+    private int resolveLimitValue(String modelKey, String property, int fallback) {
+        String propertyKey = "vector-limits." + modelKey + "." + property;
+        Integer value = environment.getProperty(propertyKey, Integer.class);
+        if (value == null || value <= 0) {
+            return fallback;
+        }
+        return value;
+    }
+
+    private void recordVectorizeLog(String source,
+                                    String modelName,
+                                    String requestUrl,
+                                    String requestPayload,
+                                    String responsePayload,
+                                    Long durationMs,
+                                    boolean success,
+                                    String errorMessage,
+                                    String apiKey,
+                                    Integer promptTokens,
+                                    Integer completionTokens,
+                                    Integer totalTokens) {
+        if (modelRequestLogService == null) {
+            return;
+        }
+        ModelRequestLog logEntry = new ModelRequestLog();
+        logEntry.setRequestSource(source);
+        logEntry.setRequestUrl(requestUrl);
+        logEntry.setRequestMethod("POST");
+        logEntry.setModelName(modelName);
+        logEntry.setRequestPayload(requestPayload);
+        logEntry.setResponsePayload(responsePayload);
+        logEntry.setStatus(success ? "SUCCESS" : "FAILED");
+        logEntry.setDurationMs(durationMs);
+        logEntry.setApiKey(apiKey);
+        logEntry.setPromptTokens(promptTokens);
+        logEntry.setCompletionTokens(completionTokens);
+        logEntry.setTotalTokens(totalTokens);
+        logEntry.setErrorMessage(errorMessage);
+        modelRequestLogService.record(logEntry);
+    }
+
+    private static class LimitSettings {
+        private final String modelKey;
+        private final int maxChars;
+        private final int maxTokens;
+
+        LimitSettings(String modelKey, int maxChars, int maxTokens) {
+            this.modelKey = modelKey;
+            this.maxChars = maxChars;
+            this.maxTokens = maxTokens;
+        }
+
+        String getModelKey() {
+            return modelKey;
+        }
+
+        int getMaxChars() {
+            return maxChars;
+        }
+
+        int getMaxTokens() {
+            return maxTokens;
+        }
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/utils/ApiAuthUtils.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/utils/ApiAuthUtils.java
new file mode 100644
index 0000000..4e7e2b4
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/utils/ApiAuthUtils.java
@@ -0,0 +1,136 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.utils
+ * @ClassName: ApiAuthUtils
+ * @Description: API认证工具类
+ * @Author: 张志锋
+ * @Date: 2025-11-05
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.utils;
+
+import com.alibaba.fastjson2.JSON;
+import com.alibaba.fastjson2.JSONObject;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import org.springframework.beans.factory.annotation.Value;
+import org.springframework.stereotype.Component;
+
+import java.nio.charset.StandardCharsets;
+import java.security.MessageDigest;
+import java.security.NoSuchAlgorithmException;
+import java.util.Base64;
+import java.util.Map;
+import java.util.UUID;
+
+/**
+ * API认证工具类
+ * 用于验证ES平台的请求认证信息
+ */
+@Component
+public class ApiAuthUtils {
+
+    private static final Logger logger = LoggerFactory.getLogger(ApiAuthUtils.class);
+
+    @Value("${private-model.appkey:}")
+    private String appKey;
+
+    /**
+     * 验证请求头中的认证信息
+     *
+     * @param xServerParam X-Server-Param头
+     * @param xCurTime X-CurTime头
+     * @param xCheckSum X-CheckSum头
+     * @param expectedAppId 期望的AppId
+     * @return 验证是否通过
+     */
+    public boolean validateAuth(String xServerParam, String xCurTime, String xCheckSum, String expectedAppId) {
+        try {
+            // 1. 解码X-Server-Param
+            String serverParamJson = new String(Base64.getDecoder().decode(xServerParam), StandardCharsets.UTF_8);
+            JSONObject serverParam = JSON.parseObject(serverParamJson);
+
+            String appid = serverParam.getString("appid");
+            String csid = serverParam.getString("csid");
+
+//            // 2. 验证appid
+//            if (!expectedAppId.equals(appid)) {
+//                logger.warn("AppId不匹配，期望: {}, 实际: {}", expectedAppId, appid);
+//                return false;
+//            }
+
+            // 3. 计算校验和
+//            String expectedCheckSum = calculateMD5(appKey + xCurTime + xServerParam);
+
+            // 4. 验证校验和
+//            if (!expectedCheckSum.equals(xCheckSum)) {
+//                logger.warn("校验和不匹配，期望: {}, 实际: {}", expectedCheckSum, xCheckSum);
+//                return false;
+//            }
+
+            // 5. 验证时间戳（允许5分钟的时间差）
+            long currentTime = System.currentTimeMillis() / 1000;
+            long requestTime = Long.parseLong(xCurTime);
+            long timeDiff = Math.abs(currentTime - requestTime);
+
+            if (timeDiff > 300) { // 5分钟
+                logger.warn("请求时间戳过期，当前时间: {}, 请求时间: {}, 时间差: {}秒",
+                           currentTime, requestTime, timeDiff);
+                return false;
+            }
+
+            logger.info("API认证通过，AppId: {}", appid);
+            return true;
+
+        } catch (Exception e) {
+            logger.error("API认证失败", e);
+            return false;
+        }
+    }
+
+    /**
+     * 计算MD5值
+     */
+    private String calculateMD5(String input) {
+        try {
+            MessageDigest md = MessageDigest.getInstance("MD5");
+            byte[] hashBytes = md.digest(input.getBytes(StandardCharsets.UTF_8));
+            StringBuilder sb = new StringBuilder();
+            for (byte b : hashBytes) {
+                sb.append(String.format("%02x", b));
+            }
+            return sb.toString();
+        } catch (NoSuchAlgorithmException e) {
+            throw new RuntimeException("MD5计算失败", e);
+        }
+    }
+
+    /**
+     * 生成X-Server-Param
+     * 用于测试或其他需要生成认证信息的地方
+     */
+    public Map<String, String> generateServerParam(String appId, String capabilityName) {
+        String uuid = UUID.randomUUID().toString().replace("-", "");
+        String capabilityName24 = getCapabilityName24(capabilityName);
+        String csid = appId + capabilityName24 + uuid;
+
+        Map<String, String> serverParam = Map.of(
+            "appid", appId,
+            "csid", csid
+        );
+
+        return serverParam;
+    }
+
+    /**
+     * 将capabilityname补齐到24位
+     */
+    private String getCapabilityName24(String rawCapabilityName) {
+        StringBuilder sb = new StringBuilder(rawCapabilityName);
+        while (sb.length() < 24) {
+            sb.append("0");
+        }
+        return sb.toString();
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/utils/FilePathOrganizer.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/utils/FilePathOrganizer.java
new file mode 100644
index 0000000..df37063
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/utils/FilePathOrganizer.java
@@ -0,0 +1,268 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.utils
+ * @ClassName: FilePathOrganizer
+ * @Description: 文件路径整理工具类，将文件路径按照公司、文件类型（技术文件、商务文件、报价文件）进行分组
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.utils;
+
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+import java.util.*;
+
+/**
+ * 文件路径整理工具类
+ * 将文件路径按照公司、文件类型进行分组
+ */
+public class FilePathOrganizer {
+
+    private static final Logger logger = LoggerFactory.getLogger(FilePathOrganizer.class);
+
+    /**
+     * 文件类型映射
+     */
+    private static final Map<String, String> FILE_TYPE_MAP = new HashMap<>();
+    
+    static {
+        FILE_TYPE_MAP.put("技术文件", "技术文件");
+        FILE_TYPE_MAP.put("商务文件", "商务文件");
+        FILE_TYPE_MAP.put("报价文件", "报价文件");
+    }
+
+    /**
+     * 整理文件路径，按照公司和文件类型分组
+     * 
+     * @param filePaths 文件路径列表
+     * @return 整理后的JSON结构：{"公司1": {"技术文件": [...], "商务文件": [...]}}
+     */
+    public static Map<String, Map<String, List<String>>> organizeByCompanyAndType(List<String> filePaths) {
+        // 结果：Map<公司名, Map<文件类型, List<文件路径>>>
+        Map<String, Map<String, List<String>>> result = new LinkedHashMap<>();
+        
+        logger.info("开始整理文件路径，共 {} 个文件", filePaths.size());
+
+        for (String filePath : filePaths) {
+            try {
+                logger.debug("处理文件路径: {}", filePath);
+                // 统一处理路径分隔符（Windows和Linux兼容）
+                String normalizedPath = filePath.replace('\\', '/');
+                
+                // 从路径中提取关键信息
+                // 路径结构: .../extracts/项目目录/公司名_xxx/文件类型-xxx/投标文件/结果文件/文件名.pdf
+                String[] pathSegments = normalizedPath.split("/");
+                
+                String companyName = null;
+                String fileType = null;
+                
+                // 从后往前查找文件类型和公司名
+                for (int i = pathSegments.length - 1; i >= 0; i--) {
+                    String segment = pathSegments[i];
+                    
+                    // 查找文件类型（技术文件、商务文件、报价文件）
+                    // 支持匹配：技术文件、技术文件-2507161938、技术文件-xxx等形式
+                    if (fileType == null) {
+                        for (String type : FILE_TYPE_MAP.keySet()) {
+                            // 精确匹配或包含文件类型关键词（注意：先精确匹配，再匹配带横线的）
+                            if (segment.equals(type) || segment.startsWith(type + "-") || 
+                                (segment.length() >= type.length() && segment.substring(0, type.length()).equals(type))) {
+                                fileType = FILE_TYPE_MAP.get(type);
+                                logger.debug("从路径段 '{}' 中识别出文件类型: {}", segment, fileType);
+                                break;
+                            }
+                        }
+                    }
+                    
+                    // 查找公司名（在文件类型之前的一个包含"公司"的段）
+                    if (fileType != null && companyName == null && i > 0) {
+                        String prevSegment = pathSegments[i - 1];
+                        if (prevSegment.contains("公司") || prevSegment.contains("企业") || 
+                            prevSegment.contains("有限") || prevSegment.contains("股份") ||
+                            prevSegment.contains("集团") || prevSegment.contains("物流") ||
+                            prevSegment.contains("供应链") || prevSegment.contains("科技") ||
+                            prevSegment.contains("管理")) {
+                            // 去掉后缀如"_虚拟标段_标包1_文件"
+                            // 提取公司名：找到包含"公司"、"企业"、"有限"、"股份"的部分，但保留完整公司名
+                            String[] parts = prevSegment.split("_");
+                            StringBuilder companyBuilder = new StringBuilder();
+                            boolean foundCompanyKeyword = false;
+                            
+                            // 改进逻辑：完整保留包含公司关键词的所有部分
+                            for (String part : parts) {
+                                // 如果是项目相关后缀，停止
+                                if (part.contains("虚拟") || part.contains("标段") || 
+                                    part.contains("标包") || part.contains("文件") ||
+                                    part.matches("\\d+")) {
+                                    // 如果已经找到了公司关键词，说明后面的都是后缀，停止
+                                    if (foundCompanyKeyword) {
+                                        break;
+                                    }
+                                    // 如果还没找到公司关键词，这个部分可能不是公司名的一部分，跳过
+                                    continue;
+                                }
+                                
+                                // 添加部分到公司名
+                                if (companyBuilder.length() > 0) {
+                                    companyBuilder.append("_");
+                                }
+                                companyBuilder.append(part);
+                                
+                                // 检查是否包含公司关键词
+                                if (part.contains("公司") || part.contains("企业") || 
+                                    part.contains("有限") || part.contains("股份")) {
+                                    foundCompanyKeyword = true;
+                                }
+                            }
+                            
+                            // 如果成功构建了公司名，使用它；否则使用第一个部分（不含后缀）
+                            if (companyBuilder.length() > 0) {
+                                companyName = companyBuilder.toString();
+                                // 如果公司名以"_"结尾，去掉最后一个"_"
+                                if (companyName.endsWith("_")) {
+                                    companyName = companyName.substring(0, companyName.length() - 1);
+                                }
+                            } else {
+                                // 如果构建失败，尝试使用第一个非后缀部分
+                                for (String part : parts) {
+                                    if (!part.contains("虚拟") && !part.contains("标段") && 
+                                        !part.contains("标包") && !part.contains("文件")) {
+                                        companyName = part;
+                                        break;
+                                    }
+                                }
+                                if (companyName == null) {
+                                    companyName = prevSegment.split("_")[0];
+                                }
+                            }
+                            
+                            logger.debug("从路径段 '{}' 中识别出公司名: {}", prevSegment, companyName);
+                            break;
+                        }
+                    }
+                }
+
+                // 如果仍然没找到，使用更宽松的匹配
+                if (companyName == null || fileType == null) {
+                    // 重新搜索，使用更宽松的条件
+                    for (int i = 0; i < pathSegments.length; i++) {
+                        String segment = pathSegments[i];
+                        
+                        // 查找文件类型（支持精确匹配或包含）
+                        if (fileType == null) {
+                            for (String type : FILE_TYPE_MAP.keySet()) {
+                                // 精确匹配或包含文件类型关键词
+                                if (segment.equals(type) || segment.startsWith(type + "-") || segment.startsWith(type)) {
+                                    fileType = FILE_TYPE_MAP.get(type);
+                                    break;
+                                }
+                            }
+                        }
+                        
+                        // 查找公司名（包含中文且可能包含公司关键词）
+                        if (fileType != null && companyName == null && i > 0) {
+                            String prevSegment = pathSegments[i - 1];
+                            // 如果上一段包含中文且不是项目目录（不包含project-）
+                            if (!prevSegment.contains("project-") && 
+                                prevSegment.matches(".*[\\u4e00-\\u9fa5].*")) {
+                                // 改进处理：尝试提取完整公司名（去掉后缀）
+                                String[] parts = prevSegment.split("_");
+                                StringBuilder companyBuilder = new StringBuilder();
+                                
+                                for (String part : parts) {
+                                    // 如果是项目相关后缀，停止
+                                    if (part.contains("虚拟") || part.contains("标段") || 
+                                        part.contains("标包") || part.contains("文件")) {
+                                        break;
+                                    }
+                                    
+                                    if (companyBuilder.length() > 0) {
+                                        companyBuilder.append("_");
+                                    }
+                                    companyBuilder.append(part);
+                                    
+                                    // 如果遇到公司关键词，可能已经找到了完整的公司名
+                                    if (part.contains("公司") || part.contains("企业") || 
+                                        part.contains("有限") || part.contains("股份")) {
+                                        // 继续查找，看是否还有后续部分（如"管理有限公司"）
+                                        // 但如果有"虚拟"、"标段"等后缀，则停止
+                                    }
+                                }
+                                
+                                if (companyBuilder.length() > 0) {
+                                    companyName = companyBuilder.toString();
+                                    if (companyName.endsWith("_")) {
+                                        companyName = companyName.substring(0, companyName.length() - 1);
+                                    }
+                                } else {
+                                    // 如果构建失败，使用第一个部分
+                                    companyName = parts[0];
+                                    if (companyName.length() < 2) {
+                                        companyName = prevSegment; // 如果太短，使用完整段
+                                    }
+                                }
+                                
+                                logger.debug("从宽松匹配中识别出公司名: {} -> {}", prevSegment, companyName);
+                                break;
+                            }
+                        }
+                    }
+                }
+
+                // 如果仍然没找到，使用默认值
+                if (companyName == null || companyName.trim().isEmpty()) {
+                    companyName = "未知公司";
+                    logger.warn("无法从路径中提取公司名: {}", filePath);
+                }
+                
+                if (fileType == null) {
+                    fileType = "其他文件";
+                    logger.warn("无法从路径中提取文件类型: {}", filePath);
+                }
+
+                // 添加到结果中
+                result.computeIfAbsent(companyName, k -> new LinkedHashMap<>())
+                      .computeIfAbsent(fileType, k -> new ArrayList<>())
+                      .add(filePath);
+                
+                logger.debug("文件已分类: 公司={}, 类型={}, 文件={}", companyName, fileType, 
+                        java.nio.file.Paths.get(filePath).getFileName().toString());
+
+            } catch (Exception e) {
+                logger.error("处理文件路径失败: {}, 错误: {}", filePath, e.getMessage(), e);
+            }
+        }
+        
+        // 统计信息
+        logger.info("文件整理完成，共识别出 {} 个公司", result.size());
+        for (Map.Entry<String, Map<String, List<String>>> entry : result.entrySet()) {
+            String companyName = entry.getKey();
+            Map<String, List<String>> fileTypes = entry.getValue();
+            int totalFiles = fileTypes.values().stream().mapToInt(List::size).sum();
+            logger.info("公司: {}, 总文件数: {}, 类型分布: {}", companyName, totalFiles, 
+                    fileTypes.entrySet().stream()
+                        .map(e -> e.getKey() + "(" + e.getValue().size() + ")")
+                        .collect(java.util.stream.Collectors.joining(", ")));
+        }
+
+        // 确保每个公司都有完整的文件类型结构（技术文件、商务文件、报价文件），即使为空也显示空数组
+        List<String> requiredFileTypes = Arrays.asList("技术文件", "商务文件", "报价文件");
+        for (Map.Entry<String, Map<String, List<String>>> entry : result.entrySet()) {
+            String companyName = entry.getKey();
+            Map<String, List<String>> companyFiles = entry.getValue();
+            for (String requiredType : requiredFileTypes) {
+                // 如果缺少某个文件类型，添加空数组
+                if (!companyFiles.containsKey(requiredType)) {
+                    companyFiles.put(requiredType, new ArrayList<>());
+                    logger.debug("为公司 '{}' 补充缺失的文件类型: {}", companyName, requiredType);
+                }
+            }
+        }
+
+        return result;
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/utils/HttpClientUtils.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/utils/HttpClientUtils.java
new file mode 100644
index 0000000..4cd8d07
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/utils/HttpClientUtils.java
@@ -0,0 +1,150 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.utils
+ * @ClassName: HttpClientUtils
+ * @Description: HTTP客户端工具类，提供HTTP GET请求功能，用于调用外部ES接口
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.utils;
+
+import org.apache.hc.client5.http.classic.methods.HttpGet;
+import org.apache.hc.client5.http.classic.methods.HttpPost;
+import org.apache.hc.client5.http.impl.classic.CloseableHttpClient;
+import org.apache.hc.client5.http.impl.classic.HttpClients;
+import org.apache.hc.core5.http.io.entity.EntityUtils;
+import org.apache.hc.core5.http.io.entity.StringEntity;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+import java.nio.charset.StandardCharsets;
+import java.util.Map;
+
+/**
+ * HTTP 请求工具类
+ */
+public class HttpClientUtils {
+    
+    /**
+     * 发送GET请求
+     *
+     * @param url 请求URL
+     * @return 响应内容
+     */
+    public static String doGet(String url) {
+        try (CloseableHttpClient httpClient = HttpClients.createDefault()) {
+            HttpGet httpGet = new HttpGet(url);
+
+            return httpClient.execute(httpGet, response -> {
+                int statusCode = response.getCode();
+                if (statusCode == 200) {
+                    return EntityUtils.toString(response.getEntity(), StandardCharsets.UTF_8);
+                } else {
+                    throw new RuntimeException("HTTP请求失败，状态码: " + statusCode);
+                }
+            });
+        } catch (Exception e) {
+            throw new RuntimeException("HTTP请求异常: " + e.getMessage(), e);
+        }
+    }
+
+    /**
+     * 发送POST JSON请求
+     *
+     * @param url 请求URL
+     * @param jsonData JSON数据
+     * @param headers 请求头
+     * @return 响应内容
+     */
+    public static String postJson(String url, String jsonData, Map<String, String> headers) {
+        try (CloseableHttpClient httpClient = HttpClients.createDefault()) {
+            HttpPost httpPost = new HttpPost(url);
+
+            // 设置请求体
+            StringEntity entity = new StringEntity(jsonData, StandardCharsets.UTF_8);
+            httpPost.setEntity(entity);
+            httpPost.setHeader("Content-Type", "application/json");
+
+            // 设置请求头
+            if (headers != null) {
+                headers.forEach(httpPost::setHeader);
+            }
+
+            return httpClient.execute(httpPost, response -> {
+                int statusCode = response.getCode();
+                String responseBody = EntityUtils.toString(response.getEntity(), StandardCharsets.UTF_8);
+
+                if (statusCode >= 200 && statusCode < 300) {
+                    return responseBody;
+                } else {
+                    throw new RuntimeException("HTTP请求失败，状态码: " + statusCode + ", 响应: " + responseBody);
+                }
+            });
+        } catch (Exception e) {
+            throw new RuntimeException("HTTP请求异常: " + e.getMessage(), e);
+        }
+    }
+
+    private static final Logger logger = LoggerFactory.getLogger(HttpClientUtils.class);
+
+    /**
+     * 下载文件（带进度日志）
+     *
+     * @param url 文件URL
+     * @return 文件字节数组
+     */
+    public static byte[] downloadFile(String url) {
+        try (CloseableHttpClient httpClient = HttpClients.createDefault()) {
+            HttpGet httpGet = new HttpGet(url);
+
+            return httpClient.execute(httpGet, response -> {
+                int statusCode = response.getCode();
+                if (statusCode == 200) {
+                    org.apache.hc.core5.http.HttpEntity entity = response.getEntity();
+                    long contentLength = entity.getContentLength();
+                    
+                    if (contentLength < 0) {
+                        logger.info("开始下载文件: {}, 大小未知", url);
+                    } else {
+                        logger.info("开始下载文件: {}, 大小: {} bytes", url, contentLength);
+                    }
+
+                    try (java.io.InputStream instream = entity.getContent();
+                         java.io.ByteArrayOutputStream buffer = new java.io.ByteArrayOutputStream()) {
+                        
+                        byte[] data = new byte[8192];
+                        int nRead;
+                        long totalRead = 0;
+                        long lastLogTime = System.currentTimeMillis();
+                        
+                        while ((nRead = instream.read(data, 0, data.length)) != -1) {
+                            buffer.write(data, 0, nRead);
+                            totalRead += nRead;
+                            
+                            long now = System.currentTimeMillis();
+                            if (now - lastLogTime > 2000) { // 每2秒打印一次日志
+                                if (contentLength > 0) {
+                                    int percent = (int) ((totalRead * 100) / contentLength);
+                                    logger.info("下载进度: {}% ({}/{})", percent, totalRead, contentLength);
+                                } else {
+                                    logger.info("已下载: {} bytes", totalRead);
+                                }
+                                lastLogTime = now;
+                            }
+                        }
+                        
+                        logger.info("下载完成，总大小: {} bytes", totalRead);
+                        return buffer.toByteArray();
+                    }
+                } else {
+                    throw new RuntimeException("文件下载失败，状态码: " + statusCode);
+                }
+            });
+        } catch (Exception e) {
+            throw new RuntimeException("文件下载异常: " + e.getMessage(), e);
+        }
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/utils/JwtUtils.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/utils/JwtUtils.java
new file mode 100644
index 0000000..4a9b506
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/utils/JwtUtils.java
@@ -0,0 +1,103 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.utils
+ * @ClassName: JwtUtils
+ * @Description: JWT工具类，提供Token生成、解析、验证等功能
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.utils;
+
+import io.jsonwebtoken.Claims;
+import io.jsonwebtoken.Jwts;
+import io.jsonwebtoken.security.Keys;
+
+import javax.crypto.SecretKey;
+import java.nio.charset.StandardCharsets;
+import java.util.Date;
+import java.util.HashMap;
+import java.util.Map;
+
+/**
+ * JWT 工具类
+ */
+public class JwtUtils {
+    
+    // JWT密钥（实际项目中应从配置文件中读取）
+    private static final String SECRET_KEY = "gdyd_zhpb_zgf_secret_key_for_jwt_token_generation_2024";
+    
+    // Token过期时间：7天
+    private static final long EXPIRATION_TIME = 7 * 24 * 60 * 60 * 1000;
+    
+    /**
+     * 生成JWT Token
+     * 
+     * @param userId 用户ID
+     * @param userName 用户名称
+     * @return Token字符串
+     */
+    public static String generateToken(String userId, String userName) {
+        Map<String, Object> claims = new HashMap<>();
+        claims.put("userId", userId);
+        claims.put("userName", userName);
+        
+        Date now = new Date();
+        Date expiration = new Date(now.getTime() + EXPIRATION_TIME);
+        
+        SecretKey key = Keys.hmacShaKeyFor(SECRET_KEY.getBytes(StandardCharsets.UTF_8));
+        
+        return Jwts.builder()
+                .claims(claims)
+                .subject(userId)
+                .issuedAt(now)
+                .expiration(expiration)
+                .signWith(key)
+                .compact();
+    }
+    
+    /**
+     * 验证并解析Token
+     * 
+     * @param token Token字符串
+     * @return Claims对象
+     */
+    public static Claims parseToken(String token) {
+        SecretKey key = Keys.hmacShaKeyFor(SECRET_KEY.getBytes(StandardCharsets.UTF_8));
+        
+        return Jwts.parser()
+                .verifyWith(key)
+                .build()
+                .parseSignedClaims(token)
+                .getPayload();
+    }
+    
+    /**
+     * 从Token中获取用户ID
+     * 
+     * @param token Token字符串
+     * @return 用户ID
+     */
+    public static String getUserIdFromToken(String token) {
+        Claims claims = parseToken(token);
+        return claims.get("userId", String.class);
+    }
+    
+    /**
+     * 检查Token是否过期
+     * 
+     * @param token Token字符串
+     * @return true表示已过期
+     */
+    public static boolean isTokenExpired(String token) {
+        try {
+            Claims claims = parseToken(token);
+            Date expiration = claims.getExpiration();
+            return expiration.before(new Date());
+        } catch (Exception e) {
+            return true;
+        }
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/utils/PdfTableExtractor.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/utils/PdfTableExtractor.java
new file mode 100644
index 0000000..50c6c61
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/utils/PdfTableExtractor.java
@@ -0,0 +1,165 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.utils
+ * @ClassName: PdfTableExtractor
+ * @Description: PDF表格提取工具类 - 使用tabula-java提取PDF中的表格并转换为结构化文本
+ * @Author: 张志锋
+ * @Date: 2025-12-11
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.utils;
+
+import org.apache.pdfbox.pdmodel.PDDocument;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+import technology.tabula.ObjectExtractor;
+import technology.tabula.Page;
+import technology.tabula.PageIterator;
+import technology.tabula.Table;
+import technology.tabula.RectangularTextContainer;
+import technology.tabula.extractors.SpreadsheetExtractionAlgorithm;
+
+import java.io.File;
+import java.util.ArrayList;
+import java.util.List;
+
+/**
+ * PDF表格提取工具类
+ * 将PDF中的表格转换为易于向量化的自然语言格式
+ */
+public class PdfTableExtractor {
+
+    private static final Logger logger = LoggerFactory.getLogger(PdfTableExtractor.class);
+
+    /**
+     * 从PDF中提取表格并转换为自然语言描述
+     *
+     * @param filePath PDF文件路径
+     * @return 表格的自然语言描述，如果没有表格则返回null
+     */
+    public static String extractTablesAsText(String filePath) {
+        StringBuilder result = new StringBuilder();
+        int tableCount = 0;
+
+        try (PDDocument document = org.apache.pdfbox.Loader.loadPDF(new File(filePath))) {
+            SpreadsheetExtractionAlgorithm algorithm = new SpreadsheetExtractionAlgorithm();
+            
+            try (ObjectExtractor extractor = new ObjectExtractor(document)) {
+                PageIterator pages = extractor.extract();
+                
+                while (pages.hasNext()) {
+                    Page page = pages.next();
+                    List<Table> tables = algorithm.extract(page);
+                    
+                    for (Table table : tables) {
+                        if (table.getRowCount() < 2) {
+                            continue; // 跳过只有一行的表格（可能只是表头）
+                        }
+                        
+                        tableCount++;
+                        String tableText = convertTableToText(table, tableCount);
+                        if (tableText != null && !tableText.isBlank()) {
+                            result.append(tableText).append("\n\n");
+                        }
+                    }
+                }
+            }
+            
+            if (tableCount > 0) {
+                logger.info("[表格提取] 从PDF中提取了 {} 个表格", tableCount);
+            } else {
+                logger.debug("[表格提取] PDF中未检测到表格");
+            }
+            
+        } catch (Exception e) {
+            logger.warn("[表格提取] 提取表格失败: {}", e.getMessage());
+        }
+
+        return result.length() > 0 ? result.toString() : null;
+    }
+
+    /**
+     * 将单个表格转换为自然语言描述
+     */
+    private static String convertTableToText(Table table, int tableIndex) {
+        StringBuilder sb = new StringBuilder();
+        
+        List<List<RectangularTextContainer>> rows = table.getRows();
+        if (rows.isEmpty()) {
+            return null;
+        }
+        
+        // 获取表头
+        List<String> headers = new ArrayList<>();
+        if (!rows.isEmpty()) {
+            for (RectangularTextContainer cell : rows.get(0)) {
+                String text = cell.getText().trim().replaceAll("\\s+", " ");
+                headers.add(text.isEmpty() ? "列" + (headers.size() + 1) : text);
+            }
+        }
+        
+        // 如果表头看起来不像表头（全是数字或空），尝试生成通用表头
+        boolean hasValidHeaders = headers.stream().anyMatch(h -> h.length() > 0 && !h.matches("\\d+"));
+        
+        // 处理数据行（从第2行开始，如果有有效表头的话）
+        int startRow = hasValidHeaders ? 1 : 0;
+        int rowNum = 0;
+        
+        for (int i = startRow; i < rows.size(); i++) {
+            List<RectangularTextContainer> row = rows.get(i);
+            rowNum++;
+            
+            // 构建该行的自然语言描述
+            StringBuilder rowDesc = new StringBuilder();
+            rowDesc.append("第").append(rowNum).append("项: ");
+            
+            List<String> items = new ArrayList<>();
+            for (int j = 0; j < row.size() && j < headers.size(); j++) {
+                String cellText = row.get(j).getText().trim().replaceAll("\\s+", " ");
+                if (!cellText.isEmpty()) {
+                    String header = j < headers.size() ? headers.get(j) : "字段" + (j + 1);
+                    items.add(header + "为\"" + cellText + "\"");
+                }
+            }
+            
+            if (!items.isEmpty()) {
+                rowDesc.append(String.join("，", items));
+                sb.append(rowDesc).append("。\n");
+            }
+        }
+        
+        if (sb.length() == 0) {
+            return null;
+        }
+        
+        // 添加表格标题
+        StringBuilder result = new StringBuilder();
+        result.append("【表格").append(tableIndex).append("】\n");
+        result.append(sb);
+        
+        return result.toString();
+    }
+
+    /**
+     * 提取表格并合并到原有文本中
+     *
+     * @param filePath PDF文件路径
+     * @param originalText 原有的PDF文本内容
+     * @return 合并后的文本（表格内容追加在后面）
+     */
+    public static String enhanceTextWithTables(String filePath, String originalText) {
+        String tableText = extractTablesAsText(filePath);
+        
+        if (tableText == null || tableText.isBlank()) {
+            return originalText;
+        }
+        
+        StringBuilder enhanced = new StringBuilder();
+        enhanced.append(originalText);
+        enhanced.append("\n\n=== 表格结构化内容 ===\n\n");
+        enhanced.append(tableText);
+        
+        return enhanced.toString();
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/utils/TaskStatusUtils.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/utils/TaskStatusUtils.java
new file mode 100644
index 0000000..d8caffa
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/utils/TaskStatusUtils.java
@@ -0,0 +1,67 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.utils
+ * @ClassName: TaskStatusUtils
+ * @Description: 任务状态工具类
+ * @Author: 张志锋
+ * @Date: 2025-12-02
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.utils;
+
+/**
+ * 任务状态工具类
+ * 用于状态的中英文转换
+ */
+public class TaskStatusUtils {
+
+    /**
+     * 将英文状态转换为"英文:中文"格式
+     * @param status 英文状态
+     * @return "英文:中文"格式的状态
+     */
+    public static String toChineseStatus(String status) {
+        if (status == null) {
+            return "unknown:未知";
+        }
+        
+        String lowerStatus = status.toLowerCase();
+        switch (lowerStatus) {
+            case "pending":
+                return "processing";
+            case "processing":
+                return "processing";
+            case "completed":
+                return "completed";
+            case "failed":
+                return "failed";
+            default:
+                return status + ":未知";
+        }
+    }
+
+    /**
+     * 将中文状态转换为英文
+     * @param chineseStatus 中文状态
+     * @return 英文状态
+     */
+    public static String toEnglishStatus(String chineseStatus) {
+        if (chineseStatus == null) {
+            return "unknown";
+        }
+        
+        switch (chineseStatus) {
+            case "待处理":
+                return "pending";
+            case "处理中":
+                return "processing";
+            case "已完成":
+                return "completed";
+            case "失败":
+                return "failed";
+            default:
+                return chineseStatus;
+        }
+    }
+}
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/utils/VectorizeUtils.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/utils/VectorizeUtils.java
new file mode 100644
index 0000000..7fe6640
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/utils/VectorizeUtils.java
@@ -0,0 +1,350 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf.utils
+ * @ClassName: VectorizeUtils
+ * @Description: 向量化工具类，调用外部向量化接口将文本转换为向量
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf.utils;
+
+import com.alibaba.fastjson2.JSON;
+import com.alibaba.fastjson2.JSONObject;
+import org.apache.hc.client5.http.classic.methods.HttpPost;
+import org.apache.hc.client5.http.impl.classic.CloseableHttpClient;
+import org.apache.hc.client5.http.impl.classic.HttpClients;
+import org.apache.hc.core5.http.io.entity.EntityUtils;
+import org.apache.hc.core5.http.io.entity.StringEntity;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+import java.nio.charset.StandardCharsets;
+import java.util.List;
+
+/**
+ * 向量化工具类
+ */
+public class VectorizeUtils {
+
+    private static final Logger logger = LoggerFactory.getLogger(VectorizeUtils.class);
+
+    /**
+     * 调用向量化接口将文本转换为向量
+     *
+     * @param text 要向量化的文本
+     * @param baseUrl 向量化服务基础URL
+     * @param apiPath 向量化接口路径
+     * @return 向量化结果，包含vector数组和dimension维度
+     * @throws Exception 向量化失败时抛出异常
+     */
+    public static VectorizeResult vectorize(String text, String baseUrl, String apiPath) throws Exception {
+        return vectorize(text, baseUrl, apiPath, null, null);
+    }
+
+    /**
+     * 调用向量化接口将文本转换为向量（支持 OpenAI 格式）
+     *
+     * @param text 要向量化的文本
+     * @param baseUrl 向量化服务基础URL
+     * @param apiPath 向量化接口路径
+     * @param apiKey API密钥（可选，用于 OpenAI 格式）
+     * @param model 模型名称（可选，用于 OpenAI 格式）
+     * @return 向量化结果，包含vector数组和dimension维度
+     * @throws Exception 向量化失败时抛出异常
+     */
+    public static VectorizeResult vectorize(String text, String baseUrl, String apiPath, String apiKey, String model) throws Exception {
+        if (text == null || text.trim().isEmpty()) {
+            throw new IllegalArgumentException("文本内容不能为空");
+        }
+
+        String url = baseUrl.endsWith("/") ? baseUrl + apiPath.substring(1) : baseUrl + apiPath;
+
+        logger.debug("调用向量化接口: {}", url);
+        logger.debug("文本长度: {} 字符", text.length());
+
+        // 配置HTTP客户端，设置超时时间（30秒连接，60秒响应）
+        org.apache.hc.core5.util.Timeout connectTimeout = org.apache.hc.core5.util.Timeout.ofSeconds(30);
+        org.apache.hc.core5.util.Timeout responseTimeout = org.apache.hc.core5.util.Timeout.ofSeconds(60);
+        org.apache.hc.client5.http.config.RequestConfig requestConfig = org.apache.hc.client5.http.config.RequestConfig.custom()
+                .setConnectTimeout(connectTimeout)
+                .setResponseTimeout(responseTimeout)
+                .build();
+        
+        long startTime = System.currentTimeMillis();
+        try (CloseableHttpClient httpClient = HttpClients.custom()
+                .setDefaultRequestConfig(requestConfig)
+                .build()) {
+            HttpPost httpPost = new HttpPost(url);
+            httpPost.setHeader("Content-Type", "application/json;charset=UTF-8");
+            
+            // 如果提供了 API Key，添加 Authorization header（OpenAI 格式）
+            if (apiKey != null && !apiKey.trim().isEmpty()) {
+                httpPost.setHeader("Authorization", "Bearer " + apiKey.trim());
+            }
+
+            // 构建请求体
+            JSONObject requestBody = new JSONObject();
+            
+            // 判断是否为 OpenAI 格式（如果提供了 model 参数）
+            if (model != null && !model.trim().isEmpty()) {
+                // OpenAI 格式
+                requestBody.put("model", model);
+                requestBody.put("input", text);
+            } else {
+                // 原有格式
+                requestBody.put("text", text);
+            }
+            
+            String requestJson = requestBody.toJSONString();
+            httpPost.setEntity(new StringEntity(requestJson, StandardCharsets.UTF_8));
+
+            logger.info("调用向量化接口: {}, 文本长度: {} 字符", url, text.length());
+
+            // 执行请求
+            return httpClient.execute(httpPost, response -> {
+                int statusCode = response.getCode();
+                String responseBody = EntityUtils.toString(response.getEntity(), StandardCharsets.UTF_8);
+
+                if (statusCode != 200) {
+                    String errorMsg = String.format("向量化接口调用失败，URL: %s, 状态码: %d, 响应: %s", url, statusCode, responseBody);
+                    logger.error(errorMsg);
+                    throw new VectorizeRequestException(
+                        errorMsg, requestJson, responseBody, url, apiKey, model,
+                        System.currentTimeMillis() - startTime, statusCode, null);
+                }
+
+                // 解析响应
+                JSONObject responseJson = JSON.parseObject(responseBody);
+                
+                // 判断响应格式：OpenAI 格式或原有格式
+                List<Number> vectorList = null;
+                Integer dimension = null;
+                
+                if (responseJson.containsKey("data") && responseJson.getJSONArray("data") != null) {
+                    // OpenAI 格式：{"data": [{"embedding": [...]}]}
+                    JSONObject dataItem = responseJson.getJSONArray("data").getJSONObject(0);
+                    if (dataItem != null && dataItem.containsKey("embedding")) {
+                        vectorList = dataItem.getList("embedding", Number.class);
+                        dimension = vectorList != null ? vectorList.size() : null;
+                    }
+                } else if (responseJson.containsKey("vector")) {
+                    // 原有格式：{"vector": [...], "dimension": ...}
+                    vectorList = responseJson.getList("vector", Number.class);
+                    dimension = responseJson.getInteger("dimension");
+                }
+                
+                if (vectorList == null || vectorList.isEmpty()) {
+                    String errorMsg = String.format("向量化接口返回格式错误，无法解析向量数据，URL: %s, 响应: %s", url, responseBody);
+                    logger.error(errorMsg);
+                    throw new RuntimeException(errorMsg);
+                }
+                
+                float[] vector = new float[vectorList.size()];
+                for (int i = 0; i < vectorList.size(); i++) {
+                    vector[i] = vectorList.get(i).floatValue();
+                }
+
+                // 获取维度
+                if (dimension == null) {
+                    dimension = vector.length;
+                }
+
+                Integer promptTokens = null;
+                Integer completionTokens = null;
+                Integer totalTokens = null;
+                JSONObject usage = responseJson.getJSONObject("usage");
+                if (usage != null) {
+                    promptTokens = usage.getInteger("prompt_tokens");
+                    completionTokens = usage.getInteger("completion_tokens");
+                    totalTokens = usage.getInteger("total_tokens");
+                }
+
+                logger.info("向量化成功，URL: {}, 维度: {}", url, dimension);
+
+                return new VectorizeResult(
+                    vector,
+                    dimension,
+                    requestJson,
+                    responseBody,
+                    url,
+                    apiKey,
+                    model,
+                    System.currentTimeMillis() - startTime,
+                    promptTokens,
+                    completionTokens,
+                    totalTokens
+                );
+            });
+        } catch (java.net.ConnectException e) {
+            String errorMsg = String.format("无法连接到向量化服务，URL: %s, 错误: %s。请检查向量化服务是否已启动（默认地址: http://localhost:8001）", url, e.getMessage());
+            logger.error(errorMsg, e);
+            throw new VectorizeRequestException(errorMsg, null, null, url, apiKey, model,
+                System.currentTimeMillis() - startTime, null, e);
+        } catch (org.apache.hc.core5.http.ConnectionClosedException e) {
+            String errorMsg = String.format("向量化服务连接已关闭，URL: %s, 错误: %s", url, e.getMessage());
+            logger.error(errorMsg, e);
+            throw new VectorizeRequestException(errorMsg, null, null, url, apiKey, model,
+                System.currentTimeMillis() - startTime, null, e);
+        } catch (java.net.SocketTimeoutException e) {
+            String errorMsg = String.format("向量化接口请求超时，URL: %s, 错误: %s", url, e.getMessage());
+            logger.error(errorMsg, e);
+            throw new VectorizeRequestException(errorMsg, null, null, url, apiKey, model,
+                System.currentTimeMillis() - startTime, null, e);
+        } catch (java.io.IOException e) {
+            String errorMsg = String.format("向量化接口网络IO异常，URL: %s, 错误: %s", url, e.getMessage());
+            logger.error(errorMsg, e);
+            throw new VectorizeRequestException(errorMsg, null, null, url, apiKey, model,
+                System.currentTimeMillis() - startTime, null, e);
+        } catch (Exception e) {
+            String errorMsg = String.format("向量化接口调用异常，URL: %s, 错误: %s", url, e.getMessage());
+            logger.error(errorMsg, e);
+            throw new VectorizeRequestException(errorMsg, null, null, url, apiKey, model,
+                System.currentTimeMillis() - startTime, null, e);
+        }
+    }
+
+    /**
+     * 向量化结果
+     */
+    public static class VectorizeResult {
+        private final float[] vector;
+        private final Integer dimension;
+        private final String requestPayload;
+        private final String responsePayload;
+        private final String requestUrl;
+        private final String apiKey;
+        private final String model;
+        private final Long durationMs;
+        private final Integer promptTokens;
+        private final Integer completionTokens;
+        private final Integer totalTokens;
+
+        public VectorizeResult(float[] vector,
+                               Integer dimension,
+                               String requestPayload,
+                               String responsePayload,
+                               String requestUrl,
+                               String apiKey,
+                               String model,
+                               Long durationMs,
+                               Integer promptTokens,
+                               Integer completionTokens,
+                               Integer totalTokens) {
+            this.vector = vector;
+            this.dimension = dimension;
+            this.requestPayload = requestPayload;
+            this.responsePayload = responsePayload;
+            this.requestUrl = requestUrl;
+            this.apiKey = apiKey;
+            this.model = model;
+            this.durationMs = durationMs;
+            this.promptTokens = promptTokens;
+            this.completionTokens = completionTokens;
+            this.totalTokens = totalTokens;
+        }
+
+        public float[] getVector() {
+            return vector;
+        }
+
+        public Integer getDimension() {
+            return dimension;
+        }
+
+        public String getRequestPayload() {
+            return requestPayload;
+        }
+
+        public String getResponsePayload() {
+            return responsePayload;
+        }
+
+        public String getRequestUrl() {
+            return requestUrl;
+        }
+
+        public String getApiKey() {
+            return apiKey;
+        }
+
+        public String getModel() {
+            return model;
+        }
+
+        public Long getDurationMs() {
+            return durationMs;
+        }
+
+        public Integer getPromptTokens() {
+            return promptTokens;
+        }
+
+        public Integer getCompletionTokens() {
+            return completionTokens;
+        }
+
+        public Integer getTotalTokens() {
+            return totalTokens;
+        }
+    }
+
+    public static class VectorizeRequestException extends RuntimeException {
+        private final String requestPayload;
+        private final String responsePayload;
+        private final String requestUrl;
+        private final String apiKey;
+        private final String model;
+        private final Long durationMs;
+        private final Integer statusCode;
+
+        public VectorizeRequestException(String message,
+                                         String requestPayload,
+                                         String responsePayload,
+                                         String requestUrl,
+                                         String apiKey,
+                                         String model,
+                                         Long durationMs,
+                                         Integer statusCode,
+                                         Throwable cause) {
+            super(message, cause);
+            this.requestPayload = requestPayload;
+            this.responsePayload = responsePayload;
+            this.requestUrl = requestUrl;
+            this.apiKey = apiKey;
+            this.model = model;
+            this.durationMs = durationMs;
+            this.statusCode = statusCode;
+        }
+
+        public String getRequestPayload() {
+            return requestPayload;
+        }
+
+        public String getResponsePayload() {
+            return responsePayload;
+        }
+
+        public String getRequestUrl() {
+            return requestUrl;
+        }
+
+        public String getApiKey() {
+            return apiKey;
+        }
+
+        public String getModel() {
+            return model;
+        }
+
+        public Long getDurationMs() {
+            return durationMs;
+        }
+
+        public Integer getStatusCode() {
+            return statusCode;
+        }
+    }
+}
+
diff --git a/src/main/java/com/zhpb/gdyd_zhpb_zgf/web/FriendlyErrorController.java b/src/main/java/com/zhpb/gdyd_zhpb_zgf/web/FriendlyErrorController.java
new file mode 100644
index 0000000..ce72ec1
--- /dev/null
+++ b/src/main/java/com/zhpb/gdyd_zhpb_zgf/web/FriendlyErrorController.java
@@ -0,0 +1,150 @@
+package com.zhpb.gdyd_zhpb_zgf.web;
+
+import jakarta.servlet.http.HttpServletRequest;
+import org.springframework.boot.web.error.ErrorAttributeOptions;
+import org.springframework.boot.web.error.ErrorAttributeOptions.Include;
+import org.springframework.boot.web.servlet.error.ErrorAttributes;
+import org.springframework.boot.web.servlet.error.ErrorController;
+import org.springframework.http.HttpHeaders;
+import org.springframework.http.HttpStatus;
+import org.springframework.http.MediaType;
+import org.springframework.http.ResponseEntity;
+import org.springframework.lang.NonNull;
+import org.springframework.stereotype.Controller;
+import org.springframework.web.bind.annotation.RequestMapping;
+import org.springframework.web.context.request.ServletWebRequest;
+
+import java.time.OffsetDateTime;
+import java.time.format.DateTimeFormatter;
+import java.util.HashMap;
+import java.util.Locale;
+import java.util.Map;
+
+/**
+ * 自定义错误控制器：在数据库等常见错误发生时，用中文进行友好提示，避免晦涩的堆栈信息。
+ */
+@Controller
+@RequestMapping("${server.error.path:${error.path:/error}}")
+public class FriendlyErrorController implements ErrorController {
+
+    private final ErrorAttributes errorAttributes;
+
+    public FriendlyErrorController(ErrorAttributes errorAttributes) {
+        this.errorAttributes = errorAttributes;
+    }
+
+    @RequestMapping(produces = MediaType.TEXT_HTML_VALUE)
+    public ResponseEntity<String> errorHtml(@NonNull HttpServletRequest request) {
+        ErrorViewModel model = buildModel(request);
+        String html = renderHtml(model);
+        return ResponseEntity.status(model.status().value())
+                .header(HttpHeaders.CONTENT_TYPE, MediaType.TEXT_HTML_VALUE)
+                .body(html);
+    }
+
+    @RequestMapping(produces = MediaType.APPLICATION_JSON_VALUE)
+    public ResponseEntity<Map<String, Object>> errorJson(@NonNull HttpServletRequest request) {
+        ErrorViewModel model = buildModel(request);
+        Map<String, Object> body = new HashMap<>();
+        body.put("status", model.status().value());
+        body.put("error", model.status().getReasonPhrase());
+        body.put("message", model.friendlyMessage());
+        body.put("timestamp", model.timestamp());
+        body.put("path", model.path());
+        return ResponseEntity.status(model.status().value()).body(body);
+    }
+
+    private ErrorViewModel buildModel(@NonNull HttpServletRequest request) {
+        ServletWebRequest webRequest = new ServletWebRequest(request);
+        Map<String, Object> attrs = errorAttributes.getErrorAttributes(webRequest,
+                ErrorAttributeOptions.defaults().including(Include.MESSAGE));
+        Throwable error = errorAttributes.getError(webRequest);
+
+        int statusCode = (int) attrs.getOrDefault("status", 500);
+        HttpStatus status = HttpStatus.resolve(statusCode);
+        if (status == null) {
+            status = HttpStatus.INTERNAL_SERVER_ERROR;
+        }
+        String path = (String) attrs.getOrDefault("path", request.getRequestURI());
+
+        String rawMessage = extractRawMessage(attrs, error);
+        String friendly = translate(rawMessage, status);
+
+        return new ErrorViewModel(status, friendly, path,
+                attrs.getOrDefault("timestamp", OffsetDateTime.now().format(DateTimeFormatter.ISO_OFFSET_DATE_TIME)).toString());
+    }
+
+    private String extractRawMessage(Map<String, Object> attrs, Throwable error) {
+        StringBuilder sb = new StringBuilder();
+        if (attrs.get("error") != null) {
+            sb.append(attrs.get("error")).append(" ");
+        }
+        if (attrs.get("message") != null) {
+            sb.append(attrs.get("message")).append(" ");
+        }
+        if (error != null && error.getMessage() != null) {
+            sb.append(error.getMessage());
+        }
+        return sb.toString();
+    }
+
+    private String translate(String origin, HttpStatus status) {
+        String lower = origin == null ? "" : origin.toLowerCase(Locale.ROOT);
+        if (lower.contains("communications link failure") || lower.contains("connection refused")
+                || lower.contains("connect timed out") || lower.contains("could not open jdbc connection")) {
+            return "无法连接到数据库，请检查数据库是否启动、账号密码是否正确、网络是否可达。";
+        }
+        if (lower.contains("password authentication failed") || lower.contains("role") && lower.contains("does not exist")) {
+            return "数据库账号或密码错误，请核对配置中的数据库凭证。";
+        }
+        if (lower.contains("database") && lower.contains("does not exist")) {
+            return "数据库不存在，请确认数据库名称是否正确并已创建。";
+        }
+        if (status.is4xxClientError()) {
+            return "请求参数有误或资源不存在，请核对输入后重试。";
+        }
+        if (status.is5xxServerError()) {
+            return "服务内部出现异常，请稍后再试或联系系统管理员。";
+        }
+        return "系统出现未知错误，请稍后再试。";
+    }
+
+    private String renderHtml(ErrorViewModel model) {
+        return """
+                <!DOCTYPE html>
+                <html lang="zh-CN">
+                <head>
+                    <meta charset="UTF-8">
+                    <title>系统提示</title>
+                    <style>
+                        body {font-family: "Microsoft YaHei", sans-serif; background: #0f172a; color: #e2e8f0; margin:0; padding:0;}
+                        .wrapper {max-width: 520px; margin: 10%% auto; background: #1e293b; border-radius: 16px; padding: 32px; text-align: center;
+                                  box-shadow: 0 20px 60px rgba(15,23,42,0.4);}
+                        h1 {font-size: 26px; margin-bottom: 12px;}
+                        p {font-size: 15px; line-height: 1.7;}
+                        .status {display:inline-block; padding: 6px 14px; border-radius: 999px; background: rgba(248,113,113,.2);
+                                 color: #fecaca; font-weight: 600; margin-bottom: 18px;}
+                        .meta {margin-top: 18px; font-size: 13px; color: #94a3b8;}
+                        .btn {display:inline-block; margin-top:20px; padding:10px 18px; border-radius:999px;
+                              background:#334155; color:#fff; text-decoration:none;}
+                    </style>
+                </head>
+                <body>
+                    <div class="wrapper">
+                        <div class="status">状态码 %s</div>
+                        <h1>系统提示</h1>
+                        <p>%s</p>
+                        <div class="meta">
+                            <div>路径：%s</div>
+                            <div>时间：%s</div>
+                        </div>
+                        <a class="btn" href="/" >返回首页</a>
+                    </div>
+                </body>
+                </html>
+                """.formatted(model.status().value(), model.friendlyMessage(), model.path(), model.timestamp());
+    }
+
+    private record ErrorViewModel(HttpStatus status, String friendlyMessage, String path, String timestamp) {}
+}
+
diff --git a/src/main/resources/application.yml b/src/main/resources/application.yml
new file mode 100644
index 0000000..ff6ade6
--- /dev/null
+++ b/src/main/resources/application.yml
@@ -0,0 +1,190 @@
+server:
+  port: 8000
+
+spring:
+  application:
+    name: gdyd_zhpb_zgf
+  
+  # PostgreSQL 单数据源配置
+  datasource:
+      driver-class-name: org.postgresql.Driver
+      url: jdbc:postgresql://45.192.102.202:32771/vector_db
+      username: postgres
+      password: 123456
+      # HikariCP 连接池配置
+      hikari:
+        maximum-pool-size: 50  # 最大连接数
+        minimum-idle: 10       # 最小空闲连接数
+        idle-timeout: 300000   # 空闲连接超时时间(5分钟)
+        max-lifetime: 1800000  # 连接最大存活时间(30分钟)
+        connection-timeout: 60000  # 连接超时时间(60秒)
+  
+  # JPA 配置（全部实体使用 PostgreSQL）
+  jpa:
+    database: postgresql
+    hibernate:
+      ddl-auto: update
+    show-sql: false
+    properties:
+      hibernate:
+        dialect: org.hibernate.dialect.PostgreSQLDialect
+        jdbc.time_zone: Asia/Shanghai
+  
+  # Redis 配置
+  data:
+    redis:
+      host: 45.192.102.202
+      port: 6379
+      password: 123456
+      database: 0
+      timeout: 6000ms
+      lettuce:
+        pool:
+          max-active: 8
+          max-wait: -1ms
+          max-idle: 8
+          min-idle: 0
+  
+  # 文件上传配置
+  servlet:
+    multipart:
+      # 单个文件最大大小 (500MB)
+      max-file-size: 500MB
+      # 整个请求最大大小 (500MB)
+      max-request-size: 500MB
+      # 文件写入磁盘的阈值 (10MB，超过此大小会写入临时文件)
+      file-size-threshold: 10MB
+
+# Umi-OCR 配置
+umi-ocr:
+  base-url: http://154.219.106.93:1224
+  enabled: false  # OCR功能开关，设为true启用，false禁用
+  options:
+    ocr-language: "models/config_chinese.txt"
+    max-pages: 50
+    interval: 5
+  download-file-types:
+    - "pdfLayered"
+    - "txtPlain"
+
+# SSO配置
+sso:
+  # ES接口基础地址
+  es-base-url: http://es-integration.es-uat-paas:8890
+  # 获取用户信息接口路径
+  getUserInfoPath: /bi/getUserInfoByTicket
+
+# 文件管理配置
+file:
+  # 文件上传存储路径
+  upload-path: ./uploads
+  # ZIP文件解压路径
+  extract-path: ./extracts
+  # 支持的文件类型
+  allowed-types: pdf,doc,docx,txt
+  # 是否只返回技术文件类型（PDF、DOC、DOCX、PPT、PPTX、TXT、JSON）
+  filter:
+    technical-only: true
+  # 文件下载地址替换配置
+  download-replace-source: 172.21.200.148:2000
+  download-replace-target: wnsp-uat-fw-gw.wnsp-fw-gw:8099
+
+# 向量化服务配置
+vectorize:
+  # 向量化接口基础地址
+  base-url: https://api.siliconflow.cn
+  # 向量化接口路径（OpenAI 格式）
+  api-path: /v1/embeddings
+  # API密钥（用于 OpenAI 格式）
+  api-key: sk-vppcdrlrltizbwluttekgorrngofgounxyxqaheystfmbelz
+  # 向量化模型（OpenAI 格式，如 text-embedding-ada-002）
+  model: BAAI/bge-large-zh-v1.5
+  # 向量维度（从接口返回的dimension字段获取，默认1536 for text-embedding-ada-002）
+  dimension: 1024
+  # 向量化并发数（控制同时处理的文件数量，建议设置为2-4，避免耗尽数据库连接）
+  concurrency: 3
+
+
+# 向量化模型阈值配置（按模型区分）
+# 注意：SiliconFlow BGE模型限制512 tokens，这里设置450留出余量
+vector-limits:
+  default:
+    max-chars: 800
+    max-tokens: 450
+  text-embedding-ada-002:
+    max-chars: 8000
+    max-tokens: 8000
+  embed-bge-large-zh-v1.5:
+    max-chars: 800
+    max-tokens: 450
+  BAAI/bge-large-zh-v1.5:
+    max-chars: 800
+    max-tokens: 450
+
+# Swagger/OpenAPI 配置
+springdoc:
+  api-docs:
+    path: /api-docs
+    enabled: true
+  swagger-ui:
+    path: /swagger-ui.html
+    enabled: true
+    tags-sorter: alpha
+    operations-sorter: alpha
+    try-it-out-enabled: true
+    display-request-duration: true
+  show-actuator: false
+  # 排除异常处理器包，避免兼容性问题
+  packages-to-exclude: com.zhpb.gdyd_zhpb_zgf.exception
+  # 不扫描ControllerAdvice类
+  use-management-port: false
+
+# Actuator 配置
+management:
+  endpoints:
+    web:
+      exposure:
+        include: health,info
+  endpoint:
+    health:
+      show-details: always
+  health:
+    indicators:
+      enabled: true
+
+# AI模型配置
+ai:
+  model-type: public  # public: 公网模型(DeepSeek), private: 私有模型
+
+# 向量化模型配置
+vectorization:
+  mode: local  # local: 本地处理, online: 在线API, office: 办公网关
+  batch-size: 10  # 批量处理大小
+
+# DeepSeek AI 配置 (公网模型)
+deepseek:
+  api-key: sk-a6c394f29fd241f6b760d31986d95776  # 建议通过环境变量设置
+  api-url: https://dashscope.aliyuncs.com/compatible-mode/v1/chat/completions
+  model: qwen3-max
+  temperature: 0.7
+  max-tokens: 4096
+
+# 私有大模型配置
+private-model:
+  url: http://IP:PORT/bigmodel_infer_gateway/v1/service  # 私有模型API地址
+  appid: ""  # 私有模型AppID
+  appkey: ""  # 私有模型AppKey
+  capabilityname: semantic0000000000000000  # 能力名称
+  model: JiuTian-75B-8K  # 私有模型名称
+  temperature: 0.1  # 温度参数
+  max-tokens: 1024  # 最大token数
+
+# 办公网关配置 (磐智宁波池)
+office-gateway:
+  uat-url: http://172.21.153.46:9050/micro-model-gateway/v1/service/uat  # UAT环境
+  prod-url: http://prod-gateway-url  # 生产环境URL
+  appid: modelway
+  appkey: 4f34396a9b876346133883461caf6596
+  capabilityname: semantic0000000000000000  # 能力名称，24位
+  embedding-model: embed-bge-large-zh-v1.5  # 向量化模型
+  timeout: 30000  # 请求超时时间(毫秒)
diff --git a/src/main/resources/banner.txt b/src/main/resources/banner.txt
new file mode 100644
index 0000000..38ab5fb
--- /dev/null
+++ b/src/main/resources/banner.txt
@@ -0,0 +1,16 @@
+
+
+  .--.--.              ____                               ___
+ /  /    '.          ,'  , `.                           ,--.'|_                            ,-.----.   ,-.----.
+|  :  /`. /       ,-+-,.' _ |                 __  ,-.   |  | :,'      ,---,.               \    /  \  \    /  \
+;  |  |--`     ,-+-. ;   , ||               ,' ,'/ /|   :  : ' :    ,'  .' |               |   :    | |   :    |
+|  :  ;_      ,--.'|'   |  ||    ,--.--.    '  | |' | .;__,'  /   ,---.'   ,    ,--.--.    |   | .\ : |   | .\ :
+ \  \    `.  |   |  ,', |  |,   /       \   |  |   ,' |  |   |    |   |    |   /       \   .   : |: | .   : |: |
+  `----.   \ |   | /  | |--'   .--.  .-. |  '  :  /   :__,'| :    :   :  .'   .--.  .-. |  |   |  \ : |   |  \ :
+  __ \  \  | |   : |  | ,       \__\/: . .  |  | '      '  : |__  :   |.'      \__\/: . .  |   : .  | |   : .  |
+ /  /`--'  / |   : |  |/        ," .--.; |  ;  : |      |  | '.'| `---'        ," .--.; |  :     |`-' :     |`-'
+'--'.     /  |   | |`-'        /  /  ,.  |  |  , ;      ;  :    ;             /  /  ,.  |  :   : :    :   : :
+  `--'---'   |   ;/           ;  :   .'   \  ---'       |  ,   /             ;  :   .'   \ |   | :    |   | :
+             '---'            |  ,     .-./              ---`-'              |  ,     .-./ `---'.|    `---'.|
+                               `--`---'                                       `--`---'       `---`      `---`
+
diff --git a/src/main/resources/db/migration/V1.1__create_prompt_template_table.sql b/src/main/resources/db/migration/V1.1__create_prompt_template_table.sql
new file mode 100644
index 0000000..862f917
--- /dev/null
+++ b/src/main/resources/db/migration/V1.1__create_prompt_template_table.sql
@@ -0,0 +1,260 @@
+-- 创建提示词模板表
+CREATE TABLE `t_prompt_template` (
+    `id` bigint(20) NOT NULL AUTO_INCREMENT COMMENT '主键ID',
+    `template_id` varchar(64) NOT NULL COMMENT '模板ID（业务ID，唯一）',
+    `template_name` varchar(256) NOT NULL COMMENT '模板名称',
+    `template_type` varchar(32) NOT NULL COMMENT '模板类型（qa-智能问答, summary-摘要, etc）',
+    `prompt_content` text NOT NULL COMMENT '提示词内容',
+    `description` text COMMENT '模板描述',
+    `is_active` tinyint(1) NOT NULL DEFAULT 1 COMMENT '是否启用',
+    `sort_order` int(11) NOT NULL DEFAULT 0 COMMENT '排序号',
+    `created_by` varchar(64) DEFAULT NULL COMMENT '创建人',
+    `updated_by` varchar(64) DEFAULT NULL COMMENT '更新人',
+    `create_time` datetime NOT NULL COMMENT '创建时间',
+    `update_time` datetime DEFAULT NULL COMMENT '更新时间',
+    PRIMARY KEY (`id`),
+    UNIQUE KEY `uk_template_id` (`template_id`),
+    UNIQUE KEY `uk_template_name` (`template_name`),
+    KEY `idx_template_type` (`template_type`),
+    KEY `idx_is_active` (`is_active`),
+    KEY `idx_sort_order` (`sort_order`)
+) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_unicode_ci COMMENT='提示词模板表';
+
+-- 初始化评标助手智能问答提示词模板
+INSERT INTO `t_prompt_template` (
+    `template_id`,
+    `template_name`,
+    `template_type`,
+    `prompt_content`,
+    `description`,
+    `is_active`,
+    `sort_order`,
+    `created_by`,
+    `create_time`
+) VALUES (
+    'tender-qa-general-001',
+    '通用评标问答',
+    'qa',
+    '你是一个专业的招标评审专家助手。请基于以下提供的招标文件内容，准确、客观地回答评审专家的问题。
+
+招标文件内容：
+{document_content}
+
+评审问题：{query}
+
+请遵循以下原则：
+1. 严格基于招标文件内容回答，不添加外部知识或个人观点
+2. 引用具体条款、页码或章节，便于核实
+3. 对于技术参数、商务条件要准确表述
+4. 如有多个相关条款，应进行归纳整理
+5. 对于文件未明确的内容，应明确说明"招标文件中未明确规定"
+6. 保持专业、客观、中立的态度
+
+评审意见：',
+    '通用评标问答提示词，适用于各类招标项目的评审问答',
+    1,
+    0,
+    'system',
+    NOW()
+);
+
+INSERT INTO `t_prompt_template` (
+    `template_id`,
+    `template_name`,
+    `template_type`,
+    `prompt_content`,
+    `description`,
+    `is_active`,
+    `sort_order`,
+    `created_by`,
+    `create_time`
+) VALUES (
+    'tender-qa-technical-002',
+    '技术评审问答',
+    'qa',
+    '你是一个专业的技术评审专家。请基于以下技术规格书和相关技术文档，回答技术评审相关问题。
+
+技术文档内容：
+{document_content}
+
+技术问题：{query}
+
+技术评审要点：
+1. 严格按照技术规格书要求进行评审
+2. 重点关注技术参数、性能指标、质量标准
+3. 识别技术风险和潜在问题
+4. 评估技术方案的可行性和先进性
+5. 检查技术指标是否满足最低要求
+6. 对于技术偏离应明确指出影响程度
+
+技术评审意见：
+- 符合性分析：
+- 技术风险评估：
+- 建议意见：',
+    '技术评审专用提示词，重点关注技术参数、性能指标等技术评审内容',
+    1,
+    1,
+    'system',
+    NOW()
+);
+
+INSERT INTO `t_prompt_template` (
+    `template_id`,
+    `template_name`,
+    `template_type`,
+    `prompt_content`,
+    `description`,
+    `is_active`,
+    `sort_order`,
+    `created_by`,
+    `create_time`
+) VALUES (
+    'tender-qa-commercial-003',
+    '商务评审问答',
+    'qa',
+    '你是一个专业的商务评审专家。请基于以下商务文件和报价文件，回答商务评审相关问题。
+
+商务文件内容：
+{document_content}
+
+商务问题：{query}
+
+商务评审要点：
+1. 严格按照招标文件商务要求进行评审
+2. 重点关注报价合理性、商务条件、付款方式
+3. 检查商务偏差及其影响
+4. 评估供应商的商务实力和信誉
+5. 审核商务条款的合规性和可行性
+6. 计算商务得分并给出评分理由
+
+商务评审意见：
+- 报价分析：
+- 商务条件评估：
+- 合规性检查：
+- 评分建议：',
+    '商务评审专用提示词，重点关注报价、商务条件、付款方式等商务评审内容',
+    1,
+    2,
+    'system',
+    NOW()
+);
+
+INSERT INTO `t_prompt_template` (
+    `template_id`,
+    `template_name`,
+    `template_type`,
+    `prompt_content`,
+    `description`,
+    `is_active`,
+    `sort_order`,
+    `created_by`,
+    `create_time`
+) VALUES (
+    'tender-qa-compliance-004',
+    '合规性检查问答',
+    'qa',
+    '你是一个专业的招标合规性审查专家。请基于招标文件和相关法律法规，回答合规性相关问题。
+
+招标文件及法律法规内容：
+{document_content}
+
+合规性问题：{query}
+
+合规性审查要点：
+1. 检查投标文件是否符合招标文件要求
+2. 审核供应商资质和相关证书
+3. 验证投标保证金、履约保证金等保证措施
+4. 审查投标文件完整性和有效性
+5. 识别潜在的违规行为和风险
+6. 确认投标有效期和投标截止时间
+
+合规性审查意见：
+- 资格审查：
+- 文件完整性：
+- 保证措施：
+- 违规风险：
+- 审查结论：',
+    '合规性检查专用提示词，重点关注投标文件的合规性和有效性审查',
+    1,
+    3,
+    'system',
+    NOW()
+);
+
+INSERT INTO `t_prompt_template` (
+    `template_id`,
+    `template_name`,
+    `template_type`,
+    `prompt_content`,
+    `description`,
+    `is_active`,
+    `sort_order`,
+    `created_by`,
+    `create_time`
+) VALUES (
+    'tender-qa-comparison-005',
+    '供应商对比问答',
+    'qa',
+    '你是一个专业的供应商对比分析专家。请基于多个供应商的投标文件，进行对比分析并回答相关问题。
+
+供应商投标文件内容：
+{document_content}
+
+对比分析问题：{query}
+
+供应商对比要点：
+1. 对比各供应商的技术方案差异
+2. 分析报价合理性和竞争力
+3. 评估供应商实力和服务能力
+4. 识别各供应商的优势和劣势
+5. 综合评价供应商的综合竞争力
+6. 提出供应商选择建议
+
+供应商对比分析：
+- 技术方案对比：
+- 商务条件对比：
+- 供应商实力评估：
+- 综合评分建议：
+- 推荐意见：',
+    '供应商对比分析专用提示词，用于多个供应商投标文件的对比评审',
+    1,
+    4,
+    'system',
+    NOW()
+);
+
+-- 初始化默认的文档摘要提示词模板
+INSERT INTO `t_prompt_template` (
+    `template_id`,
+    `template_name`,
+    `template_type`,
+    `prompt_content`,
+    `description`,
+    `is_active`,
+    `sort_order`,
+    `created_by`,
+    `create_time`,
+    `update_time`
+) VALUES (
+    'default-summary-template-001',
+    '默认文档摘要提示词',
+    'summary',
+    '你是一个专业的文档摘要助手。请对以下文档章节进行结构化摘要。
+
+章节标题：{title}
+
+请对该章节内容进行准确、客观的总结。要求：
+1. 总结要准确反映章节的核心要点
+2. 保持客观中立，不添加主观评价
+3. 突出关键信息和技术参数
+4. 结构清晰，逻辑连贯
+5. 控制在适当长度
+
+总结内容：',
+    '默认的文档摘要提示词模板，用于指导AI如何对文档章节进行结构化摘要',
+    1,
+    10,
+    'system',
+    NOW(),
+    NOW()
+);
diff --git a/src/main/resources/db/migration/V1.2__create_qa_history_tables.sql b/src/main/resources/db/migration/V1.2__create_qa_history_tables.sql
new file mode 100644
index 0000000..b49db31
--- /dev/null
+++ b/src/main/resources/db/migration/V1.2__create_qa_history_tables.sql
@@ -0,0 +1,56 @@
+-- 创建QA会话表
+CREATE TABLE IF NOT EXISTS `t_qa_session` (
+    `id` bigint NOT NULL AUTO_INCREMENT COMMENT '主键ID',
+    `session_id` varchar(100) NOT NULL COMMENT '会话ID',
+    `project_id` varchar(50) NOT NULL COMMENT '项目ID',
+    `user_id` varchar(50) NOT NULL COMMENT '用户ID',
+    `first_query` text COMMENT '首次问题内容',
+    `start_time` datetime NOT NULL COMMENT '会话开始时间',
+    `last_message_time` datetime DEFAULT NULL COMMENT '最后消息时间',
+    `message_count` int DEFAULT '0' COMMENT '消息总数',
+    `status` varchar(20) DEFAULT 'active' COMMENT '会话状态：active-活跃，closed-关闭',
+    `create_time` datetime NOT NULL DEFAULT CURRENT_TIMESTAMP COMMENT '创建时间',
+    `update_time` datetime NOT NULL DEFAULT CURRENT_TIMESTAMP ON UPDATE CURRENT_TIMESTAMP COMMENT '更新时间',
+    PRIMARY KEY (`id`),
+    UNIQUE KEY `uk_session_id` (`session_id`),
+    KEY `idx_project_user` (`project_id`, `user_id`),
+    KEY `idx_start_time` (`start_time`)
+) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_unicode_ci COMMENT='QA会话表';
+
+-- 创建QA消息表
+CREATE TABLE IF NOT EXISTS `t_qa_message` (
+    `id` bigint NOT NULL AUTO_INCREMENT COMMENT '主键ID',
+    `message_id` varchar(100) NOT NULL COMMENT '消息ID',
+    `session_id` varchar(100) NOT NULL COMMENT '会话ID',
+    `sender` varchar(20) NOT NULL COMMENT '发送者：user-用户，ai-AI',
+    `content` text NOT NULL COMMENT '消息内容',
+    `sources` text COMMENT '来源信息JSON',
+    `file_ids` text COMMENT '相关文件ID列表JSON',
+    `timestamp` datetime NOT NULL COMMENT '消息时间戳',
+    `create_time` datetime NOT NULL DEFAULT CURRENT_TIMESTAMP COMMENT '创建时间',
+    PRIMARY KEY (`id`),
+    UNIQUE KEY `uk_message_id` (`message_id`),
+    KEY `idx_session_timestamp` (`session_id`, `timestamp`),
+    KEY `idx_sender` (`sender`)
+) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_unicode_ci COMMENT='QA消息表';
+
+-- 创建文件目录表
+CREATE TABLE IF NOT EXISTS `t_file_directory` (
+    `id` bigint NOT NULL AUTO_INCREMENT COMMENT '主键ID',
+    `file_id` varchar(100) NOT NULL COMMENT '文件ID',
+    `chapter_id` varchar(100) NOT NULL COMMENT '章节ID',
+    `title` text NOT NULL COMMENT '章节标题',
+    `page` varchar(50) COMMENT '页码信息',
+    `start_page` int COMMENT '起始页码',
+    `end_page` int COMMENT '结束页码',
+    `parent_chapter_id` varchar(100) COMMENT '父章节ID',
+    `level` int DEFAULT '1' COMMENT '层级',
+    `sort_order` int DEFAULT '0' COMMENT '排序',
+    `summary` text COMMENT '章节摘要',
+    `create_time` datetime NOT NULL DEFAULT CURRENT_TIMESTAMP COMMENT '创建时间',
+    `update_time` datetime NOT NULL DEFAULT CURRENT_TIMESTAMP ON UPDATE CURRENT_TIMESTAMP COMMENT '更新时间',
+    PRIMARY KEY (`id`),
+    UNIQUE KEY `uk_file_chapter` (`file_id`, `chapter_id`),
+    KEY `idx_file_id` (`file_id`),
+    KEY `idx_parent_chapter` (`parent_chapter_id`)
+) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_unicode_ci COMMENT='文件目录表';
diff --git a/src/main/resources/db/migration/V1.4__create_comparison_task_table.sql b/src/main/resources/db/migration/V1.4__create_comparison_task_table.sql
new file mode 100644
index 0000000..7fc95b7
--- /dev/null
+++ b/src/main/resources/db/migration/V1.4__create_comparison_task_table.sql
@@ -0,0 +1,22 @@
+-- 创建对比分析任务表
+CREATE TABLE `t_comparison_task` (
+    `id` bigint(20) NOT NULL AUTO_INCREMENT COMMENT '主键ID',
+    `task_id` varchar(64) NOT NULL COMMENT '任务ID（业务ID，唯一）',
+    `project_id` varchar(64) NOT NULL COMMENT '项目ID',
+    `user_id` varchar(64) NOT NULL COMMENT '用户ID',
+    `vendor_ids` text COMMENT '供应商ID列表（JSON格式）',
+    `dimensions` text COMMENT '对比维度（JSON格式）',
+    `status` varchar(32) NOT NULL COMMENT '任务状态：pending-待处理, processing-处理中, completed-已完成, failed-失败',
+    `result` longtext COMMENT '对比结果（JSON格式）',
+    `error_message` text COMMENT '错误信息',
+    `progress` int(11) NOT NULL DEFAULT 0 COMMENT '处理进度（0-100）',
+    `create_time` datetime NOT NULL COMMENT '创建时间',
+    `update_time` datetime DEFAULT NULL COMMENT '更新时间',
+    `completed_time` datetime DEFAULT NULL COMMENT '完成时间',
+    PRIMARY KEY (`id`),
+    UNIQUE KEY `uk_task_id` (`task_id`),
+    KEY `idx_project_id` (`project_id`),
+    KEY `idx_user_id` (`user_id`),
+    KEY `idx_status` (`status`),
+    KEY `idx_create_time` (`create_time`)
+) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_unicode_ci COMMENT='对比分析任务表';
diff --git a/src/main/resources/db/migration/V1.5__create_user_log_table.sql b/src/main/resources/db/migration/V1.5__create_user_log_table.sql
new file mode 100644
index 0000000..6281859
--- /dev/null
+++ b/src/main/resources/db/migration/V1.5__create_user_log_table.sql
@@ -0,0 +1,25 @@
+-- 创建用户操作日志表
+CREATE TABLE `t_user_log` (
+    `id` bigint(20) NOT NULL AUTO_INCREMENT COMMENT '主键ID',
+    `user_id` varchar(64) NOT NULL COMMENT '用户ID',
+    `user_name` varchar(128) DEFAULT NULL COMMENT '用户名',
+    `project_id` varchar(64) DEFAULT NULL COMMENT '项目ID',
+    `operation_type` varchar(64) NOT NULL COMMENT '操作类型（UPLOAD_FILE-上传文件, QA_QUESTION-智能问答, QA_MESSAGE-问答消息, VIEW_CATALOG-查看目录, CATALOG_SUMMARY-目录总结, MODEL_REPLY-大模型回复等）',
+    `operation_detail` text COMMENT '操作详情',
+    `related_id` varchar(64) DEFAULT NULL COMMENT '关联ID（如文件ID、会话ID等）',
+    `ip_address` varchar(64) DEFAULT NULL COMMENT '操作IP地址',
+    `user_agent` varchar(512) DEFAULT NULL COMMENT '用户代理',
+    `create_time` datetime NOT NULL DEFAULT CURRENT_TIMESTAMP COMMENT '创建时间',
+    PRIMARY KEY (`id`),
+    KEY `idx_user_id` (`user_id`),
+    KEY `idx_project_id` (`project_id`),
+    KEY `idx_operation_type` (`operation_type`),
+    KEY `idx_create_time` (`create_time`)
+) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_unicode_ci COMMENT='用户操作日志表';
+
+-- 初始化一些常见的操作类型注释
+INSERT INTO `t_user_log` (`user_id`, `user_name`, `operation_type`, `operation_detail`, `create_time`) 
+VALUES ('system', '系统账户', 'SYSTEM_INIT', '创建用户操作日志表', NOW());
+
+-- 删除初始化数据
+DELETE FROM `t_user_log` WHERE `user_id` = 'system';
\ No newline at end of file
diff --git a/src/main/resources/db/migration/V1__create_ocr_result_table.sql b/src/main/resources/db/migration/V1__create_ocr_result_table.sql
new file mode 100644
index 0000000..4c9b7fb
--- /dev/null
+++ b/src/main/resources/db/migration/V1__create_ocr_result_table.sql
@@ -0,0 +1,62 @@
+-- 创建OCR结果表 (PostgreSQL语法)
+CREATE TABLE IF NOT EXISTS t_ocr_result (
+    id BIGSERIAL PRIMARY KEY,
+    file_id BIGINT NOT NULL,
+    project_id VARCHAR(64) NOT NULL,
+    supplier_id VARCHAR(64),
+    ocr_task_id VARCHAR(100),
+    status VARCHAR(32) NOT NULL DEFAULT 'processing',
+    recognized_text TEXT,
+    download_url TEXT,
+    download_file_name VARCHAR(512),
+    processed_pages INTEGER,
+    total_pages INTEGER,
+    error_message TEXT,
+    start_time TIMESTAMP,
+    end_time TIMESTAMP,
+    create_time TIMESTAMP NOT NULL DEFAULT CURRENT_TIMESTAMP,
+    update_time TIMESTAMP NOT NULL DEFAULT CURRENT_TIMESTAMP,
+
+    -- 外键约束
+    CONSTRAINT fk_ocr_file FOREIGN KEY (file_id) REFERENCES t_file(id) ON DELETE CASCADE
+);
+
+-- 创建索引
+CREATE INDEX IF NOT EXISTS idx_ocr_file_id ON t_ocr_result(file_id);
+CREATE INDEX IF NOT EXISTS idx_ocr_project_id ON t_ocr_result(project_id);
+CREATE INDEX IF NOT EXISTS idx_ocr_supplier_id ON t_ocr_result(supplier_id);
+CREATE INDEX IF NOT EXISTS idx_ocr_status ON t_ocr_result(status);
+CREATE INDEX IF NOT EXISTS idx_ocr_create_time ON t_ocr_result(create_time);
+
+-- 创建更新时间触发器函数
+CREATE OR REPLACE FUNCTION update_updated_at_column()
+RETURNS TRIGGER AS $$
+BEGIN
+    NEW.update_time = CURRENT_TIMESTAMP;
+    RETURN NEW;
+END;
+$$ language 'plpgsql';
+
+-- 创建触发器
+CREATE TRIGGER update_ocr_result_updated_at
+    BEFORE UPDATE ON t_ocr_result
+    FOR EACH ROW
+    EXECUTE FUNCTION update_updated_at_column();
+
+-- 添加注释 (PostgreSQL方式)
+COMMENT ON TABLE t_ocr_result IS 'OCR识别结果表';
+COMMENT ON COLUMN t_ocr_result.file_id IS '文件ID，关联t_file表';
+COMMENT ON COLUMN t_ocr_result.project_id IS '项目ID';
+COMMENT ON COLUMN t_ocr_result.supplier_id IS '供应商ID';
+COMMENT ON COLUMN t_ocr_result.ocr_task_id IS 'OCR任务ID';
+COMMENT ON COLUMN t_ocr_result.status IS '识别状态（processing-处理中, success-成功, failed-失败）';
+COMMENT ON COLUMN t_ocr_result.recognized_text IS '识别文本内容';
+COMMENT ON COLUMN t_ocr_result.download_url IS '下载链接（OCR结果文件下载链接）';
+COMMENT ON COLUMN t_ocr_result.download_file_name IS '下载文件名';
+COMMENT ON COLUMN t_ocr_result.processed_pages IS '处理页数';
+COMMENT ON COLUMN t_ocr_result.total_pages IS '总页数';
+COMMENT ON COLUMN t_ocr_result.error_message IS '错误信息';
+COMMENT ON COLUMN t_ocr_result.start_time IS '开始时间';
+COMMENT ON COLUMN t_ocr_result.end_time IS '完成时间';
+COMMENT ON COLUMN t_ocr_result.create_time IS '创建时间';
+COMMENT ON COLUMN t_ocr_result.update_time IS '更新时间';
diff --git a/src/main/resources/logback-spring.xml b/src/main/resources/logback-spring.xml
new file mode 100644
index 0000000..8e4ea9d
--- /dev/null
+++ b/src/main/resources/logback-spring.xml
@@ -0,0 +1,38 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<configuration>
+    <!-- 引入Spring Boot默认配置 -->
+    <include resource="org/springframework/boot/logging/logback/defaults.xml"/>
+
+    <!-- 控制台输出（UTF-8编码，支持中文） -->
+    <appender name="CONSOLE" class="ch.qos.logback.core.ConsoleAppender">
+        <encoder>
+            <charset>UTF-8</charset>
+            <pattern>%d{HH:mm:ss} [%thread] %-5level %logger{30} - %msg%n</pattern>
+        </encoder>
+    </appender>
+
+    <!-- 数据库日志Appender（通过Spring初始化） -->
+    <appender name="DATABASE" class="com.zhpb.gdyd_zhpb_zgf.config.DatabaseLogAppender">
+    </appender>
+
+    <!-- 根日志级别 -->
+    <root level="INFO">
+        <appender-ref ref="CONSOLE"/>
+        <appender-ref ref="DATABASE"/>
+    </root>
+
+    <!-- 应用日志级别 -->
+    <logger name="com.zhpb.gdyd_zhpb_zgf" level="DEBUG"/>
+    
+    <!-- 抑制Hibernate和JPA的日志输出 -->
+    <logger name="org.hibernate" level="WARN"/>
+    <logger name="org.hibernate.SQL" level="OFF"/>
+    <logger name="org.hibernate.type" level="OFF"/>
+    <logger name="org.hibernate.type.descriptor.sql" level="OFF"/>
+    
+    <!-- Spring框架日志 -->
+    <logger name="org.springframework" level="INFO"/>
+    <logger name="org.springframework.data" level="WARN"/>
+</configuration>
+
+
diff --git a/src/main/resources/static/view/files.html b/src/main/resources/static/view/files.html
new file mode 100644
index 0000000..3dfa524
--- /dev/null
+++ b/src/main/resources/static/view/files.html
@@ -0,0 +1,240 @@
+<!DOCTYPE html>
+<html lang="zh-CN">
+<head>
+    <meta charset="UTF-8">
+    <title>文件列表</title>
+    <style>
+        body {
+            font-family: -apple-system, BlinkMacSystemFont, 'Segoe UI', Roboto, 'Helvetica Neue', Arial, sans-serif;
+            margin: 0;
+            padding: 20px;
+            background-color: #f5f5f5;
+            color: #333;
+        }
+        .container {
+            max-width: 1200px;
+            margin: 0 auto;
+            background: white;
+            padding: 30px;
+            border-radius: 8px;
+            box-shadow: 0 2px 10px rgba(0,0,0,0.1);
+        }
+        h2 {
+            color: #2c3e50;
+            margin-bottom: 30px;
+            text-align: center;
+        }
+        .filter {
+            background: #f8f9fa;
+            padding: 20px;
+            border-radius: 6px;
+            margin-bottom: 30px;
+            display: flex;
+            gap: 15px;
+            align-items: flex-end;
+            flex-wrap: wrap;
+        }
+        label {
+            display: flex;
+            flex-direction: column;
+            font-weight: 500;
+            color: #555;
+        }
+        input {
+            padding: 10px;
+            border: 1px solid #ddd;
+            border-radius: 4px;
+            font-size: 14px;
+            margin-top: 5px;
+        }
+        input:focus {
+            outline: none;
+            border-color: #007bff;
+            box-shadow: 0 0 0 2px rgba(0,123,255,0.25);
+        }
+        button {
+            padding: 10px 20px;
+            background-color: #007bff;
+            color: white;
+            border: none;
+            border-radius: 4px;
+            cursor: pointer;
+            font-size: 14px;
+            transition: background-color 0.2s;
+            align-self: flex-end;
+        }
+        button:hover {
+            background-color: #0056b3;
+        }
+
+        .clear-btn {
+            background-color: #6c757d;
+            margin-left: 10px;
+        }
+
+        .clear-btn:hover {
+            background-color: #5a6268;
+        }
+        table {
+            border-collapse: collapse;
+            width: 100%;
+            margin-top: 20px;
+            box-shadow: 0 1px 3px rgba(0,0,0,0.1);
+        }
+        th, td {
+            padding: 12px;
+            text-align: left;
+            border-bottom: 1px solid #ddd;
+        }
+        th {
+            background-color: #f8f9fa;
+            font-weight: 600;
+            color: #2c3e50;
+        }
+        tr:hover {
+            background-color: #f8f9fa;
+        }
+        .download-link {
+            color: #007bff;
+            text-decoration: none;
+        }
+        .download-link:hover {
+            text-decoration: underline;
+        }
+        .no-data {
+            text-align: center;
+            padding: 40px;
+            color: #666;
+            font-style: italic;
+        }
+        .file-size {
+            color: #28a745;
+            font-weight: 500;
+        }
+        .status-uploaded {
+            color: #17a2b8;
+        }
+        .status-processed {
+            color: #28a745;
+        }
+        .status-failed {
+            color: #dc3545;
+        }
+    </style>
+</head>
+<body>
+    <div class="container">
+        <h2>文件列表</h2>
+        <div class="filter">
+            <label>项目ID：<input id="projectId" placeholder="可选"></label>
+            <label>供应商ID：<input id="supplierId" placeholder="可选"></label>
+            <button onclick="loadFiles()">查询</button>
+            <button onclick="clearFileFilters()" class="clear-btn">清空</button>
+        </div>
+        <table>
+    <thead>
+    <tr>
+        <th>ID</th>
+        <th>文件名</th>
+        <th>文件大小</th>
+        <th>项目ID</th>
+        <th>供应商ID</th>
+        <th>状态</th>
+        <th>OCR状态</th>
+        <th>操作</th>
+    </tr>
+    </thead>
+                    <tbody id="fileTable">
+                        <tr>
+                            <td colspan="8" class="no-data">点击查询按钮加载文件列表</td>
+                        </tr>
+                    </tbody>
+        </table>
+        <p><a href="/view/index.html">返回首页</a></p>
+    </div>
+
+<script>
+async function loadFiles() {
+    const projectId = document.getElementById('projectId').value.trim();
+    const supplierId = document.getElementById('supplierId').value.trim();
+    const params = new URLSearchParams();
+    if (projectId) params.append('projectId', projectId);
+    if (supplierId) params.append('supplierId', supplierId);
+    const res = await fetch('/view/api/v1/files/list' + (params.toString() ? '?' + params.toString() : ''));
+    const data = await res.json();
+    const tbody = document.getElementById('fileTable');
+    tbody.innerHTML = '';
+    if (!data || data.length === 0) {
+        tbody.innerHTML = '<tr><td colspan="8" class="no-data">无数据</td></tr>';
+        return;
+    }
+    // 为每个文件获取OCR状态
+    const filesWithOcr = await Promise.all(data.map(async (f) => {
+        try {
+            // 只有PDF文件才需要OCR状态
+            if (f.fileExtension && f.fileExtension.toLowerCase() === 'pdf') {
+                const ocrRes = await fetch(`/view/api/v1/files/ocr/file/${f.id}`);
+                const ocrData = await ocrRes.json();
+                f.ocrResult = ocrData.code === 200 ? ocrData.data : null;
+            }
+        } catch (error) {
+            console.warn(`获取文件 ${f.id} 的OCR状态失败:`, error);
+            f.ocrResult = null;
+        }
+        return f;
+    }));
+
+    filesWithOcr.forEach(f => {
+        const tr = document.createElement('tr');
+
+        const fileStatusClass = f.status === 'uploaded' ? 'status-uploaded' :
+                               f.status === 'processed' ? 'status-processed' :
+                               f.status === 'failed' ? 'status-failed' : '';
+
+        // OCR状态显示
+        let ocrStatusDisplay = '-';
+        if (f.fileExtension && f.fileExtension.toLowerCase() === 'pdf') {
+            if (f.ocrResult) {
+                const ocrStatus = f.ocrResult.status;
+                const ocrClass = ocrStatus === 'success' ? 'status-processed' :
+                                ocrStatus === 'failed' ? 'status-failed' :
+                                ocrStatus === 'processing' ? 'status-uploaded' : '';
+                ocrStatusDisplay = `<span class="${ocrClass}">${ocrStatus}</span>`;
+            } else {
+                ocrStatusDisplay = '<span class="status-uploaded">未开始</span>';
+            }
+        }
+
+        tr.innerHTML = `
+          <td>${f.id}</td>
+          <td>${f.fileName || ''}</td>
+          <td class="file-size">${f.fileSize ? formatFileSize(f.fileSize) : ''}</td>
+          <td>${f.projectId || ''}</td>
+          <td>${f.supplierId || '-'}</td>
+          <td><span class="${fileStatusClass}">${f.status || 'unknown'}</span></td>
+          <td>${ocrStatusDisplay}</td>
+          <td><a href="/view${f.downloadUrl}" class="download-link" target="_blank">下载</a></td>
+        `;
+        tbody.appendChild(tr);
+    });
+}
+
+function formatFileSize(bytes) {
+    if (!bytes) return '';
+    const sizes = ['Bytes', 'KB', 'MB', 'GB'];
+    const i = Math.floor(Math.log(bytes) / Math.log(1024));
+    return (bytes / Math.pow(1024, i)).toFixed(2) + ' ' + sizes[i];
+}
+
+function clearFileFilters() {
+    document.getElementById('projectId').value = '';
+    document.getElementById('supplierId').value = '';
+    const tbody = document.getElementById('fileTable');
+    tbody.innerHTML = '<tr><td colspan="8" class="no-data">点击查询按钮加载文件列表</td></tr>';
+}
+
+loadFiles();
+</script>
+</body>
+</html>
+
diff --git a/src/main/resources/static/view/health.html b/src/main/resources/static/view/health.html
new file mode 100644
index 0000000..0f2f825
--- /dev/null
+++ b/src/main/resources/static/view/health.html
@@ -0,0 +1,1248 @@
+<!DOCTYPE html>
+<html lang="zh-CN">
+<head>
+    <meta charset="UTF-8">
+    <meta name="viewport" content="width=device-width, initial-scale=1.0">
+    <title>系统状态监控 - 智慧评标-主观分助手</title>
+    <style>
+        * {
+            margin: 0;
+            padding: 0;
+            box-sizing: border-box;
+        }
+
+        body {
+            font-family: -apple-system, BlinkMacSystemFont, 'Segoe UI', 'Roboto', 'Helvetica Neue', Arial, sans-serif;
+            background: #0f172a;
+            color: #e2e8f0;
+            height: 100vh;
+            overflow: hidden;
+            padding: 16px;
+        }
+
+        .container {
+            max-width: 100%;
+            height: 100%;
+            display: flex;
+            flex-direction: column;
+            margin: 0 auto;
+        }
+
+        .header {
+            text-align: center;
+            margin-bottom: 16px;
+            padding: 12px 0;
+            flex-shrink: 0;
+        }
+
+        h1 {
+            font-size: 24px;
+            font-weight: 700;
+            color: #f1f5f9;
+            margin-bottom: 4px;
+        }
+
+        .subtitle {
+            color: #94a3b8;
+            font-size: 14px;
+        }
+
+        .header-actions {
+            margin-top: 12px;
+            display: flex;
+            justify-content: center;
+            gap: 12px;
+            flex-wrap: wrap;
+        }
+
+        .nav-btn {
+            display: inline-flex;
+            align-items: center;
+            gap: 6px;
+            padding: 10px 18px;
+            border-radius: 999px;
+            border: 1px solid #475569;
+            background: rgba(148, 163, 184, 0.1);
+            color: #e2e8f0;
+            font-size: 13px;
+            font-weight: 500;
+            text-decoration: none;
+            transition: all 0.2s ease;
+        }
+
+        .nav-btn:hover {
+            background: rgba(148, 163, 184, 0.2);
+            transform: translateY(-1px);
+        }
+
+        .status-summary {
+            display: grid;
+            grid-template-columns: repeat(5, 1fr);
+            gap: 12px;
+            margin-bottom: 16px;
+            flex-shrink: 0;
+        }
+
+        .summary-card {
+            background: #1e293b;
+            border-radius: 8px;
+            padding: 16px;
+            border: 1px solid #334155;
+        }
+
+        .summary-title {
+            color: #94a3b8;
+            font-size: 11px;
+            margin-bottom: 8px;
+            text-transform: uppercase;
+            letter-spacing: 0.5px;
+        }
+
+        .summary-value {
+            font-size: 24px;
+            font-weight: 700;
+            margin-bottom: 6px;
+        }
+
+        .summary-status {
+            display: flex;
+            align-items: center;
+            gap: 8px;
+            font-size: 14px;
+        }
+
+        .status-dot {
+            width: 8px;
+            height: 8px;
+            border-radius: 50%;
+            animation: pulse 2s infinite;
+        }
+
+        .status-dot.online {
+            background: #10b981;
+        }
+
+        .status-dot.offline {
+            background: #ef4444;
+        }
+
+        @keyframes pulse {
+            0%, 100% {
+                opacity: 1;
+            }
+            50% {
+                opacity: 0.5;
+            }
+        }
+
+        .content-grid {
+            display: grid;
+            grid-template-columns: 1fr 1fr;
+            gap: 16px;
+            flex: 1;
+            min-height: 0;
+            overflow: hidden;
+        }
+
+        .section {
+            background: #1e293b;
+            border-radius: 8px;
+            padding: 16px;
+            border: 1px solid #334155;
+            display: flex;
+            flex-direction: column;
+            overflow: hidden;
+        }
+
+        .section-title {
+            font-size: 16px;
+            font-weight: 600;
+            color: #f1f5f9;
+            margin-bottom: 12px;
+            display: flex;
+            align-items: center;
+            gap: 8px;
+            flex-shrink: 0;
+        }
+
+        .section-icon {
+            font-size: 20px;
+        }
+
+        .section-content {
+            flex: 1;
+            overflow-y: auto;
+            overflow-x: hidden;
+            padding-right: 8px;
+        }
+
+        .section-content::-webkit-scrollbar {
+            width: 6px;
+        }
+
+        .section-content::-webkit-scrollbar-track {
+            background: #0f172a;
+            border-radius: 3px;
+        }
+
+        .section-content::-webkit-scrollbar-thumb {
+            background: #334155;
+            border-radius: 3px;
+        }
+
+        .section-content::-webkit-scrollbar-thumb:hover {
+            background: #475569;
+        }
+
+        .service-item {
+            display: flex;
+            align-items: center;
+            justify-content: space-between;
+            padding: 16px;
+            background: #0f172a;
+            border-radius: 8px;
+            margin-bottom: 12px;
+            border: 1px solid #1e293b;
+            transition: all 0.2s ease;
+        }
+
+        .service-item:hover {
+            border-color: #334155;
+            background: #1e293b;
+        }
+
+        .service-info {
+            display: flex;
+            align-items: center;
+            gap: 12px;
+            flex: 1;
+        }
+
+        .service-icon {
+            font-size: 18px;
+            width: 36px;
+            height: 36px;
+            display: flex;
+            align-items: center;
+            justify-content: center;
+            background: #1e293b;
+            border-radius: 6px;
+            flex-shrink: 0;
+        }
+
+        .service-details {
+            flex: 1;
+        }
+
+        .service-name {
+            font-weight: 600;
+            color: #f1f5f9;
+            margin-bottom: 2px;
+            font-size: 14px;
+        }
+
+        .service-desc {
+            color: #94a3b8;
+            font-size: 12px;
+        }
+
+        .service-status {
+            display: flex;
+            align-items: center;
+            gap: 8px;
+        }
+
+        .status-badge {
+            padding: 6px 12px;
+            border-radius: 6px;
+            font-size: 12px;
+            font-weight: 600;
+            text-transform: uppercase;
+            letter-spacing: 0.5px;
+        }
+
+        .status-operational {
+            background: rgba(16, 185, 129, 0.2);
+            color: #10b981;
+            border: 1px solid rgba(16, 185, 129, 0.3);
+        }
+
+        .status-degraded {
+            background: rgba(251, 191, 36, 0.2);
+            color: #fbbf24;
+            border: 1px solid rgba(251, 191, 36, 0.3);
+        }
+
+        .status-down {
+            background: rgba(239, 68, 68, 0.2);
+            color: #ef4444;
+            border: 1px solid rgba(239, 68, 68, 0.3);
+        }
+
+        .status-checking {
+            background: rgba(148, 163, 184, 0.2);
+            color: #94a3b8;
+            border: 1px solid rgba(148, 163, 184, 0.3);
+        }
+
+        .response-time {
+            color: #64748b;
+            font-size: 12px;
+            margin-left: 8px;
+        }
+
+        .footer {
+            text-align: center;
+            margin-top: 16px;
+            flex-shrink: 0;
+        }
+
+        .back-btn {
+            display: inline-flex;
+            align-items: center;
+            gap: 8px;
+            padding: 10px 20px;
+            background: #334155;
+            color: #f1f5f9;
+            text-decoration: none;
+            border-radius: 8px;
+            font-size: 14px;
+            font-weight: 500;
+            transition: all 0.3s ease;
+            border: 1px solid #475569;
+        }
+
+        .back-btn:hover {
+            background: #475569;
+            transform: translateY(-1px);
+        }
+
+        .loading {
+            text-align: center;
+            color: #94a3b8;
+            padding: 40px;
+        }
+
+        .detail-info {
+            margin-top: 8px;
+            font-size: 12px;
+            color: #64748b;
+        }
+
+        @media (max-width: 1200px) {
+            .status-summary {
+                grid-template-columns: repeat(3, 1fr);
+            }
+        }
+
+        @media (max-width: 768px) {
+            body {
+                overflow-y: auto;
+            }
+
+            .status-summary {
+                grid-template-columns: repeat(2, 1fr);
+            }
+
+            .content-grid {
+                grid-template-columns: 1fr;
+            }
+
+            .section-content {
+                max-height: 300px;
+            }
+
+            .service-item {
+                flex-direction: column;
+                align-items: flex-start;
+                gap: 8px;
+            }
+
+            .service-status {
+                width: 100%;
+                justify-content: space-between;
+            }
+        }
+    </style>
+</head>
+<body>
+    <div class="container">
+        <div class="header">
+            <h1>系统状态监控</h1>
+            <div class="subtitle">智慧评标-主观分助手 实时服务状态</div>
+            <div class="header-actions">
+                <a class="nav-btn" href="/logs.html" target="_blank">📜 查看系统日志</a>
+            </div>
+        </div>
+
+        <!-- 状态概览 -->
+        <div class="status-summary" id="statusSummary">
+            <div class="summary-card">
+                <div class="summary-title">所有系统</div>
+                <div class="summary-value" id="allSystems">-</div>
+                <div class="summary-status">
+                    <span class="status-dot online" id="allSystemsDot"></span>
+                    <span id="allSystemsText">检查中...</span>
+                </div>
+            </div>
+            <div class="summary-card">
+                <div class="summary-title">API接口</div>
+                <div class="summary-value" id="apiStatus">-</div>
+                <div class="summary-status">
+                    <span class="status-dot online" id="apiStatusDot"></span>
+                    <span id="apiStatusText">检查中...</span>
+                </div>
+            </div>
+            <div class="summary-card">
+                <div class="summary-title">数据库</div>
+                <div class="summary-value" id="dbStatus">-</div>
+                <div class="summary-status">
+                    <span class="status-dot online" id="dbStatusDot"></span>
+                    <span id="dbStatusText">检查中...</span>
+                </div>
+            </div>
+            <div class="summary-card">
+                <div class="summary-title">缓存服务</div>
+                <div class="summary-value" id="cacheStatus">-</div>
+                <div class="summary-status">
+                    <span class="status-dot online" id="cacheStatusDot"></span>
+                    <span id="cacheStatusText">检查中...</span>
+                </div>
+            </div>
+            <div class="summary-card">
+                <div class="summary-title">接口访问次数</div>
+                <div class="summary-value" id="apiCount">-</div>
+                <div class="summary-status">
+                    <span class="status-dot online"></span>
+                    <span>实时统计</span>
+                </div>
+            </div>
+        </div>
+
+        <!-- 内容区域 -->
+        <div class="content-grid">
+            <!-- 系统组件 -->
+            <div class="section">
+                <div class="section-title">
+                    <span class="section-icon">🔧</span>
+                    <span>系统组件</span>
+                </div>
+                <div class="section-content" id="systemComponents">
+                    <div class="loading">正在检查系统组件...</div>
+                </div>
+            </div>
+
+            <!-- API接口 -->
+            <div class="section">
+                <div class="section-title">
+                    <span class="section-icon">🌐</span>
+                    <span>API接口</span>
+                </div>
+                <div class="section-content" id="apiEndpoints">
+                    <div class="loading">正在检查API接口...</div>
+                </div>
+            </div>
+        </div>
+
+        <!-- 底部按钮 -->
+        <div class="footer">
+            <a href="/" class="back-btn">
+                <span>←</span>
+                <span>返回首页</span>
+            </a>
+        </div>
+    </div>
+
+    <script>
+        // API接口列表
+        const apiEndpoints = [
+            {
+                name: 'SSO登录接口',
+                path: '/api/v1/sso/login',
+                method: 'GET',
+                description: '单点登录接口，验证ticket并生成JWT Token',
+                icon: '🔐'
+            },
+            {
+                name: '首页',
+                path: '/',
+                method: 'GET',
+                description: '系统首页',
+                icon: '🏠'
+            },
+            {
+                name: '健康检查',
+                path: '/health',
+                method: 'GET',
+                description: '系统健康检查页面',
+                icon: '❤️'
+            },
+            {
+                name: 'Swagger文档',
+                path: '/swagger-ui.html',
+                method: 'GET',
+                description: 'API交互式文档',
+                icon: '📄'
+            },
+            {
+                name: 'API文档JSON',
+                path: '/api-docs',
+                method: 'GET',
+                description: 'OpenAPI规范文档',
+                icon: '📋'
+            },
+            {
+                name: '文件上传接口',
+                path: '/api/v1/files/uploadzip',
+                method: 'POST',
+                description: '上传ZIP压缩包并自动解压',
+                icon: '📦'
+            },
+            {
+                name: 'PDF文件预览',
+                path: '/api/v1/files/preview/pdf/test-file-id',
+                method: 'GET',
+                description: '在线预览PDF文件',
+                icon: '📄'
+            },
+            {
+                name: 'Word文档预览',
+                path: '/api/v1/files/preview/word/test-file-id',
+                method: 'GET',
+                description: '预览Word文档（转换为PDF后预览）',
+                icon: '📝'
+            },
+            {
+                name: '向量化状态统计',
+                path: '/api/v1/files/vectorization/stats',
+                method: 'GET',
+                description: '查询系统中所有文件的向量化状态统计',
+                icon: '📊'
+            },
+            {
+                name: '向量化健康检查测试',
+                path: '/api/v1/files/test/vectorization-health',
+                method: 'GET',
+                description: '测试向量化服务的健康检查功能',
+                icon: '❤️'
+            },
+            {
+                name: '项目供应商和文件信息',
+                path: '/api/v1/files/project/test-project-id/suppliers-files',
+                method: 'GET',
+                description: '查询项目相关的供应商和文件信息',
+                icon: '📋'
+            },
+            {
+                name: '智能问答',
+                path: '/api/v1/qa/sendMessage',
+                method: 'POST',
+                description: '基于文档的智能问答接口',
+                icon: '🤖'
+            },
+            {
+                name: '智能问答测试页面',
+                path: '/qa-test.html',
+                method: 'GET',
+                description: '智能问答功能测试页面',
+                icon: '💬'
+            },
+            {
+                name: '查询项目文件列表',
+                path: '/api/v1/files/project/7654/files',
+                method: 'GET',
+                description: '根据项目ID查询文件列表信息',
+                icon: '📂'
+            },
+            {
+                name: '提示词模板管理',
+                path: '/api/v1/prompt-templates',
+                method: 'GET',
+                description: '提示词模板管理接口',
+                icon: '⚙️'
+            },
+            {
+                name: '提示词配置页面',
+                path: '/prompt-templates.html',
+                method: 'GET',
+                description: '提示词模板配置管理页面',
+                icon: '📝'
+            },
+            {
+                name: '获取提示词模板详情',
+                path: '/api/v1/prompt-templates/{templateId}',
+                method: 'GET',
+                description: '获取指定提示词模板的详细信息',
+                icon: '📋'
+            },
+            {
+                name: '创建提示词模板',
+                path: '/api/v1/prompt-templates',
+                method: 'POST',
+                description: '创建新的提示词模板',
+                icon: '➕'
+            },
+            {
+                name: '更新提示词模板',
+                path: '/api/v1/prompt-templates/{templateId}',
+                method: 'PUT',
+                description: '更新指定提示词模板',
+                icon: '✏️'
+            },
+            {
+                name: '删除提示词模板',
+                path: '/api/v1/prompt-templates/{templateId}',
+                method: 'DELETE',
+                description: '删除指定提示词模板',
+                icon: '🗑️'
+            },
+            {
+                name: '切换提示词模板状态',
+                path: '/api/v1/prompt-templates/{templateId}/status',
+                method: 'PUT',
+                description: '启用或禁用提示词模板',
+                icon: '🔄'
+            },
+            {
+                name: '文件目录结构',
+                path: '/api/v1/summary/directory',
+                method: 'GET',
+                description: '获取指定文件的目录结构',
+                icon: '📑'
+            },
+            {
+                name: '章节摘要',
+                path: '/api/v1/summary/chapter',
+                method: 'GET',
+                description: '获取指定章节的摘要内容',
+                icon: '📖'
+            },
+            {
+                name: '目录处理',
+                path: '/api/v1/summary/directory/process',
+                method: 'POST',
+                description: '手动触发文件目录提取和摘要生成',
+                icon: '⚙️'
+            },
+            {
+                name: '目录状态检查',
+                path: '/api/v1/summary/directory/status',
+                method: 'GET',
+                description: '检查文件的目录提取状态',
+                icon: '🔍'
+            },
+            {
+                name: '目录数量统计',
+                path: '/api/v1/summary/directory/count',
+                method: 'GET',
+                description: '获取系统中目录的总数量',
+                icon: '📊'
+            },
+            {
+                name: '供应商对比-获取供应商',
+                path: '/api/v1/comparison/vendors',
+                method: 'GET',
+                description: '获取指定项目的供应商及其文件信息',
+                icon: '🏢'
+            },
+            {
+                name: '供应商对比-启动任务',
+                path: '/api/v1/comparison/start',
+                method: 'POST',
+                description: '启动供应商横向对比分析任务',
+                icon: '🚀'
+            },
+            {
+                name: '供应商对比-历史任务',
+                path: '/api/v1/comparison/history/tasks',
+                method: 'GET',
+                description: '获取对比分析任务历史记录',
+                icon: '📜'
+            },
+            {
+                name: '供应商对比-任务结果',
+                path: '/api/v1/comparison/result',
+                method: 'GET',
+                description: '获取对比分析任务的处理结果',
+                icon: '📋'
+            },
+            {
+                name: 'QA历史会话',
+                path: '/api/v1/qa/history/sessions',
+                method: 'GET',
+                description: '获取智能问答的历史会话记录',
+                icon: '💬'
+            },
+            {
+                name: 'QA历史消息',
+                path: '/api/v1/qa/history/messages',
+                method: 'GET',
+                description: '获取指定会话的历史消息记录',
+                icon: '📨'
+            },
+            {
+                name: '文件向量化状态',
+                path: '/api/v1/files/vectorization/status/{file_id}',
+                method: 'GET',
+                description: '查询单个文件的向量化状态',
+                icon: '🔍'
+            },
+            {
+                name: '项目文件向量化状态',
+                path: '/api/v1/files/vectorization/project/{project_id}',
+                method: 'GET',
+                description: '查询项目下所有文件的向量化状态',
+                icon: '📂'
+            },
+            {
+                name: '系统统计-总数',
+                path: '/api/v1/statistics/total',
+                method: 'GET',
+                description: '获取系统各类数据的总数统计',
+                icon: '📊'
+            },
+            {
+                name: '系统统计-计数',
+                path: '/api/v1/statistics/count',
+                method: 'GET',
+                description: '获取系统各类数据的计数统计',
+                icon: '🔢'
+            },
+            {
+                name: '目录测试页面',
+                path: '/summary-test.html',
+                method: 'GET',
+                description: '文件目录和摘要功能测试页面',
+                icon: '📑'
+            },
+            {
+                name: '对比测试页面',
+                path: '/comparison-test.html',
+                method: 'GET',
+                description: '供应商对比分析功能测试页面',
+                icon: '⚖️'
+            }
+        ];
+
+        // 检查系统组件健康状态
+        async function checkSystemComponents() {
+            const container = document.getElementById('systemComponents');
+            container.innerHTML = '';
+
+            try {
+                const response = await fetch('/actuator/health', {
+                    method: 'GET',
+                    headers: {
+                        'Accept': 'application/json'
+                    }
+                });
+                
+                if (!response.ok) {
+                    throw new Error(`HTTP ${response.status}: ${response.statusText}`);
+                }
+                
+                const data = await response.json();
+
+                const components = [];
+                
+                // 总体状态
+                components.push({
+                    name: '应用状态',
+                    status: data.status === 'UP' ? 'operational' : 'down',
+                    icon: '🏠',
+                    description: '应用程序运行状态',
+                    details: data.status
+                });
+
+                // 组件状态
+                const normalizeComponentKey = (key) => {
+                    const postgresKeys = ['db', 'pgsqlHealthIndicator', 'postgreSQL', 'postgreSQLHealthIndicator', 'pgsql', 'pgsqlDataSource'];
+                    if (postgresKeys.includes(key)) {
+                        return 'db';
+                    }
+                    return key;
+                };
+
+                if (data.components) {
+                    const renderedComponents = new Set();
+                    const componentMap = {
+                        'db': { name: 'PostgreSQL数据库', icon: '🐘' },
+                        'pgsqlHealthIndicator': { name: 'PostgreSQL数据库', icon: '🐘' },
+                        'postgreSQL': { name: 'PostgreSQL数据库', icon: '🐘' },
+                        'postgreSQLHealthIndicator': { name: 'PostgreSQL数据库', icon: '🐘' },
+                        'pgsql': { name: 'PostgreSQL数据库', icon: '🐘' },
+                        'pgsqlDataSource': { name: 'PostgreSQL数据库', icon: '🐘' },
+                        'diskSpace': { name: '磁盘空间', icon: '💾' },
+                        'redis': { name: 'Redis缓存', icon: '⚡' },
+                        'datasource': { name: '数据源连接池', icon: '🔌' },
+                        'ping': { name: '连接检测', icon: '📡' }
+                    };
+
+                    for (const [key, value] of Object.entries(data.components)) {
+                        if (key !== 'ping') {
+                            const normalizedKey = normalizeComponentKey(key);
+                            if (renderedComponents.has(normalizedKey)) {
+                                continue;
+                            }
+                            renderedComponents.add(normalizedKey);
+
+                            const info = componentMap[normalizedKey] || componentMap[key] || { name: key, icon: '📊' };
+                            components.push({
+                                name: info.name,
+                                status: value.status === 'UP' ? 'operational' : 'down',
+                                icon: info.icon,
+                                description: getComponentDescription(normalizedKey, value),
+                                details: formatComponentDetails(normalizedKey, value)
+                            });
+                        }
+                    }
+                }
+
+                // 更新状态概览
+                updateSummary(components, data);
+
+                // 清空容器并渲染组件列表
+                if (container) {
+                    container.innerHTML = '';
+                    components.forEach(component => {
+                        container.innerHTML += createServiceItem(component);
+                    });
+                }
+
+            } catch (error) {
+                console.error('检查系统组件失败:', error);
+                container.innerHTML = `<div class="service-item">
+                    <div class="service-info">
+                        <span class="service-icon">⚠️</span>
+                        <div class="service-details">
+                            <div class="service-name">无法获取系统状态</div>
+                            <div class="service-desc">错误: ${error.message || '连接健康检查端点失败'}</div>
+                        </div>
+                    </div>
+                    <div class="service-status">
+                        <span class="status-badge status-down">异常</span>
+                    </div>
+                </div>`;
+                
+                // 更新概览状态
+                document.getElementById('allSystems').textContent = '错误';
+                document.getElementById('allSystems').style.color = '#ef4444';
+                document.getElementById('allSystemsText').textContent = '检查失败';
+                document.getElementById('allSystemsDot').className = 'status-dot offline';
+            }
+        }
+
+        // 检查API接口状态
+        async function checkApiEndpoints() {
+            console.log('API接口总数:', apiEndpoints.length);
+            const container = document.getElementById('apiEndpoints');
+            if (!container) return;
+            container.innerHTML = '<div class="loading">正在检查API接口...</div>';
+
+            let operationalCount = 0;
+            let totalCount = apiEndpoints.length;
+            const results = [];
+
+            const checkPromises = apiEndpoints.map(async (endpoint) => {
+                const startTime = performance.now();
+                try {
+                    const controller = new AbortController();
+                    const timeoutId = setTimeout(() => controller.abort(), 3000); // 减少超时时间到3秒
+                    
+                    let response;
+                    // 对于POST接口（如文件上传），使用OPTIONS请求来检查接口是否存在
+                    // 对于其他接口，使用HEAD请求来检查接口是否存在，不触发业务逻辑
+                    const isPostEndpoint = endpoint.method === 'POST';
+                    
+                    if (isPostEndpoint) {
+                        // POST接口使用OPTIONS请求检查
+                        try {
+                            response = await fetch(endpoint.path, { 
+                                method: 'OPTIONS',
+                                signal: controller.signal
+                            });
+                        } catch (optionsError) {
+                            // 如果OPTIONS失败，使用HEAD尝试
+                            try {
+                                response = await fetch(endpoint.path, { 
+                                    method: 'HEAD',
+                                    signal: controller.signal
+                                });
+                            } catch (headError) {
+                                // 如果都失败，最后尝试GET（但只检查404）
+                                response = await fetch(endpoint.path, { 
+                                    method: 'GET',
+                                    signal: controller.signal
+                                });
+                            }
+                        }
+                    } else {
+                        // GET接口使用HEAD请求检查
+                        try {
+                            response = await fetch(endpoint.path, { 
+                                method: 'HEAD',
+                                signal: controller.signal
+                            });
+                        } catch (headError) {
+                            // 如果HEAD请求失败（可能不支持），尝试OPTIONS请求
+                            try {
+                                response = await fetch(endpoint.path, { 
+                                    method: 'OPTIONS',
+                                    signal: controller.signal
+                                });
+                            } catch (optionsError) {
+                                // 如果HEAD和OPTIONS都失败，最后尝试GET请求（但只检查404）
+                                response = await fetch(endpoint.path, { 
+                                    method: 'GET',
+                                    signal: controller.signal
+                                });
+                            }
+                        }
+                    }
+                    
+                    clearTimeout(timeoutId);
+                    const endTime = performance.now();
+                    const responseTime = Math.round(endTime - startTime);
+                    
+                    // 判断接口是否可用
+                    // 只要服务器能响应（不管什么状态码），就认为接口存在且能工作
+                    // 真正的异常是网络错误、超时等无法连接的情况
+                    const isOperational = response.status >= 200 && response.status < 600; // 2xx-5xx状态码都算能响应
+                    
+                    const status = isOperational ? 'operational' : 'degraded';
+                    
+                    results.push({
+                        name: endpoint.name,
+                        status: status,
+                        icon: endpoint.icon,
+                        description: endpoint.description,
+                        details: `${endpoint.method} ${endpoint.path} (${response.status})`,
+                        responseTime: responseTime
+                    });
+                    
+                    if (status === 'operational') {
+                        operationalCount++;
+                    }
+                } catch (error) {
+                    console.error(`检查接口 ${endpoint.name} 失败:`, error);
+                    results.push({
+                        name: endpoint.name,
+                        status: 'down',
+                        icon: endpoint.icon,
+                        description: endpoint.description,
+                        details: `${endpoint.method} ${endpoint.path}`,
+                        error: error.name === 'AbortError' ? '请求超时' : '连接失败'
+                    });
+                }
+            });
+
+            await Promise.all(checkPromises);
+            
+            // 重新计算正常数量（确保准确性）
+            operationalCount = results.filter(r => r.status === 'operational').length;
+            
+            // 清空容器并渲染所有结果
+            container.innerHTML = '';
+            results.forEach(result => {
+                container.innerHTML += createServiceItem(result);
+            });
+
+            // 更新API状态概览
+            const apiStatusDot = document.getElementById('apiStatusDot');
+            const apiStatusText = document.getElementById('apiStatusText');
+            const apiStatus = document.getElementById('apiStatus');
+            
+            if (operationalCount === totalCount) {
+                apiStatusDot.className = 'status-dot online';
+                apiStatusText.textContent = '全部正常';
+                apiStatus.textContent = operationalCount + '/' + totalCount;
+                apiStatus.style.color = '#10b981';
+            } else if (operationalCount > 0) {
+                apiStatusDot.className = 'status-dot degraded';
+                apiStatusText.textContent = '部分异常';
+                apiStatus.textContent = operationalCount + '/' + totalCount;
+                apiStatus.style.color = '#fbbf24';
+            } else {
+                apiStatusDot.className = 'status-dot offline';
+                apiStatusText.textContent = '全部异常';
+                apiStatus.textContent = '0/' + totalCount;
+                apiStatus.style.color = '#ef4444';
+            }
+        }
+
+        function createServiceItem(service) {
+            const statusMap = {
+                'operational': { text: '正常', class: 'status-operational' },
+                'degraded': { text: '降级', class: 'status-degraded' },
+                'down': { text: '异常', class: 'status-down' },
+                'checking': { text: '检查中', class: 'status-checking' }
+            };
+
+            const status = statusMap[service.status] || statusMap['checking'];
+
+            return `
+                <div class="service-item">
+                    <div class="service-info">
+                        <div class="service-icon">${service.icon}</div>
+                        <div class="service-details">
+                            <div class="service-name">${service.name}</div>
+                            <div class="service-desc">${service.description || ''}</div>
+                            ${service.details ? `<div class="detail-info">${service.details}</div>` : ''}
+                            ${service.error ? `<div class="detail-info" style="color:#ef4444;">错误: ${service.error}</div>` : ''}
+                        </div>
+                    </div>
+                    <div class="service-status">
+                        <span class="status-badge ${status.class}">${status.text}</span>
+                        ${service.responseTime ? `<span class="response-time">${service.responseTime}ms</span>` : ''}
+                    </div>
+                </div>
+            `;
+        }
+
+        function getComponentDescription(key, data) {
+            const descriptions = {
+                'db': 'PostgreSQL数据库连接状态',
+                'pgsqlHealthIndicator': 'PostgreSQL数据库连接状态',
+                'postgreSQL': 'PostgreSQL数据库连接状态',
+                'postgreSQLHealthIndicator': 'PostgreSQL数据库连接状态',
+                'pgsql': 'PostgreSQL数据库连接状态',
+                'pgsqlDataSource': 'PostgreSQL数据库连接状态',
+                'diskSpace': '服务器磁盘空间使用情况',
+                'redis': 'Redis缓存服务连接状态',
+                'datasource': '数据源连接池状态'
+            };
+            return descriptions[key] || '系统组件状态';
+        }
+
+        function formatComponentDetails(key, data) {
+            if (!data.details) return '';
+            
+            const detailMap = {
+                'diskSpace': () => {
+                    const free = Math.round(data.details.free / 1024 / 1024 / 1024);
+                    const total = Math.round(data.details.total / 1024 / 1024 / 1024);
+                    return `可用: ${free}GB / 总计: ${total}GB`;
+                },
+                'redis': () => {
+                    return data.details.version ? `版本: ${data.details.version}` : '';
+                },
+                'pgsqlHealthIndicator': () => {
+                    if (data.details.version) {
+                        return `版本: ${data.details.version}`;
+                    }
+                    if (data.details.status) {
+                        return data.details.status;
+                    }
+                    return '';
+                },
+                'postgreSQL': () => {
+                    if (data.details.version) {
+                        return `版本: ${data.details.version}`;
+                    }
+                    if (data.details.status) {
+                        return data.details.status;
+                    }
+                    return '';
+                },
+                'postgreSQLHealthIndicator': () => {
+                    if (data.details.version) {
+                        return `版本: ${data.details.version}`;
+                    }
+                    if (data.details.status) {
+                        return data.details.status;
+                    }
+                    return '';
+                }
+            };
+
+            const formatter = detailMap[key];
+            return formatter ? formatter() : '';
+        }
+
+        function updateSummary(components, data) {
+            const allOperational = components.every(c => c.status === 'operational');
+            const dbComponent = components.find(c => c.name.includes('数据库'));
+            const cacheComponent = components.find(c => c.name.includes('Redis') || c.name.includes('缓存'));
+
+            // 所有系统状态
+            const allSystemsDot = document.getElementById('allSystemsDot');
+            const allSystemsText = document.getElementById('allSystemsText');
+            const allSystems = document.getElementById('allSystems');
+            
+            if (allOperational) {
+                allSystemsDot.className = 'status-dot online';
+                allSystemsText.textContent = '全部正常';
+                allSystems.textContent = '100%';
+                allSystems.style.color = '#10b981';
+            } else {
+                allSystemsDot.className = 'status-dot offline';
+                allSystemsText.textContent = '部分异常';
+                const operationalCount = components.filter(c => c.status === 'operational').length;
+                allSystems.textContent = Math.round(operationalCount / components.length * 100) + '%';
+                allSystems.style.color = '#fbbf24';
+            }
+
+            // 数据库状态
+            if (dbComponent) {
+                const dbStatusDot = document.getElementById('dbStatusDot');
+                const dbStatusText = document.getElementById('dbStatusText');
+                const dbStatus = document.getElementById('dbStatus');
+                
+                if (dbComponent.status === 'operational') {
+                    dbStatusDot.className = 'status-dot online';
+                    dbStatusText.textContent = '正常';
+                    dbStatus.textContent = '正常';
+                    dbStatus.style.color = '#10b981';
+                } else {
+                    dbStatusDot.className = 'status-dot offline';
+                    dbStatusText.textContent = '异常';
+                    dbStatus.textContent = '异常';
+                    dbStatus.style.color = '#ef4444';
+                }
+            }
+
+            // 缓存状态
+            if (cacheComponent) {
+                const cacheStatusDot = document.getElementById('cacheStatusDot');
+                const cacheStatusText = document.getElementById('cacheStatusText');
+                const cacheStatus = document.getElementById('cacheStatus');
+                
+                if (cacheComponent.status === 'operational') {
+                    cacheStatusDot.className = 'status-dot online';
+                    cacheStatusText.textContent = '正常';
+                    cacheStatus.textContent = '正常';
+                    cacheStatus.style.color = '#10b981';
+                } else {
+                    cacheStatusDot.className = 'status-dot offline';
+                    cacheStatusText.textContent = '异常';
+                    cacheStatus.textContent = '异常';
+                    cacheStatus.style.color = '#ef4444';
+                }
+            }
+        }
+
+        // 初始化检查
+        async function init() {
+            try {
+                console.log('开始初始化健康检查...');
+                
+                // 检查必要的DOM元素是否存在
+                const systemComponents = document.getElementById('systemComponents');
+                const apiEndpoints = document.getElementById('apiEndpoints');
+                
+                if (!systemComponents || !apiEndpoints) {
+                    console.error('必要的DOM元素不存在，延迟重试...');
+                    setTimeout(init, 500);
+                    return;
+                }
+                
+                // 并行执行两个检查
+                await Promise.all([
+                    checkSystemComponents().catch(err => {
+                        console.error('系统组件检查失败:', err);
+                        const container = document.getElementById('systemComponents');
+                        if (container) {
+                            container.innerHTML = `<div class="service-item">
+                                <div class="service-info">
+                                    <span class="service-icon">⚠️</span>
+                                    <div class="service-details">
+                                        <div class="service-name">检查失败</div>
+                                        <div class="service-desc">${err.message || '未知错误'}</div>
+                                    </div>
+                                </div>
+                                <div class="service-status">
+                                    <span class="status-badge status-down">异常</span>
+                                </div>
+                            </div>`;
+                        }
+                    }),
+                    checkApiEndpoints().catch(err => {
+                        console.error('API接口检查失败:', err);
+                        const container = document.getElementById('apiEndpoints');
+                        if (container) {
+                            container.innerHTML = `<div class="service-item">
+                                <div class="service-info">
+                                    <span class="service-icon">⚠️</span>
+                                    <div class="service-details">
+                                        <div class="service-name">检查失败</div>
+                                        <div class="service-desc">${err.message || '未知错误'}</div>
+                                    </div>
+                                </div>
+                                <div class="service-status">
+                                    <span class="status-badge status-down">异常</span>
+                                </div>
+                            </div>`;
+                        }
+                    }),
+                    loadApiCount().catch(err => {
+                        console.error('加载API访问次数失败:', err);
+                    })
+                ]);
+                console.log('健康检查完成');
+            } catch (error) {
+                console.error('初始化失败:', error);
+            }
+        }
+
+        // 确保DOM完全加载后再执行
+        function startCheck() {
+            if (document.readyState === 'loading') {
+                document.addEventListener('DOMContentLoaded', () => {
+                    setTimeout(init, 200);
+                });
+            } else {
+                // DOM已经加载完成，延迟一点确保所有元素都存在
+                setTimeout(init, 200);
+            }
+        }
+
+        // 获取API访问次数
+        async function loadApiCount() {
+            try {
+                const response = await fetch('/api/v1/statistics/total', {
+                    method: 'GET',
+                    headers: {
+                        'Accept': 'application/json'
+                    }
+                });
+
+                if (response.ok) {
+                    const result = await response.json();
+                    const count = result.data || 0;
+                    const apiCountElement = document.getElementById('apiCount');
+                    if (apiCountElement) {
+                        // 格式化数字，添加千分位分隔符
+                        apiCountElement.textContent = count.toLocaleString('zh-CN');
+                        apiCountElement.style.color = '#10b981';
+                    }
+                } else {
+                    const apiCountElement = document.getElementById('apiCount');
+                    if (apiCountElement) {
+                        apiCountElement.textContent = 'N/A';
+                        apiCountElement.style.color = '#ef4444';
+                    }
+                }
+            } catch (error) {
+                console.error('获取API访问次数失败:', error);
+                const apiCountElement = document.getElementById('apiCount');
+                if (apiCountElement) {
+                    apiCountElement.textContent = 'N/A';
+                    apiCountElement.style.color = '#ef4444';
+                }
+            }
+        }
+
+        // 启动检查
+        startCheck();
+
+        // 每60秒自动刷新系统状态和API接口状态（减少刷新频率，避免频繁请求）
+        setInterval(() => {
+            Promise.all([
+                checkSystemComponents().catch(err => console.error('刷新系统组件失败:', err)),
+                checkApiEndpoints().catch(err => console.error('刷新API接口失败:', err)),
+                loadApiCount().catch(err => console.error('刷新API访问次数失败:', err))
+            ]);
+        }, 60000); // 改为60秒刷新一次
+    </script>
+</body>
+</html>
diff --git a/src/main/resources/static/view/index.html b/src/main/resources/static/view/index.html
new file mode 100644
index 0000000..dd5ba83
--- /dev/null
+++ b/src/main/resources/static/view/index.html
@@ -0,0 +1,288 @@
+<!DOCTYPE html>
+<html lang="zh-CN">
+<head>
+    <meta charset="UTF-8">
+    <meta name="viewport" content="width=device-width, initial-scale=1.0">
+    <title>智慧评标-主观分助手</title>
+    <style>
+        * {
+            margin: 0;
+            padding: 0;
+            box-sizing: border-box;
+        }
+
+        body {
+            font-family: -apple-system, BlinkMacSystemFont, 'Segoe UI', 'Roboto', 'Helvetica Neue', Arial, sans-serif;
+            background: linear-gradient(135deg, #667eea 0%, #764ba2 100%);
+            min-height: 100vh;
+            display: flex;
+            justify-content: center;
+            align-items: center;
+            padding: 20px;
+        }
+
+        .container {
+            background: white;
+            border-radius: 20px;
+            padding: 60px 80px;
+            box-shadow: 0 20px 60px rgba(0, 0, 0, 0.3);
+            max-width: 900px;
+            width: 100%;
+            text-align: center;
+        }
+
+        .header {
+            display: flex;
+            align-items: center;
+            justify-content: center;
+            gap: 20px;
+            margin-bottom: 20px;
+        }
+
+        .rocket-icon {
+            width: 48px;
+            height: 48px;
+            background: linear-gradient(135deg, #ff6b6b 0%, #4ecdc4 100%);
+            border-radius: 50%;
+            display: flex;
+            align-items: center;
+            justify-content: center;
+            font-size: 28px;
+            animation: float 3s ease-in-out infinite;
+        }
+
+        @keyframes float {
+            0%, 100% {
+                transform: translateY(0);
+            }
+            50% {
+                transform: translateY(-10px);
+            }
+        }
+
+        h1 {
+            font-size: 36px;
+            font-weight: 700;
+            color: #2c3e50;
+            margin-bottom: 10px;
+        }
+
+        .version {
+            font-size: 16px;
+            color: #7f8c8d;
+            margin-bottom: 40px;
+        }
+
+        .buttons {
+            display: flex;
+            justify-content: center;
+            gap: 20px;
+            margin-bottom: 40px;
+            flex-wrap: wrap;
+        }
+
+        .btn {
+            display: inline-flex;
+            align-items: center;
+            gap: 10px;
+            padding: 14px 28px;
+            background: linear-gradient(135deg, #667eea 0%, #764ba2 100%);
+            color: white;
+            text-decoration: none;
+            border-radius: 10px;
+            font-size: 16px;
+            font-weight: 500;
+            transition: all 0.3s ease;
+            border: none;
+            cursor: pointer;
+        }
+
+        .btn:hover {
+            transform: translateY(-2px);
+            box-shadow: 0 10px 20px rgba(102, 126, 234, 0.4);
+        }
+
+        .btn:active {
+            transform: translateY(0);
+        }
+
+        .btn-icon {
+            font-size: 18px;
+        }
+
+        .status-bar {
+            background: linear-gradient(135deg, #56ab2f 0%, #a8e063 100%);
+            border-radius: 12px;
+            padding: 16px 24px;
+            display: flex;
+            align-items: center;
+            justify-content: center;
+            gap: 10px;
+            color: white;
+            font-weight: 500;
+            font-size: 16px;
+        }
+
+        .status-icon {
+            font-size: 20px;
+        }
+
+        @media (max-width: 768px) {
+            .container {
+                padding: 40px 30px;
+            }
+
+            h1 {
+                font-size: 28px;
+            }
+
+            .buttons {
+                flex-direction: column;
+            }
+
+            .btn {
+                width: 100%;
+                justify-content: center;
+            }
+
+            .file-manager {
+                margin-top: 20px;
+            }
+
+            .file-filter {
+                display: flex;
+                gap: 10px;
+                align-items: center;
+            }
+
+            .file-filter input {
+                flex: 1;
+                padding: 8px;
+            }
+        }
+
+    </style>
+</head>
+<body>
+    <div class="container">
+        <div class="header">
+            <div class="rocket-icon">🚀</div>
+            <h1>智慧评标-主观分助手</h1>
+        </div>
+        <div class="version">版本 1.0.0</div>
+        
+        <div class="buttons">
+            <a href="/swagger-ui.html" class="btn" target="_blank">
+                <span class="btn-icon">📄</span>
+                <span>API文档</span>
+            </a>
+            <a href="/api-docs" class="btn" target="_blank">
+                <span class="btn-icon">📋</span>
+                <span>API文档JSON</span>
+            </a>
+            <a href="/view/qa-test.html" class="btn" target="_blank">
+                <span class="btn-icon">💬</span>
+                <span>智能问答测试</span>
+            </a>
+            <a href="/view/prompt-templates.html" class="btn" target="_blank">
+                <span class="btn-icon">🤖</span>
+                <span>提示词配置</span>
+            </a>
+            <a href="/view/qa-history.html" class="btn" target="_blank">
+                <span class="btn-icon">📚</span>
+                <span>问答历史</span>
+            </a>
+            <a href="/view/summary-test.html" class="btn" target="_blank">
+                <span class="btn-icon">📄</span>
+                <span>目录总结测试</span>
+            </a>
+            <a href="/view/health.html" class="btn" target="_blank">
+                <span class="btn-icon">❤️</span>
+                <span>健康检查</span>
+            </a>
+            <a href="/view/logs.html" class="btn" target="_blank">
+                <span class="btn-icon">📜</span>
+                <span>系统日志</span>
+            </a>
+            <a href="/view/files.html" class="btn">
+                <span class="btn-icon">📁</span>
+                <span>文件管理</span>
+            </a>
+        </div>
+        
+        <div class="status-bar">
+            <span class="status-icon">✓</span>
+            <span>服务运行正常</span>
+        </div>
+
+    </div>
+
+    <script>
+        // 检查服务健康状态
+        fetch('/actuator/health')
+            .then(response => response.json())
+            .then(data => {
+                const statusBar = document.querySelector('.status-bar');
+                if (data.status === 'UP') {
+                    // 检查向量化服务状态
+                    const vectorizationHealth = data.details?.vectorization;
+                    if (vectorizationHealth && vectorizationHealth.status === 'UP') {
+                        const completionRate = vectorizationHealth.details?.completion_rate || '未知';
+                        const totalFiles = vectorizationHealth.details?.total_files || 0;
+                        statusBar.innerHTML = '<span class="status-icon">✓</span><span>服务运行正常（向量化完成率: ' + completionRate + ', 文件数: ' + totalFiles + '）</span>';
+                        statusBar.style.background = 'linear-gradient(135deg, #56ab2f 0%, #a8e063 100%)';
+                    } else if (vectorizationHealth && vectorizationHealth.status === 'DOWN') {
+                        statusBar.innerHTML = '<span class="status-icon">⚠</span><span>向量化服务异常</span>';
+                        statusBar.style.background = 'linear-gradient(135deg, #f093fb 0%, #f5576c 100%)';
+                    } else {
+                        statusBar.innerHTML = '<span class="status-icon">✓</span><span>服务运行正常</span>';
+                        statusBar.style.background = 'linear-gradient(135deg, #56ab2f 0%, #a8e063 100%)';
+                    }
+                } else {
+                    statusBar.innerHTML = '<span class="status-icon">✗</span><span>服务异常</span>';
+                    statusBar.style.background = 'linear-gradient(135deg, #eb3349 0%, #f45c43 100%)';
+                }
+            })
+            .catch(() => {
+                const statusBar = document.querySelector('.status-bar');
+                statusBar.innerHTML = '<span class="status-icon">⚠</span><span>无法检查服务状态</span>';
+                statusBar.style.background = 'linear-gradient(135deg, #f093fb 0%, #f5576c 100%)';
+            });
+
+        // 添加向量化状态详情显示
+        function showVectorizationDetails() {
+            fetch('/api/v1/files/vectorization/stats')
+                .then(response => response.json())
+                .then(data => {
+                    if (data.code === 0 && data.data) {
+                        const stats = data.data;
+                        alert(
+                            '向量化统计信息:\\n' +
+                            '总文件数: ' + stats.totalFiles + '\\n' +
+                            '已完成: ' + stats.completedFiles + '\\n' +
+                            '处理中: ' + stats.processingFiles + '\\n' +
+                            '失败: ' + stats.failedFiles + '\\n' +
+                            '未开始: ' + stats.notStartedFiles + '\\n' +
+                            '完成率: ' + (stats.completionRate * 100).toFixed(1) + '%'
+                        );
+                    } else {
+                        alert('获取向量化统计信息失败');
+                    }
+                })
+                .catch(() => {
+                    alert('无法获取向量化统计信息');
+                });
+        }
+
+        // 为状态栏添加点击事件，显示详情
+        document.addEventListener('DOMContentLoaded', function() {
+            const statusBar = document.querySelector('.status-bar');
+            statusBar.style.cursor = 'pointer';
+            statusBar.title = '点击查看向量化详情';
+            statusBar.addEventListener('click', showVectorizationDetails);
+        });
+
+    </script>
+</body>
+</html>
+
diff --git a/src/main/resources/static/view/logs.html b/src/main/resources/static/view/logs.html
new file mode 100644
index 0000000..b456c52
--- /dev/null
+++ b/src/main/resources/static/view/logs.html
@@ -0,0 +1,749 @@
+<!DOCTYPE html>
+<html lang="zh-CN">
+<head>
+    <meta charset="UTF-8">
+    <meta name="viewport" content="width=device-width, initial-scale=1.0">
+    <title>系统日志查询中心</title>
+    <style>
+        * {
+            margin: 0;
+            padding: 0;
+            box-sizing: border-box;
+        }
+
+        :root {
+            color-scheme: dark;
+        }
+
+        body {
+            font-family: -apple-system, BlinkMacSystemFont, 'Segoe UI', 'Roboto', 'Helvetica Neue', Arial, sans-serif;
+            background: #0f172a;
+            color: #e2e8f0;
+            min-height: 100vh;
+            padding: 20px;
+        }
+
+        .container {
+            max-width: 1400px;
+            margin: 0 auto;
+            display: flex;
+            flex-direction: column;
+            gap: 16px;
+        }
+
+        header {
+            display: flex;
+            justify-content: space-between;
+            align-items: flex-start;
+            flex-wrap: wrap;
+            gap: 16px;
+        }
+
+        header h1 {
+            font-size: 26px;
+            font-weight: 700;
+            color: #f8fafc;
+        }
+
+        header p {
+            color: #94a3b8;
+            margin-top: 4px;
+            font-size: 14px;
+        }
+
+        .header-actions {
+            display: flex;
+            gap: 10px;
+            flex-wrap: wrap;
+        }
+
+        .btn {
+            padding: 10px 18px;
+            border-radius: 999px;
+            border: 1px solid #475569;
+            background: rgba(148, 163, 184, 0.12);
+            color: #e2e8f0;
+            font-size: 13px;
+            font-weight: 500;
+            cursor: pointer;
+            transition: all 0.2s ease;
+            text-decoration: none;
+            display: inline-flex;
+            align-items: center;
+            gap: 6px;
+        }
+
+        .btn.primary {
+            background: linear-gradient(120deg, #2563eb, #9333ea);
+            border: none;
+        }
+
+        .btn:hover {
+            transform: translateY(-1px);
+            filter: brightness(1.1);
+        }
+
+        .stats-grid {
+            display: grid;
+            grid-template-columns: repeat(auto-fit, minmax(200px, 1fr));
+            gap: 12px;
+        }
+
+        .card {
+            background: #1e293b;
+            border: 1px solid #334155;
+            border-radius: 12px;
+            padding: 16px;
+        }
+
+        .card-title {
+            font-size: 13px;
+            text-transform: uppercase;
+            letter-spacing: .08em;
+            color: #94a3b8;
+            margin-bottom: 10px;
+        }
+
+        .card-value {
+            font-size: 28px;
+            font-weight: 700;
+            color: #f8fafc;
+        }
+
+        .card-desc {
+            font-size: 12px;
+            color: #94a3b8;
+            margin-top: 6px;
+        }
+
+        .filter-card {
+            display: flex;
+            flex-wrap: wrap;
+            gap: 12px;
+            align-items: flex-end;
+        }
+
+        .filter-group {
+            display: flex;
+            flex-direction: column;
+            gap: 6px;
+            min-width: 180px;
+            flex: 1;
+        }
+
+        label {
+            font-size: 12px;
+            color: #94a3b8;
+            letter-spacing: 0.05em;
+        }
+
+        input, select {
+            background: #0f172a;
+            border: 1px solid #334155;
+            border-radius: 8px;
+            color: #f8fafc;
+            padding: 10px 12px;
+            font-size: 14px;
+            width: 100%;
+        }
+
+        input:focus, select:focus {
+            outline: none;
+            border-color: #6366f1;
+            box-shadow: 0 0 0 1px rgba(99, 102, 241, 0.4);
+        }
+
+        .table-card {
+            background: #1e293b;
+            border: 1px solid #334155;
+            border-radius: 12px;
+            padding: 0;
+            overflow: hidden;
+        }
+
+        .table-header {
+            display: flex;
+            justify-content: space-between;
+            align-items: center;
+            padding: 16px 20px;
+            border-bottom: 1px solid #334155;
+            gap: 12px;
+            flex-wrap: wrap;
+        }
+
+        .table-header h2 {
+            font-size: 16px;
+            color: #f8fafc;
+        }
+
+        .table-header small {
+            color: #94a3b8;
+            font-size: 13px;
+        }
+
+        .table-wrapper {
+            overflow-x: auto;
+        }
+
+        table {
+            width: 100%;
+            border-collapse: collapse;
+            font-size: 13px;
+        }
+
+        thead {
+            background: #0f172a;
+        }
+
+        th, td {
+            padding: 12px 16px;
+            text-align: left;
+            border-bottom: 1px solid #1f2a3d;
+        }
+
+        tbody tr:hover {
+            background: rgba(148, 163, 184, 0.05);
+        }
+
+        .level-badge {
+            padding: 4px 10px;
+            border-radius: 999px;
+            font-size: 12px;
+            font-weight: 600;
+            text-transform: uppercase;
+            letter-spacing: 0.05em;
+            border: 1px solid transparent;
+        }
+
+        .level-ERROR {
+            color: #fecaca;
+            background: rgba(248, 113, 113, 0.15);
+            border-color: rgba(248, 113, 113, 0.2);
+        }
+
+        .level-WARN {
+            color: #fde68a;
+            background: rgba(251, 191, 36, 0.15);
+            border-color: rgba(251, 191, 36, 0.2);
+        }
+
+        .level-INFO {
+            color: #bfdbfe;
+            background: rgba(59, 130, 246, 0.15);
+            border-color: rgba(59, 130, 246, 0.2);
+        }
+
+        .level-DEBUG {
+            color: #c4b5fd;
+            background: rgba(99, 102, 241, 0.15);
+            border-color: rgba(99, 102, 241, 0.2);
+        }
+
+        .level-OUT, .level-TRACE {
+            color: #cbd5f5;
+            background: rgba(148, 163, 184, 0.2);
+            border-color: rgba(148, 163, 184, 0.25);
+        }
+
+        .msg-short {
+            color: #cbd5f5;
+            font-size: 13px;
+            line-height: 1.4;
+            max-width: 520px;
+            white-space: nowrap;
+            overflow: hidden;
+            text-overflow: ellipsis;
+        }
+
+        .logger, .thread {
+            color: #94a3b8;
+            font-size: 12px;
+        }
+
+        .detail-btn {
+            padding: 6px 12px;
+            border-radius: 6px;
+            border: 1px solid #475569;
+            background: transparent;
+            color: #cbd5f5;
+            cursor: pointer;
+            font-size: 12px;
+        }
+
+        .pagination {
+            display: flex;
+            justify-content: space-between;
+            align-items: center;
+            padding: 12px 20px;
+            flex-wrap: wrap;
+            gap: 10px;
+        }
+
+        .pagination-info {
+            color: #94a3b8;
+            font-size: 13px;
+        }
+
+        .toggle {
+            display: inline-flex;
+            align-items: center;
+            gap: 6px;
+            font-size: 12px;
+            color: #94a3b8;
+        }
+
+        .toggle input {
+            width: auto;
+        }
+
+        .empty {
+            text-align: center;
+            padding: 60px 20px;
+            color: #94a3b8;
+        }
+
+        .modal {
+            position: fixed;
+            inset: 0;
+            background: rgba(15, 23, 42, 0.8);
+            display: flex;
+            align-items: center;
+            justify-content: center;
+            padding: 20px;
+            z-index: 10;
+        }
+
+        .modal-content {
+            background: #0f172a;
+            border-radius: 12px;
+            border: 1px solid #334155;
+            max-width: 800px;
+            width: 100%;
+            max-height: 90vh;
+            overflow-y: auto;
+            padding: 20px;
+        }
+
+        .modal-header {
+            display: flex;
+            justify-content: space-between;
+            align-items: center;
+            margin-bottom: 16px;
+        }
+
+        .modal-close {
+            background: transparent;
+            border: none;
+            color: #94a3b8;
+            font-size: 22px;
+            cursor: pointer;
+        }
+
+        .modal-section {
+            margin-bottom: 16px;
+        }
+
+        .modal-section h3 {
+            font-size: 14px;
+            color: #94a3b8;
+            text-transform: uppercase;
+            letter-spacing: 0.08em;
+            margin-bottom: 6px;
+        }
+
+        .code-block {
+            background: #020617;
+            border-radius: 8px;
+            border: 1px solid #1f2937;
+            padding: 12px;
+            font-family: Consolas, 'Fira Code', 'SFMono-Regular', Menlo, monospace;
+            font-size: 12px;
+            white-space: pre-wrap;
+            color: #e2e8f0;
+        }
+
+        .hidden {
+            display: none;
+        }
+
+        @media (max-width: 768px) {
+            header {
+                flex-direction: column;
+            }
+
+            .filter-group {
+                min-width: 100%;
+            }
+
+            .table-wrapper {
+                overflow-x: auto;
+            }
+        }
+    </style>
+</head>
+<body>
+    <div class="container">
+        <header>
+            <div>
+                <h1>系统日志查询中心</h1>
+                <p>查看、筛选并排查系统运行日志，支持多条件过滤与详情查看</p>
+            </div>
+            <div class="header-actions">
+                <a class="btn" href="/health.html">返回健康检查</a>
+                <button class="btn" id="refreshBtn">刷新日志</button>
+            </div>
+        </header>
+
+        <section class="stats-grid" id="statsGrid">
+            <div class="card">
+                <div class="card-title">日志总量</div>
+                <div class="card-value" id="statTotal">-</div>
+                <div class="card-desc">全部记录数量</div>
+            </div>
+            <div class="card">
+                <div class="card-title">最近24小时</div>
+                <div class="card-value" id="stat24h">-</div>
+                <div class="card-desc">24小时内新增日志</div>
+            </div>
+            <div class="card">
+                <div class="card-title">最新日志时间</div>
+                <div class="card-value" style="font-size:18px" id="statLatest">-</div>
+                <div class="card-desc">最近一次日志写入</div>
+            </div>
+            <div class="card">
+                <div class="card-title">错误/告警</div>
+                <div class="card-value" id="statErrorWarn">-</div>
+                <div class="card-desc">ERROR + WARN 总计</div>
+            </div>
+        </section>
+
+        <section class="card">
+            <form class="filter-card" id="filterForm">
+                <div class="filter-group">
+                    <label for="level">日志级别</label>
+                    <select id="level" name="level">
+                        <option value="ALL">全部</option>
+                        <option value="ERROR">ERROR</option>
+                        <option value="WARN">WARN</option>
+                        <option value="INFO">INFO</option>
+                        <option value="DEBUG">DEBUG</option>
+                        <option value="TRACE">TRACE</option>
+                        <option value="OUT">OUT</option>
+                    </select>
+                </div>
+                <div class="filter-group">
+                    <label for="keyword">关键字</label>
+                    <input type="text" id="keyword" placeholder="级别 / 内容 / logger / 线程">
+                </div>
+                <div class="filter-group">
+                    <label for="startTime">开始时间</label>
+                    <input type="datetime-local" id="startTime">
+                </div>
+                <div class="filter-group">
+                    <label for="endTime">结束时间</label>
+                    <input type="datetime-local" id="endTime">
+                </div>
+                <div class="filter-group" style="min-width:120px;">
+                    <label for="pageSize">每页数量</label>
+                    <select id="pageSize">
+                        <option value="20">20</option>
+                        <option value="50">50</option>
+                        <option value="100">100</option>
+                    </select>
+                </div>
+                <div class="filter-group" style="min-width:160px;">
+                    <button type="submit" class="btn primary" style="width:100%;">查询日志</button>
+                    <button type="button" class="btn" id="resetBtn" style="width:100%; margin-top:6px;">重置条件</button>
+                </div>
+            </form>
+        </section>
+
+        <section class="table-card">
+            <div class="table-header">
+                <div>
+                    <h2>日志结果列表</h2>
+                    <small id="resultDesc">共 0 条记录</small>
+                </div>
+                <label class="toggle">
+                    <input type="checkbox" id="autoRefreshToggle">
+                    自动刷新 (30s)
+                </label>
+            </div>
+
+            <div class="table-wrapper">
+                <table>
+                    <thead>
+                        <tr>
+                            <th>时间</th>
+                            <th>级别</th>
+                            <th>来源 / 线程</th>
+                            <th>概要</th>
+                            <th style="width:90px;">操作</th>
+                        </tr>
+                    </thead>
+                    <tbody id="logTableBody">
+                        <tr><td colspan="5" class="empty">正在加载日志...</td></tr>
+                    </tbody>
+                </table>
+            </div>
+
+            <div class="pagination">
+                <div>
+                    <button class="btn" id="prevPageBtn">上一页</button>
+                    <button class="btn" id="nextPageBtn">下一页</button>
+                </div>
+                <div class="pagination-info" id="paginationInfo">第 0 / 0 页</div>
+            </div>
+        </section>
+    </div>
+
+    <div class="modal hidden" id="logModal">
+        <div class="modal-content">
+            <div class="modal-header">
+                <div>
+                    <div class="level-badge" id="modalLevel">INFO</div>
+                    <div class="logger" id="modalTime">-</div>
+                </div>
+                <button class="modal-close" id="modalCloseBtn">&times;</button>
+            </div>
+            <div class="modal-section">
+                <h3>基本信息</h3>
+                <div class="code-block" id="modalBasic"></div>
+            </div>
+            <div class="modal-section">
+                <h3>日志内容</h3>
+                <div class="code-block" id="modalContent"></div>
+            </div>
+            <div class="modal-section hidden" id="modalStackSection">
+                <h3>异常堆栈</h3>
+                <div class="code-block" id="modalStack"></div>
+            </div>
+        </div>
+    </div>
+
+    <script>
+        const state = {
+            page: 0,
+            size: 20,
+            level: 'ALL',
+            keyword: '',
+            startTime: '',
+            endTime: '',
+            autoRefresh: false,
+            timer: null,
+            currentLogs: []
+        };
+
+        const levelCounts = {};
+
+        document.addEventListener('DOMContentLoaded', () => {
+            document.getElementById('filterForm').addEventListener('submit', handleFilterSubmit);
+            document.getElementById('resetBtn').addEventListener('click', resetFilters);
+            document.getElementById('prevPageBtn').addEventListener('click', () => changePage(-1));
+            document.getElementById('nextPageBtn').addEventListener('click', () => changePage(1));
+            document.getElementById('pageSize').addEventListener('change', handlePageSizeChange);
+            document.getElementById('refreshBtn').addEventListener('click', () => fetchLogs(true));
+            document.getElementById('autoRefreshToggle').addEventListener('change', handleAutoRefreshToggle);
+            document.getElementById('modalCloseBtn').addEventListener('click', closeModal);
+            document.getElementById('logModal').addEventListener('click', (e) => {
+                if (e.target.id === 'logModal') {
+                    closeModal();
+                }
+            });
+            fetchStats();
+            fetchLogs();
+        });
+
+        function handleFilterSubmit(event) {
+            event.preventDefault();
+            state.page = 0;
+            state.level = document.getElementById('level').value;
+            state.keyword = document.getElementById('keyword').value.trim();
+            state.startTime = document.getElementById('startTime').value;
+            state.endTime = document.getElementById('endTime').value;
+            state.size = parseInt(document.getElementById('pageSize').value, 10);
+            fetchLogs(true);
+        }
+
+        function resetFilters() {
+            document.getElementById('level').value = 'ALL';
+            document.getElementById('keyword').value = '';
+            document.getElementById('startTime').value = '';
+            document.getElementById('endTime').value = '';
+            document.getElementById('pageSize').value = '20';
+            state.page = 0;
+            state.level = 'ALL';
+            state.keyword = '';
+            state.startTime = '';
+            state.endTime = '';
+            state.size = 20;
+            fetchLogs(true);
+        }
+
+        function changePage(delta) {
+            const newPage = state.page + delta;
+            if (newPage < 0) return;
+            state.page = newPage;
+            fetchLogs();
+        }
+
+        function handlePageSizeChange(event) {
+            state.size = parseInt(event.target.value, 10);
+            state.page = 0;
+            fetchLogs(true);
+        }
+
+        function handleAutoRefreshToggle(event) {
+            state.autoRefresh = event.target.checked;
+            if (state.autoRefresh) {
+                state.timer = setInterval(fetchLogs, 30000);
+            } else if (state.timer) {
+                clearInterval(state.timer);
+                state.timer = null;
+            }
+        }
+
+        async function fetchLogs(showLoading = false) {
+            const tbody = document.getElementById('logTableBody');
+            if (showLoading) {
+                tbody.innerHTML = '<tr><td colspan="5" class="empty">查询中...</td></tr>';
+            }
+            const params = new URLSearchParams();
+            params.append('page', state.page);
+            params.append('size', state.size);
+            if (state.level && state.level !== 'ALL') params.append('level', state.level);
+            if (state.keyword) params.append('keyword', state.keyword);
+            if (state.startTime) params.append('startTime', state.startTime);
+            if (state.endTime) params.append('endTime', state.endTime);
+
+            try {
+                const response = await fetch(`/api/logs?${params.toString()}`);
+                if (!response.ok) {
+                    throw new Error('请求失败: ' + response.statusText);
+                }
+                const data = await response.json();
+                renderLogs(data);
+            } catch (error) {
+                console.error('加载日志失败:', error);
+                tbody.innerHTML = `<tr><td colspan="5" class="empty">加载失败：${error.message}</td></tr>`;
+            }
+        }
+
+        async function fetchStats() {
+            try {
+                const response = await fetch('/api/logs/stats');
+                if (!response.ok) {
+                    throw new Error('获取统计失败');
+                }
+                const data = await response.json();
+                document.getElementById('statTotal').textContent = formatNumber(data.total || 0);
+                document.getElementById('stat24h').textContent = formatNumber(data.last24Hours || 0);
+                if (data.latest && data.latest.createTime) {
+                    document.getElementById('statLatest').textContent = formatDateTime(data.latest.createTime);
+                } else {
+                    document.getElementById('statLatest').textContent = '-';
+                }
+                const errorCount = (data.levelStats && data.levelStats.ERROR) || 0;
+                const warnCount = (data.levelStats && data.levelStats.WARN) || 0;
+                document.getElementById('statErrorWarn').textContent = formatNumber(errorCount + warnCount);
+            } catch (error) {
+                console.warn('统计信息获取失败:', error);
+            }
+        }
+
+        function renderLogs(data) {
+            const tbody = document.getElementById('logTableBody');
+            state.currentLogs = data.content || [];
+            const content = state.currentLogs;
+
+            document.getElementById('resultDesc').textContent = `共 ${formatNumber(data.totalElements || 0)} 条记录`;
+            document.getElementById('paginationInfo').textContent =
+                `第 ${content.length === 0 ? 0 : (data.page + 1)} / ${data.totalPages || 0} 页`;
+
+            if (!content.length) {
+                tbody.innerHTML = '<tr><td colspan="5" class="empty">未找到符合条件的日志</td></tr>';
+                return;
+            }
+
+            tbody.innerHTML = '';
+            content.forEach((log, index) => {
+                const row = document.createElement('tr');
+                row.innerHTML = `
+                    <td>
+                        <div>${formatDateTime(log.createTime)}</div>
+                        <div class="logger">${log.traceId || ''}</div>
+                    </td>
+                    <td><span class="level-badge level-${log.logLevel || 'INFO'}">${log.logLevel || '-'}</span></td>
+                    <td>
+                        <div class="logger">${log.loggerName || '-'}</div>
+                        <div class="thread">${log.threadName || ''}</div>
+                    </td>
+                    <td>
+                        <div class="msg-short" title="${escapeHtml(log.logContent || '')}">
+                            ${escapeHtml(log.logContent || '')}
+                        </div>
+                    </td>
+                    <td>
+                        <button class="detail-btn" data-index="${index}">详情</button>
+                    </td>
+                `;
+                row.querySelector('.detail-btn').addEventListener('click', () => openModal(index));
+                tbody.appendChild(row);
+            });
+        }
+
+        function openModal(index) {
+            const log = state.currentLogs[index];
+            if (!log) return;
+            document.getElementById('modalLevel').textContent = log.logLevel || 'INFO';
+            document.getElementById('modalLevel').className = `level-badge level-${log.logLevel || 'INFO'}`;
+            document.getElementById('modalTime').textContent = formatDateTime(log.createTime);
+            document.getElementById('modalBasic').textContent = [
+                `Logger : ${log.loggerName || '-'}`,
+                `线程   : ${log.threadName || '-'}`,
+                `类/方法: ${log.className || '-'}#${log.methodName || '-'} (${log.lineNumber || '-'})`,
+                `TraceId: ${log.traceId || '-'}`,
+                `应用   : ${log.applicationName || '-'}`,
+                `用户   : ${log.userId || '-'}`
+            ].join('\n');
+            document.getElementById('modalContent').textContent = log.fullMessage || log.logContent || '';
+            if (log.exceptionStack) {
+                document.getElementById('modalStack').textContent = log.exceptionStack;
+                document.getElementById('modalStackSection').classList.remove('hidden');
+            } else {
+                document.getElementById('modalStackSection').classList.add('hidden');
+            }
+            document.getElementById('logModal').classList.remove('hidden');
+        }
+
+        function closeModal() {
+            document.getElementById('logModal').classList.add('hidden');
+        }
+
+        function formatDateTime(value) {
+            if (!value) return '-';
+            try {
+                const date = new Date(value);
+                if (Number.isNaN(date.getTime())) return value;
+                return date.toLocaleString('zh-CN', { hour12: false });
+            } catch {
+                return value;
+            }
+        }
+
+        function formatNumber(num) {
+            return (num || 0).toLocaleString();
+        }
+
+        function escapeHtml(text) {
+            return text
+                .replace(/&/g, '&amp;')
+                .replace(/</g, '&lt;')
+                .replace(/>/g, '&gt;')
+                .replace(/"/g, '&quot;')
+                .replace(/'/g, '&#39;');
+        }
+    </script>
+</body>
+</html>
+
diff --git a/src/main/resources/static/view/prompt-templates.html b/src/main/resources/static/view/prompt-templates.html
new file mode 100644
index 0000000..d404836
--- /dev/null
+++ b/src/main/resources/static/view/prompt-templates.html
@@ -0,0 +1,626 @@
+<!DOCTYPE html>
+<html lang="zh-CN">
+<head>
+    <meta charset="UTF-8">
+    <meta name="viewport" content="width=device-width, initial-scale=1.0">
+    <title>提示词模板配置</title>
+    <style>
+        body {
+            font-family: 'Microsoft YaHei', Arial, sans-serif;
+            margin: 0;
+            padding: 20px;
+            background-color: #f5f5f5;
+        }
+        .container {
+            max-width: 1200px;
+            margin: 0 auto;
+            background: white;
+            border-radius: 8px;
+            box-shadow: 0 2px 10px rgba(0,0,0,0.1);
+            overflow: hidden;
+        }
+        .header {
+            background: linear-gradient(135deg, #667eea 0%, #764ba2 100%);
+            color: white;
+            padding: 20px;
+            text-align: center;
+        }
+        .header h1 {
+            margin: 0;
+            font-size: 28px;
+        }
+        .header p {
+            margin: 10px 0 0 0;
+            opacity: 0.9;
+        }
+        .content {
+            padding: 30px;
+        }
+        .toolbar {
+            display: flex;
+            justify-content: space-between;
+            align-items: center;
+            margin-bottom: 20px;
+            padding: 15px;
+            background: #f8f9fa;
+            border-radius: 6px;
+        }
+        .search-box {
+            display: flex;
+            gap: 10px;
+            align-items: center;
+        }
+        .search-box select, .search-box input {
+            padding: 8px 12px;
+            border: 1px solid #ddd;
+            border-radius: 4px;
+            font-size: 14px;
+        }
+        .btn {
+            padding: 8px 16px;
+            border: none;
+            border-radius: 4px;
+            cursor: pointer;
+            font-size: 14px;
+            transition: all 0.3s ease;
+        }
+        .btn-primary {
+            background: #007bff;
+            color: white;
+        }
+        .btn-primary:hover {
+            background: #0056b3;
+        }
+        .btn-success {
+            background: #28a745;
+            color: white;
+        }
+        .btn-success:hover {
+            background: #218838;
+        }
+        .btn-danger {
+            background: #dc3545;
+            color: white;
+        }
+        .btn-danger:hover {
+            background: #c82333;
+        }
+        .btn-secondary {
+            background: #6c757d;
+            color: white;
+        }
+        .btn-secondary:hover {
+            background: #5a6268;
+        }
+        .template-list {
+            display: grid;
+            grid-template-columns: repeat(auto-fill, minmax(350px, 1fr));
+            gap: 20px;
+        }
+        .template-card {
+            border: 1px solid #ddd;
+            border-radius: 8px;
+            padding: 20px;
+            background: white;
+            box-shadow: 0 2px 4px rgba(0,0,0,0.1);
+            transition: transform 0.2s, box-shadow 0.2s;
+        }
+        .template-card:hover {
+            transform: translateY(-2px);
+            box-shadow: 0 4px 8px rgba(0,0,0,0.15);
+        }
+        .template-header {
+            display: flex;
+            justify-content: space-between;
+            align-items: flex-start;
+            margin-bottom: 15px;
+        }
+        .template-title {
+            font-size: 18px;
+            font-weight: bold;
+            color: #333;
+            margin: 0;
+        }
+        .template-status {
+            padding: 4px 8px;
+            border-radius: 12px;
+            font-size: 12px;
+            font-weight: bold;
+            text-transform: uppercase;
+        }
+        .status-active {
+            background: #d4edda;
+            color: #155724;
+        }
+        .status-inactive {
+            background: #f8d7da;
+            color: #721c24;
+        }
+        .template-meta {
+            color: #666;
+            font-size: 14px;
+            margin-bottom: 10px;
+        }
+        .template-content {
+            color: #555;
+            font-size: 14px;
+            line-height: 1.5;
+            margin-bottom: 15px;
+            max-height: 80px;
+            overflow: hidden;
+            text-overflow: ellipsis;
+        }
+        .template-actions {
+            display: flex;
+            gap: 8px;
+            justify-content: flex-end;
+        }
+        .btn-sm {
+            padding: 4px 8px;
+            font-size: 12px;
+        }
+        .modal {
+            display: none;
+            position: fixed;
+            top: 0;
+            left: 0;
+            width: 100%;
+            height: 100%;
+            background: rgba(0,0,0,0.5);
+            z-index: 1000;
+        }
+        .modal-content {
+            position: absolute;
+            top: 50%;
+            left: 50%;
+            transform: translate(-50%, -50%);
+            background: white;
+            padding: 30px;
+            border-radius: 8px;
+            width: 90%;
+            max-width: 600px;
+            max-height: 80vh;
+            overflow-y: auto;
+        }
+        .modal-header {
+            display: flex;
+            justify-content: space-between;
+            align-items: center;
+            margin-bottom: 20px;
+            padding-bottom: 15px;
+            border-bottom: 1px solid #ddd;
+        }
+        .modal-title {
+            margin: 0;
+            font-size: 20px;
+            color: #333;
+        }
+        .close {
+            background: none;
+            border: none;
+            font-size: 24px;
+            cursor: pointer;
+            color: #666;
+        }
+        .form-group {
+            margin-bottom: 20px;
+        }
+        .form-group label {
+            display: block;
+            margin-bottom: 5px;
+            font-weight: 500;
+            color: #333;
+        }
+        .form-group input, .form-group select, .form-group textarea {
+            width: 100%;
+            padding: 10px;
+            border: 1px solid #ddd;
+            border-radius: 4px;
+            font-size: 14px;
+            box-sizing: border-box;
+        }
+        .form-group textarea {
+            resize: vertical;
+            min-height: 120px;
+        }
+        .form-row {
+            display: grid;
+            grid-template-columns: 1fr 1fr;
+            gap: 20px;
+        }
+        .form-actions {
+            display: flex;
+            gap: 10px;
+            justify-content: flex-end;
+            margin-top: 30px;
+            padding-top: 20px;
+            border-top: 1px solid #ddd;
+        }
+        .loading {
+            text-align: center;
+            padding: 40px;
+            color: #666;
+        }
+        .error-message {
+            background: #f8d7da;
+            color: #721c24;
+            padding: 10px;
+            border-radius: 4px;
+            margin-bottom: 15px;
+        }
+        .success-message {
+            background: #d4edda;
+            color: #155724;
+            padding: 10px;
+            border-radius: 4px;
+            margin-bottom: 15px;
+        }
+    </style>
+</head>
+<body>
+    <div class="container">
+        <div class="header">
+            <h1>🤖 提示词模板配置</h1>
+            <p>管理系统智能问答的提示词模板</p>
+        </div>
+
+        <div class="content">
+            <div class="toolbar">
+                <div class="search-box">
+                    <select id="templateTypeFilter">
+                        <option value="">全部类型</option>
+                        <option value="qa">智能问答</option>
+                        <option value="summary">摘要</option>
+                    </select>
+                    <input type="text" id="searchInput" placeholder="搜索模板名称...">
+                    <button class="btn btn-secondary" onclick="searchTemplates()">搜索</button>
+                </div>
+                <button class="btn btn-primary" onclick="openCreateModal()">+ 新建模板</button>
+            </div>
+
+            <div id="messageContainer"></div>
+
+            <div id="templateList" class="template-list">
+                <div class="loading">正在加载模板列表...</div>
+            </div>
+        </div>
+    </div>
+
+    <!-- 创建/编辑模板模态框 -->
+    <div id="templateModal" class="modal">
+        <div class="modal-content">
+            <div class="modal-header">
+                <h2 id="modalTitle" class="modal-title">新建提示词模板</h2>
+                <button class="close" onclick="closeModal()">&times;</button>
+            </div>
+
+            <form id="templateForm">
+                <input type="hidden" id="templateId" name="templateId">
+
+                <div class="form-row">
+                    <div class="form-group">
+                        <label for="templateName">模板名称 *</label>
+                        <input type="text" id="templateName" name="templateName" required>
+                    </div>
+                    <div class="form-group">
+                        <label for="templateType">模板类型 *</label>
+                        <select id="templateType" name="templateType" required>
+                            <option value="qa">智能问答</option>
+                            <option value="summary">摘要</option>
+                        </select>
+                    </div>
+                </div>
+
+                <div class="form-group">
+                    <label for="promptContent">提示词内容 *</label>
+                    <textarea id="promptContent" name="promptContent" placeholder="请输入提示词内容，支持变量替换：{query} - 用户问题，{document_content} - 文档内容" required></textarea>
+                </div>
+
+                <div class="form-group">
+                    <label for="description">模板描述</label>
+                    <textarea id="description" name="description" placeholder="请输入模板的详细描述"></textarea>
+                </div>
+
+                <div class="form-row">
+                    <div class="form-group">
+                        <label for="sortOrder">排序号</label>
+                        <input type="number" id="sortOrder" name="sortOrder" value="0">
+                    </div>
+                    <div class="form-group">
+                        <label for="isActive">是否启用</label>
+                        <select id="isActive" name="isActive">
+                            <option value="true">启用</option>
+                            <option value="false">禁用</option>
+                        </select>
+                    </div>
+                </div>
+
+                <div class="form-actions">
+                    <button type="button" class="btn btn-secondary" onclick="closeModal()">取消</button>
+                    <button type="submit" class="btn btn-primary">保存</button>
+                </div>
+            </form>
+        </div>
+    </div>
+
+    <script>
+        // 全局变量
+        let currentTemplates = [];
+        let currentTemplateId = null;
+
+        // 页面加载完成后的初始化
+        document.addEventListener('DOMContentLoaded', function() {
+            loadTemplates();
+            setupEventListeners();
+        });
+
+        // 设置事件监听器
+        function setupEventListeners() {
+            document.getElementById('templateTypeFilter').addEventListener('change', filterTemplates);
+            document.getElementById('searchInput').addEventListener('input', filterTemplates);
+            document.getElementById('templateForm').addEventListener('submit', handleFormSubmit);
+        }
+
+        // 加载模板列表
+        async function loadTemplates() {
+            try {
+                const response = await fetch('/api/v1/prompt-templates');
+                if (!response.ok) {
+                    throw new Error(`HTTP ${response.status}: ${response.statusText}`);
+                }
+                currentTemplates = await response.json();
+                renderTemplates(currentTemplates);
+            } catch (error) {
+                console.error('加载模板列表失败:', error);
+                showMessage('加载模板列表失败: ' + error.message, 'error');
+                renderTemplates([]);
+            }
+        }
+
+        // 渲染模板列表
+        function renderTemplates(templates) {
+            const container = document.getElementById('templateList');
+
+            if (templates.length === 0) {
+                container.innerHTML = '<div class="loading">暂无模板数据</div>';
+                return;
+            }
+
+            container.innerHTML = templates.map(template => `
+                <div class="template-card">
+                    <div class="template-header">
+                        <h3 class="template-title">${escapeHtml(template.templateName)}</h3>
+                        <span class="template-status ${template.isActive ? 'status-active' : 'status-inactive'}">
+                            ${template.isActive ? '启用' : '禁用'}
+                        </span>
+                    </div>
+                    <div class="template-meta">
+                        类型: ${getTemplateTypeText(template.templateType)} |
+                        创建时间: ${new Date(template.createTime).toLocaleString()}
+                    </div>
+                    <div class="template-content">
+                        ${escapeHtml(template.description || '暂无描述')}
+                    </div>
+                    <div class="template-actions">
+                        <button class="btn btn-sm btn-secondary" onclick="viewTemplate('${template.templateId}')">查看</button>
+                        <button class="btn btn-sm btn-primary" onclick="editTemplate('${template.templateId}')">编辑</button>
+                        <button class="btn btn-sm ${template.isActive ? 'btn-danger' : 'btn-success'}"
+                                onclick="toggleTemplateStatus('${template.templateId}', ${!template.isActive})">
+                            ${template.isActive ? '禁用' : '启用'}
+                        </button>
+                        <button class="btn btn-sm btn-danger" onclick="deleteTemplate('${template.templateId}')">删除</button>
+                    </div>
+                </div>
+            `).join('');
+        }
+
+        // 过滤模板
+        function filterTemplates() {
+            const typeFilter = document.getElementById('templateTypeFilter').value;
+            const searchText = document.getElementById('searchInput').value.toLowerCase();
+
+            const filtered = currentTemplates.filter(template => {
+                const matchesType = !typeFilter || template.templateType === typeFilter;
+                const matchesSearch = !searchText ||
+                    template.templateName.toLowerCase().includes(searchText) ||
+                    (template.description && template.description.toLowerCase().includes(searchText));
+
+                return matchesType && matchesSearch;
+            });
+
+            renderTemplates(filtered);
+        }
+
+        // 搜索模板
+        function searchTemplates() {
+            filterTemplates();
+        }
+
+        // 打开创建模态框
+        function openCreateModal() {
+            currentTemplateId = null;
+            document.getElementById('modalTitle').textContent = '新建提示词模板';
+            document.getElementById('templateForm').reset();
+            document.getElementById('templateModal').style.display = 'block';
+        }
+
+        // 编辑模板
+        function editTemplate(templateId) {
+            const template = currentTemplates.find(t => t.templateId === templateId);
+            if (!template) {
+                showMessage('模板不存在', 'error');
+                return;
+            }
+
+            currentTemplateId = templateId;
+            document.getElementById('modalTitle').textContent = '编辑提示词模板';
+
+            // 填充表单
+            document.getElementById('templateId').value = template.templateId;
+            document.getElementById('templateName').value = template.templateName;
+            document.getElementById('templateType').value = template.templateType;
+            document.getElementById('promptContent').value = template.promptContent;
+            document.getElementById('description').value = template.description || '';
+            document.getElementById('sortOrder').value = template.sortOrder;
+            document.getElementById('isActive').value = template.isActive.toString();
+
+            document.getElementById('templateModal').style.display = 'block';
+        }
+
+        // 查看模板
+        function viewTemplate(templateId) {
+            const template = currentTemplates.find(t => t.templateId === templateId);
+            if (!template) {
+                showMessage('模板不存在', 'error');
+                return;
+            }
+
+            // 显示详细内容
+            const content = `
+提示词内容：
+${template.promptContent}
+
+描述：
+${template.description || '暂无描述'}
+
+创建时间：${new Date(template.createTime).toLocaleString()}
+更新时间：${new Date(template.updateTime).toLocaleString()}
+创建人：${template.createdBy}
+更新人：${template.updatedBy}
+            `;
+
+            alert(content);
+        }
+
+        // 切换模板状态
+        async function toggleTemplateStatus(templateId, isActive) {
+            try {
+                const response = await fetch(`/api/v1/prompt-templates/${templateId}/status?isActive=${isActive}&userId=admin`, {
+                    method: 'PUT'
+                });
+
+                if (!response.ok) {
+                    throw new Error(`HTTP ${response.status}: ${response.statusText}`);
+                }
+
+                showMessage(`模板${isActive ? '启用' : '禁用'}成功`, 'success');
+                await loadTemplates();
+            } catch (error) {
+                console.error('切换模板状态失败:', error);
+                showMessage('切换模板状态失败: ' + error.message, 'error');
+            }
+        }
+
+        // 删除模板
+        async function deleteTemplate(templateId) {
+            if (!confirm('确定要删除这个模板吗？此操作不可撤销。')) {
+                return;
+            }
+
+            try {
+                const response = await fetch(`/api/v1/prompt-templates/${templateId}`, {
+                    method: 'DELETE'
+                });
+
+                if (!response.ok) {
+                    throw new Error(`HTTP ${response.status}: ${response.statusText}`);
+                }
+
+                showMessage('模板删除成功', 'success');
+                await loadTemplates();
+            } catch (error) {
+                console.error('删除模板失败:', error);
+                showMessage('删除模板失败: ' + error.message, 'error');
+            }
+        }
+
+        // 处理表单提交
+        async function handleFormSubmit(event) {
+            event.preventDefault();
+
+            const formData = new FormData(event.target);
+            const templateData = {
+                templateId: formData.get('templateId'),
+                templateName: formData.get('templateName'),
+                templateType: formData.get('templateType'),
+                promptContent: formData.get('promptContent'),
+                description: formData.get('description'),
+                sortOrder: parseInt(formData.get('sortOrder')) || 0,
+                isActive: formData.get('isActive') === 'true'
+            };
+
+            try {
+                let response;
+                if (currentTemplateId) {
+                    // 更新模板
+                    response = await fetch(`/api/v1/prompt-templates/${currentTemplateId}?userId=admin`, {
+                        method: 'PUT',
+                        headers: {
+                            'Content-Type': 'application/json'
+                        },
+                        body: JSON.stringify(templateData)
+                    });
+                } else {
+                    // 创建模板
+                    response = await fetch('/api/v1/prompt-templates?userId=admin', {
+                        method: 'POST',
+                        headers: {
+                            'Content-Type': 'application/json'
+                        },
+                        body: JSON.stringify(templateData)
+                    });
+                }
+
+                if (!response.ok) {
+                    const errorData = await response.json();
+                    throw new Error(errorData.message || `HTTP ${response.status}: ${response.statusText}`);
+                }
+
+                const result = await response.json();
+                showMessage(`模板${currentTemplateId ? '更新' : '创建'}成功`, 'success');
+                closeModal();
+                await loadTemplates();
+
+            } catch (error) {
+                console.error('保存模板失败:', error);
+                showMessage('保存模板失败: ' + error.message, 'error');
+            }
+        }
+
+        // 关闭模态框
+        function closeModal() {
+            document.getElementById('templateModal').style.display = 'none';
+            document.getElementById('templateForm').reset();
+            currentTemplateId = null;
+        }
+
+        // 显示消息
+        function showMessage(message, type) {
+            const container = document.getElementById('messageContainer');
+            const className = type === 'error' ? 'error-message' : 'success-message';
+            container.innerHTML = `<div class="${className}">${message}</div>`;
+
+            // 3秒后自动隐藏
+            setTimeout(() => {
+                container.innerHTML = '';
+            }, 3000);
+        }
+
+        // 获取模板类型文本
+        function getTemplateTypeText(type) {
+            const typeMap = {
+                'qa': '智能问答',
+                'summary': '摘要'
+            };
+            return typeMap[type] || type;
+        }
+
+        // HTML转义
+        function escapeHtml(text) {
+            const div = document.createElement('div');
+            div.textContent = text;
+            return div.innerHTML;
+        }
+    </script>
+</body>
+</html>
diff --git a/src/main/resources/static/view/qa-history.html b/src/main/resources/static/view/qa-history.html
new file mode 100644
index 0000000..65db6cc
--- /dev/null
+++ b/src/main/resources/static/view/qa-history.html
@@ -0,0 +1,465 @@
+<!DOCTYPE html>
+<html lang="zh-CN">
+<head>
+    <meta charset="UTF-8">
+    <meta name="viewport" content="width=device-width, initial-scale=1.0">
+    <title>问答历史记录</title>
+    <style>
+        body {
+            font-family: 'Microsoft YaHei', Arial, sans-serif;
+            margin: 0;
+            padding: 20px;
+            background-color: #f5f5f5;
+            line-height: 1.6;
+        }
+        .container {
+            max-width: 1400px;
+            margin: 0 auto;
+            background: white;
+            border-radius: 8px;
+            box-shadow: 0 2px 10px rgba(0,0,0,0.1);
+            overflow: hidden;
+        }
+        .header {
+            background: linear-gradient(135deg, #667eea 0%, #764ba2 100%);
+            color: white;
+            padding: 20px;
+            text-align: center;
+        }
+        .header h1 {
+            margin: 0;
+            font-size: 28px;
+        }
+        .header p {
+            margin: 10px 0 0 0;
+            opacity: 0.9;
+        }
+        .content {
+            padding: 30px;
+        }
+        .section {
+            margin-bottom: 30px;
+            padding: 20px;
+            border: 1px solid #e0e0e0;
+            border-radius: 8px;
+            background: #fafafa;
+        }
+        .section h3 {
+            margin-top: 0;
+            color: #333;
+            border-bottom: 2px solid #667eea;
+            padding-bottom: 10px;
+        }
+        .form-group {
+            margin-bottom: 20px;
+        }
+        .form-group label {
+            display: block;
+            margin-bottom: 5px;
+            font-weight: 600;
+            color: #333;
+        }
+        .form-group input, .form-group select {
+            width: 100%;
+            padding: 12px;
+            border: 1px solid #ddd;
+            border-radius: 6px;
+            font-size: 14px;
+            box-sizing: border-box;
+        }
+        .btn {
+            padding: 10px 20px;
+            border: none;
+            border-radius: 6px;
+            cursor: pointer;
+            font-size: 14px;
+            font-weight: 500;
+            transition: all 0.3s ease;
+            display: inline-flex;
+            align-items: center;
+            gap: 8px;
+        }
+        .btn-primary {
+            background: #667eea;
+            color: white;
+        }
+        .btn-primary:hover {
+            background: #5a6fd8;
+            transform: translateY(-1px);
+        }
+        .btn-success {
+            background: #28a745;
+            color: white;
+        }
+        .btn-success:hover {
+            background: #218838;
+        }
+        .btn-secondary {
+            background: #6c757d;
+            color: white;
+        }
+        .btn-secondary:hover {
+            background: #5a6268;
+        }
+        .btn-group {
+            display: flex;
+            gap: 10px;
+            margin-top: 10px;
+        }
+        .sessions-list {
+            max-height: 400px;
+            overflow-y: auto;
+            border: 1px solid #ddd;
+            border-radius: 4px;
+            background: white;
+        }
+        .session-item {
+            display: flex;
+            justify-content: space-between;
+            align-items: center;
+            padding: 15px;
+            border-bottom: 1px solid #eee;
+            cursor: pointer;
+            transition: background-color 0.3s ease;
+        }
+        .session-item:hover {
+            background-color: #f8f9fa;
+        }
+        .session-item:last-child {
+            border-bottom: none;
+        }
+        .session-info {
+            flex: 1;
+        }
+        .session-id {
+            font-weight: 500;
+            color: #333;
+            margin-bottom: 5px;
+        }
+        .session-meta {
+            font-size: 12px;
+            color: #666;
+        }
+        .session-query {
+            font-size: 14px;
+            color: #555;
+            margin-top: 5px;
+            overflow: hidden;
+            text-overflow: ellipsis;
+            white-space: nowrap;
+        }
+        .messages-area {
+            margin-top: 20px;
+            padding: 20px;
+            border: 1px solid #ddd;
+            border-radius: 6px;
+            background: #f8f9fa;
+            min-height: 300px;
+        }
+        .message-item {
+            margin-bottom: 20px;
+            padding: 15px;
+            border-radius: 8px;
+            position: relative;
+        }
+        .message-user {
+            background: #e3f2fd;
+            border-left: 4px solid #2196f3;
+            margin-left: 20px;
+        }
+        .message-ai {
+            background: #f3e5f5;
+            border-left: 4px solid #9c27b0;
+            margin-right: 20px;
+        }
+        .message-header {
+            display: flex;
+            justify-content: space-between;
+            margin-bottom: 10px;
+        }
+        .message-sender {
+            font-weight: 600;
+            color: #333;
+        }
+        .message-time {
+            font-size: 12px;
+            color: #666;
+        }
+        .message-content {
+            line-height: 1.6;
+            color: #333;
+            white-space: pre-wrap;
+        }
+        .loading {
+            text-align: center;
+            padding: 40px;
+            color: #666;
+        }
+        .loading::after {
+            content: '';
+            display: inline-block;
+            width: 20px;
+            height: 20px;
+            border: 2px solid #f3f3f3;
+            border-top: 2px solid #667eea;
+            border-radius: 50%;
+            animation: spin 1s linear infinite;
+            margin-left: 10px;
+        }
+        @keyframes spin {
+            0% { transform: rotate(0deg); }
+            100% { transform: rotate(360deg); }
+        }
+        .error-message {
+            background: #f8d7da;
+            color: #721c24;
+            padding: 15px;
+            border-radius: 6px;
+            border-left: 4px solid #dc3545;
+            margin-bottom: 20px;
+        }
+        .success-message {
+            background: #d4edda;
+            color: #155724;
+            padding: 15px;
+            border-radius: 6px;
+            border-left: 4px solid #28a745;
+            margin-bottom: 20px;
+        }
+        .empty-state {
+            text-align: center;
+            padding: 60px 20px;
+            color: #666;
+        }
+        .empty-state h4 {
+            margin: 0 0 10px 0;
+            color: #333;
+        }
+    </style>
+</head>
+<body>
+    <div class="container">
+        <div class="header">
+            <h1>📚 问答历史记录</h1>
+            <p>查看和管理智能问答的历史会话记录</p>
+        </div>
+
+        <div class="content">
+            <!-- 消息提示区域 -->
+            <div id="messageContainer"></div>
+
+            <!-- 查询条件 -->
+            <div class="section">
+                <h3>🔍 查询条件</h3>
+                <div class="form-group">
+                    <label for="projectId">项目ID *</label>
+                    <input type="text" id="projectId" placeholder="输入项目ID，例如：project-001" value="project-001">
+                </div>
+                <div class="form-group">
+                    <label for="userId">用户ID *</label>
+                    <input type="text" id="userId" placeholder="输入用户ID，例如：user-001" value="user-zhangsan">
+                </div>
+                <div class="btn-group">
+                    <button class="btn btn-primary" onclick="loadSessions()">
+                        🔄 加载会话列表
+                    </button>
+                    <button class="btn btn-secondary" onclick="clearAll()">
+                        🧹 清空显示
+                    </button>
+                </div>
+            </div>
+
+            <!-- 会话列表 -->
+            <div class="section">
+                <h3>💬 历史会话</h3>
+                <div id="sessionsContainer" class="sessions-list">
+                    <div class="loading">请先输入查询条件并点击"加载会话列表"</div>
+                </div>
+            </div>
+
+            <!-- 消息历史 -->
+            <div class="section">
+                <h3>📝 消息详情</h3>
+                <div id="messagesContainer" class="messages-area">
+                    <div class="empty-state">
+                        <h4>请选择一个会话</h4>
+                        <p>点击上方会话列表中的任意会话，即可查看详细的消息记录</p>
+                    </div>
+                </div>
+            </div>
+        </div>
+    </div>
+
+    <script>
+        // 全局变量
+        let currentSessions = [];
+
+        // 页面加载完成后的初始化
+        document.addEventListener('DOMContentLoaded', function() {
+            // 自动加载会话列表
+            loadSessions();
+        });
+
+        // 加载会话列表
+        async function loadSessions() {
+            const projectId = document.getElementById('projectId').value.trim();
+            const userId = document.getElementById('userId').value.trim();
+
+            if (!projectId || !userId) {
+                showMessage('请输入项目ID和用户ID', 'error');
+                return;
+            }
+
+            const container = document.getElementById('sessionsContainer');
+
+            try {
+                container.innerHTML = '<div class="loading">正在加载会话列表...</div>';
+
+                const response = await fetch(`/api/v1/qa/history/sessions?projectId=${encodeURIComponent(projectId)}&userId=${encodeURIComponent(userId)}`);
+                if (!response.ok) {
+                    throw new Error(`HTTP ${response.status}: ${response.statusText}`);
+                }
+
+                const result = await response.json();
+                if (result.code !== 0) {
+                    throw new Error(result.message || '获取会话列表失败');
+                }
+
+                currentSessions = result.data.sessions;
+                renderSessions();
+
+                if (currentSessions.length === 0) {
+                    container.innerHTML = '<div class="empty-state"><h4>暂无会话记录</h4><p>该用户在此项目下还没有进行过问答对话</p></div>';
+                }
+
+                showMessage(`找到 ${currentSessions.length} 个历史会话`, 'success');
+
+            } catch (error) {
+                console.error('加载会话列表失败:', error);
+                container.innerHTML = '<div class="error-message">加载失败，请检查网络连接和参数设置</div>';
+                showMessage('加载会话列表失败: ' + error.message, 'error');
+            }
+        }
+
+        // 渲染会话列表
+        function renderSessions() {
+            const container = document.getElementById('sessionsContainer');
+
+            if (currentSessions.length === 0) {
+                container.innerHTML = '<div class="empty-state"><h4>暂无会话记录</h4><p>该用户在此项目下还没有进行过问答对话</p></div>';
+                return;
+            }
+
+            container.innerHTML = currentSessions.map(session => `
+                <div class="session-item" onclick="loadMessages('${session.sessionId}')">
+                    <div class="session-info">
+                        <div class="session-id">会话: ${session.sessionId}</div>
+                        <div class="session-meta">开始时间: ${formatDateTime(session.startTime)}</div>
+                        <div class="session-query">${escapeHtml(session.firstQuery || '无标题')}</div>
+                    </div>
+                    <button class="btn btn-success btn-sm">
+                        查看详情
+                    </button>
+                </div>
+            `).join('');
+        }
+
+        // 加载消息详情
+        async function loadMessages(sessionId) {
+            const container = document.getElementById('messagesContainer');
+
+            try {
+                container.innerHTML = '<div class="loading">正在加载消息记录...</div>';
+
+                const response = await fetch(`/api/v1/qa/history/messages?sessionId=${encodeURIComponent(sessionId)}`);
+                if (!response.ok) {
+                    if (response.status === 404) {
+                        container.innerHTML = '<div class="empty-state"><h4>会话不存在</h4><p>该会话可能已被删除或不存在</p></div>';
+                        return;
+                    }
+                    throw new Error(`HTTP ${response.status}: ${response.statusText}`);
+                }
+
+                const result = await response.json();
+                if (result.code !== 0) {
+                    throw new Error(result.message || '获取消息记录失败');
+                }
+
+                renderMessages(result.data.messages);
+
+            } catch (error) {
+                console.error('加载消息记录失败:', error);
+                container.innerHTML = `<div class="error-message"><strong>加载失败:</strong> ${error.message}</div>`;
+            }
+        }
+
+        // 渲染消息列表
+        function renderMessages(messages) {
+            const container = document.getElementById('messagesContainer');
+
+            if (messages.length === 0) {
+                container.innerHTML = '<div class="empty-state"><h4>暂无消息记录</h4><p>该会话还没有消息记录</p></div>';
+                return;
+            }
+
+            container.innerHTML = messages.map(message => `
+                <div class="message-item message-${message.sender}">
+                    <div class="message-header">
+                        <span class="message-sender">${message.sender === 'user' ? '👤 用户' : '🤖 AI助手'}</span>
+                        <span class="message-time">${formatDateTime(message.timestamp)}</span>
+                    </div>
+                    <div class="message-content">${escapeHtml(message.content)}</div>
+                    ${message.sources ? `
+                        <div style="margin-top: 10px; padding: 8px; background: rgba(255,255,255,0.7); border-radius: 4px;">
+                            <small style="color: #666;">包含 ${Array.isArray(message.sources) ? message.sources.length : 0} 个信息来源</small>
+                        </div>
+                    ` : ''}
+                </div>
+            `).join('');
+        }
+
+        // 清空所有显示
+        function clearAll() {
+            document.getElementById('sessionsContainer').innerHTML = '<div style="padding: 20px; color: #666;">请点击"加载会话列表"查看历史记录</div>';
+            document.getElementById('messagesContainer').innerHTML = '<div class="empty-state"><h4>请选择一个会话</h4><p>点击上方会话列表中的任意会话，即可查看详细的消息记录</p></div>';
+            currentSessions = [];
+        }
+
+        // 显示消息
+        function showMessage(message, type) {
+            const container = document.getElementById('messageContainer');
+            const className = type === 'error' ? 'error-message' : 'success-message';
+            container.innerHTML = `<div class="${className}">${message}</div>`;
+
+            // 3秒后自动隐藏
+            setTimeout(() => {
+                container.innerHTML = '';
+            }, 3000);
+        }
+
+        // 格式化日期时间
+        function formatDateTime(dateTimeStr) {
+            try {
+                const date = new Date(dateTimeStr);
+                return date.toLocaleString('zh-CN', {
+                    year: 'numeric',
+                    month: '2-digit',
+                    day: '2-digit',
+                    hour: '2-digit',
+                    minute: '2-digit',
+                    second: '2-digit'
+                });
+            } catch (e) {
+                return dateTimeStr;
+            }
+        }
+
+        // HTML转义
+        function escapeHtml(text) {
+            const div = document.createElement('div');
+            div.textContent = text || '';
+            return div.innerHTML;
+        }
+    </script>
+</body>
+</html>
diff --git a/src/main/resources/static/view/qa-test.html b/src/main/resources/static/view/qa-test.html
new file mode 100644
index 0000000..75e83eb
--- /dev/null
+++ b/src/main/resources/static/view/qa-test.html
@@ -0,0 +1,679 @@
+<!DOCTYPE html>
+<html lang="zh-CN">
+<head>
+    <meta charset="UTF-8">
+    <meta name="viewport" content="width=device-width, initial-scale=1.0">
+    <title>智能问答测试</title>
+    <style>
+        body {
+            font-family: 'Microsoft YaHei', Arial, sans-serif;
+            margin: 0;
+            padding: 20px;
+            background-color: #f5f5f5;
+            line-height: 1.6;
+        }
+        .container {
+            max-width: 1200px;
+            margin: 0 auto;
+            background: white;
+            border-radius: 8px;
+            box-shadow: 0 2px 10px rgba(0,0,0,0.1);
+            overflow: hidden;
+        }
+        .header {
+            background: linear-gradient(135deg, #667eea 0%, #764ba2 100%);
+            color: white;
+            padding: 20px;
+            text-align: center;
+        }
+        .header h1 {
+            margin: 0;
+            font-size: 28px;
+        }
+        .header p {
+            margin: 10px 0 0 0;
+            opacity: 0.9;
+        }
+        .content {
+            padding: 30px;
+        }
+        .section {
+            margin-bottom: 30px;
+            padding: 20px;
+            border: 1px solid #e0e0e0;
+            border-radius: 8px;
+            background: #fafafa;
+        }
+        .section h3 {
+            margin-top: 0;
+            color: #333;
+            border-bottom: 2px solid #667eea;
+            padding-bottom: 10px;
+        }
+        .form-group {
+            margin-bottom: 20px;
+        }
+        .form-group label {
+            display: block;
+            margin-bottom: 5px;
+            font-weight: 600;
+            color: #333;
+        }
+        .form-group input, .form-group select, .form-group textarea {
+            width: 100%;
+            padding: 12px;
+            border: 1px solid #ddd;
+            border-radius: 6px;
+            font-size: 14px;
+            box-sizing: border-box;
+        }
+        .form-group textarea {
+            resize: vertical;
+            min-height: 80px;
+        }
+        .file-list {
+            max-height: 200px;
+            overflow-y: auto;
+            border: 1px solid #ddd;
+            border-radius: 4px;
+            padding: 10px;
+            background: white;
+        }
+        .file-item {
+            display: flex;
+            justify-content: space-between;
+            align-items: center;
+            padding: 8px;
+            border-bottom: 1px solid #eee;
+        }
+        .file-item:last-child {
+            border-bottom: none;
+        }
+        .file-info {
+            flex: 1;
+        }
+        .file-name {
+            font-weight: 500;
+            color: #333;
+        }
+        .file-meta {
+            font-size: 12px;
+            color: #666;
+        }
+        .btn {
+            padding: 10px 20px;
+            border: none;
+            border-radius: 6px;
+            cursor: pointer;
+            font-size: 14px;
+            font-weight: 500;
+            transition: all 0.3s ease;
+            display: inline-flex;
+            align-items: center;
+            gap: 8px;
+        }
+        .btn-primary {
+            background: #667eea;
+            color: white;
+        }
+        .btn-primary:hover {
+            background: #5a6fd8;
+            transform: translateY(-1px);
+        }
+        .btn-success {
+            background: #28a745;
+            color: white;
+        }
+        .btn-success:hover {
+            background: #218838;
+        }
+        .btn-danger {
+            background: #dc3545;
+            color: white;
+        }
+        .btn-danger:hover {
+            background: #c82333;
+        }
+        .btn-secondary {
+            background: #6c757d;
+            color: white;
+        }
+        .btn-secondary:hover {
+            background: #5a6268;
+        }
+        .btn-group {
+            display: flex;
+            gap: 10px;
+            margin-top: 10px;
+        }
+        .response-area {
+            margin-top: 20px;
+            padding: 20px;
+            border: 1px solid #ddd;
+            border-radius: 6px;
+            background: #f8f9fa;
+            min-height: 200px;
+        }
+        .response-header {
+            display: flex;
+            justify-content: space-between;
+            align-items: center;
+            margin-bottom: 15px;
+        }
+        .response-title {
+            font-weight: 600;
+            color: #333;
+            margin: 0;
+        }
+        .response-time {
+            font-size: 12px;
+            color: #666;
+        }
+        .response-content {
+            line-height: 1.6;
+            color: #333;
+            white-space: pre-wrap;
+        }
+        .loading {
+            text-align: center;
+            padding: 40px;
+            color: #666;
+        }
+        .loading::after {
+            content: '';
+            display: inline-block;
+            width: 20px;
+            height: 20px;
+            border: 2px solid #f3f3f3;
+            border-top: 2px solid #667eea;
+            border-radius: 50%;
+            animation: spin 1s linear infinite;
+            margin-left: 10px;
+        }
+        @keyframes spin {
+            0% { transform: rotate(0deg); }
+            100% { transform: rotate(360deg); }
+        }
+        .error-message {
+            background: #f8d7da;
+            color: #721c24;
+            padding: 15px;
+            border-radius: 6px;
+            border-left: 4px solid #dc3545;
+            margin-bottom: 20px;
+        }
+        .success-message {
+            background: #d4edda;
+            color: #155724;
+            padding: 15px;
+            border-radius: 6px;
+            border-left: 4px solid #28a745;
+            margin-bottom: 20px;
+        }
+        .sources-section {
+            margin-top: 20px;
+            padding-top: 20px;
+            border-top: 1px solid #dee2e6;
+        }
+        .sources-title {
+            font-weight: 600;
+            color: #333;
+            margin-bottom: 10px;
+        }
+        .source-item {
+            background: white;
+            padding: 12px;
+            margin-bottom: 8px;
+            border-radius: 4px;
+            border: 1px solid #e0e0e0;
+        }
+        .source-header {
+            display: flex;
+            justify-content: space-between;
+            margin-bottom: 5px;
+        }
+        .source-file {
+            font-weight: 500;
+            color: #333;
+        }
+        .source-meta {
+            font-size: 12px;
+            color: #666;
+        }
+        .source-snippet {
+            font-size: 13px;
+            color: #555;
+            margin-top: 5px;
+            padding: 8px;
+            background: #f8f9fa;
+            border-radius: 3px;
+        }
+        .tab-container {
+            margin-bottom: 20px;
+        }
+        .tab-buttons {
+            display: flex;
+            border-bottom: 1px solid #dee2e6;
+            margin-bottom: 20px;
+        }
+        .tab-btn {
+            padding: 10px 20px;
+            border: none;
+            background: none;
+            cursor: pointer;
+            font-size: 14px;
+            border-bottom: 2px solid transparent;
+            transition: all 0.3s ease;
+        }
+        .tab-btn.active {
+            color: #667eea;
+            border-bottom-color: #667eea;
+        }
+        .tab-btn:hover {
+            color: #667eea;
+        }
+        .tab-content {
+            display: none;
+        }
+        .tab-content.active {
+            display: block;
+        }
+    </style>
+</head>
+<body>
+    <div class="container">
+        <div class="header">
+            <h1>🤖 智能问答测试</h1>
+            <p>测试基于文档的智能问答功能</p>
+        </div>
+
+        <div class="content">
+            <!-- 消息提示区域 -->
+            <div id="messageContainer"></div>
+
+            <!-- 标签页容器 -->
+            <div class="tab-container">
+                <div class="tab-buttons">
+                    <button class="tab-btn active" onclick="switchTab('test')">测试问答</button>
+                    <button class="tab-btn" onclick="switchTab('manage')">文件管理</button>
+                </div>
+
+                <!-- 测试问答标签页 -->
+                <div id="testTab" class="tab-content active">
+                    <div class="section">
+                        <h3>📋 项目和文件选择</h3>
+                        <div class="form-group">
+                            <label for="projectId">项目ID *</label>
+                            <input type="text" id="projectId" placeholder="输入项目ID，例如：project-001" value="project-001">
+                        </div>
+                        <div class="form-group">
+                            <label for="userId">用户ID *</label>
+                            <input type="text" id="userId" placeholder="输入用户ID，例如：user-001" value="user-001">
+                        </div>
+                        <div class="form-group">
+                            <label for="sessionId">会话ID (可选)</label>
+                            <input type="text" id="sessionId" placeholder="输入会话ID，用于多轮对话">
+                        </div>
+                        <div class="form-group">
+                            <label>选择文件 *</label>
+                            <div id="selectedFiles" class="file-list">
+                                <div class="loading">请先从文件管理标签页选择文件...</div>
+                            </div>
+                        </div>
+                    </div>
+
+                    <div class="section">
+                        <h3>❓ 问题输入</h3>
+                        <div class="form-group">
+                            <label for="query">输入您的问题 *</label>
+                            <textarea id="query" placeholder="例如：这个项目的预算有多少？
+这个供应商的技术方案是否符合要求？
+供应商A和供应商B的报价有什么差异？" required>这个项目的预算有多少？</textarea>
+                        </div>
+                        <div class="btn-group">
+                            <button class="btn btn-primary" onclick="sendQuestion()">
+                                🚀 发送问题
+                            </button>
+                            <button class="btn btn-secondary" onclick="clearResponse()">
+                                🧹 清空结果
+                            </button>
+                        </div>
+                    </div>
+
+                    <div class="section">
+                        <h3>📝 问答结果</h3>
+                        <div id="responseArea" class="response-area">
+                            <div class="loading">等待您提出问题...</div>
+                        </div>
+                    </div>
+                </div>
+
+                <!-- 文件管理标签页 -->
+                <div id="manageTab" class="tab-content">
+                    <div class="section">
+                        <h3>📁 文件管理</h3>
+                        <div class="form-group">
+                            <label for="projectFilter">按项目筛选</label>
+                            <input type="text" id="projectFilter" placeholder="输入项目ID筛选文件" value="project-001">
+                        </div>
+                        <div class="btn-group">
+                            <button class="btn btn-success" onclick="loadFiles()">
+                                🔄 加载文件列表
+                            </button>
+                        </div>
+                        <div class="form-group">
+                            <label>可用文件列表</label>
+                            <div id="availableFiles" class="file-list">
+                                <div class="loading">点击"加载文件列表"查看可用文件...</div>
+                            </div>
+                        </div>
+                    </div>
+                </div>
+            </div>
+        </div>
+    </div>
+
+    <script>
+        // 全局变量
+        let availableFiles = [];
+        let selectedFiles = [];
+
+        // 页面加载完成后的初始化
+        document.addEventListener('DOMContentLoaded', function() {
+            // 自动加载文件列表
+            loadFiles();
+        });
+
+        // 切换标签页
+        function switchTab(tabName) {
+            // 隐藏所有标签页内容
+            document.querySelectorAll('.tab-content').forEach(tab => {
+                tab.classList.remove('active');
+            });
+            document.querySelectorAll('.tab-btn').forEach(btn => {
+                btn.classList.remove('active');
+            });
+
+            // 显示选中的标签页
+            document.getElementById(tabName + 'Tab').classList.add('active');
+            event.target.classList.add('active');
+        }
+
+        // 加载文件列表
+        async function loadFiles() {
+            const projectId = document.getElementById('projectFilter').value || 'project-001';
+            const container = document.getElementById('availableFiles');
+
+            try {
+                container.innerHTML = '<div class="loading">正在加载文件列表...</div>';
+
+                // 调用实际的文件列表API
+                const response = await fetch(`/api/v1/files/project/${projectId}/files`);
+                if (!response.ok) {
+                    throw new Error(`HTTP ${response.status}: ${response.statusText}`);
+                }
+
+                const result = await response.json();
+                if (result.code !== 0) {
+                    throw new Error(result.message || '获取文件列表失败');
+                }
+
+                // 转换API响应为文件列表
+                availableFiles = result.data.files.map(file => ({
+                    fileId: file.fileId,
+                    fileName: file.fileName,
+                    projectId: projectId,
+                    fileSize: file.fileSize,
+                    createTime: file.uploadTime
+                }));
+
+                renderAvailableFiles();
+                showMessage(`文件列表加载成功，共 ${availableFiles.length} 个文件`, 'success');
+
+            } catch (error) {
+                console.error('加载文件列表失败:', error);
+                showMessage('文件列表加载失败，请检查API配置', 'error');
+            }
+        }
+
+        // 渲染可用文件列表
+        function renderAvailableFiles() {
+            const container = document.getElementById('availableFiles');
+
+            if (availableFiles.length === 0) {
+                container.innerHTML = '<div style="padding: 20px; color: #666;">暂无文件数据</div>';
+                return;
+            }
+
+            container.innerHTML = availableFiles.map(file => `
+                <div class="file-item">
+                    <div class="file-info">
+                        <div class="file-name">${escapeHtml(file.fileName)}</div>
+                        <div class="file-meta">
+                            ID: ${file.fileId} |
+                            大小: ${formatFileSize(file.fileSize)} |
+                            时间: ${file.createTime}
+                        </div>
+                    </div>
+                    <button class="btn btn-success btn-sm"
+                            onclick="selectFile('${file.fileId}')">
+                        选择
+                    </button>
+                </div>
+            `).join('');
+
+            updateSelectedFilesDisplay();
+        }
+
+        // 选择文件
+        function selectFile(fileId) {
+            const file = availableFiles.find(f => f.fileId === fileId);
+            if (!file) return;
+
+            // 检查是否已选择
+            if (selectedFiles.find(f => f.fileId === fileId)) {
+                showMessage('文件已选择', 'error');
+                return;
+            }
+
+            selectedFiles.push(file);
+            updateSelectedFilesDisplay();
+            showMessage(`已选择文件: ${file.fileName}`, 'success');
+        }
+
+        // 取消选择文件
+        function deselectFile(fileId) {
+            selectedFiles = selectedFiles.filter(f => f.fileId !== fileId);
+            updateSelectedFilesDisplay();
+            showMessage('文件已取消选择', 'success');
+        }
+
+        // 更新已选文件显示
+        function updateSelectedFilesDisplay() {
+            const container = document.getElementById('selectedFiles');
+
+            if (selectedFiles.length === 0) {
+                container.innerHTML = '<div style="padding: 20px; color: #666;">尚未选择任何文件，请从文件管理标签页选择文件</div>';
+                return;
+            }
+
+            container.innerHTML = selectedFiles.map(file => `
+                <div class="file-item">
+                    <div class="file-info">
+                        <div class="file-name">${escapeHtml(file.fileName)}</div>
+                        <div class="file-meta">
+                            ID: ${file.fileId} | 大小: ${formatFileSize(file.fileSize)}
+                        </div>
+                    </div>
+                    <button class="btn btn-danger btn-sm"
+                            onclick="deselectFile('${file.fileId}')">
+                        移除
+                    </button>
+                </div>
+            `).join('');
+        }
+
+        // 发送问答请求
+        async function sendQuestion() {
+            const projectId = document.getElementById('projectId').value.trim();
+            const userId = document.getElementById('userId').value.trim();
+            const sessionId = document.getElementById('sessionId').value.trim();
+            const query = document.getElementById('query').value.trim();
+
+            // 验证输入
+            if (!projectId) {
+                showMessage('请输入项目ID', 'error');
+                return;
+            }
+            if (!userId) {
+                showMessage('请输入用户ID', 'error');
+                return;
+            }
+            if (!query) {
+                showMessage('请输入问题', 'error');
+                return;
+            }
+            if (selectedFiles.length === 0) {
+                showMessage('请至少选择一个文件', 'error');
+                return;
+            }
+
+            const requestData = {
+                projectId: projectId,
+                userId: userId,
+                query: query,
+                fileIds: selectedFiles.map(f => f.fileId)
+            };
+
+            if (sessionId) {
+                requestData.sessionId = sessionId;
+            }
+
+            const responseArea = document.getElementById('responseArea');
+            responseArea.innerHTML = '<div class="loading">正在思考中，请稍候...</div>';
+
+            try {
+                const response = await fetch('/api/v1/qa/sendMessage', {
+                    method: 'POST',
+                    headers: {
+                        'Content-Type': 'application/json'
+                    },
+                    body: JSON.stringify(requestData)
+                });
+
+                if (!response.ok) {
+                    const errorData = await response.json();
+                    throw new Error(errorData.message || `HTTP ${response.status}: ${response.statusText}`);
+                }
+
+                const result = await response.json();
+
+                // 显示回答结果
+                displayResponse(result.data);
+
+                showMessage('问答请求成功', 'success');
+
+            } catch (error) {
+                console.error('问答请求失败:', error);
+                responseArea.innerHTML = `
+                    <div class="error-message">
+                        <strong>请求失败:</strong> ${error.message}
+                        <br><br>
+                        <strong>建议检查:</strong>
+                        <ul>
+                            <li>确保应用程序正在运行</li>
+                            <li>检查DeepSeek API配置是否正确</li>
+                            <li>确认提示词模板已启用</li>
+                            <li>查看控制台日志了解详细信息</li>
+                        </ul>
+                    </div>
+                `;
+                showMessage('问答请求失败: ' + error.message, 'error');
+            }
+        }
+
+        // 显示回答结果
+        function displayResponse(data) {
+            const responseArea = document.getElementById('responseArea');
+
+            let html = `
+                <div class="response-header">
+                    <h4 class="response-title">🤖 AI回答</h4>
+                    <span class="response-time">${new Date().toLocaleString()}</span>
+                </div>
+                <div class="response-content">${escapeHtml(data.answer)}</div>
+            `;
+
+            // 如果有来源信息，显示来源
+            if (data.sources && data.sources.length > 0) {
+                html += `
+                    <div class="sources-section">
+                        <h5 class="sources-title">📚 信息来源</h5>
+                        ${data.sources.map(source => `
+                            <div class="source-item">
+                                <div class="source-header">
+                                    <span class="source-file">${escapeHtml(source.fileName)}</span>
+                                    <span class="source-meta">
+                                        供应商: ${escapeHtml(source.vendorName || '未知')} |
+                                        相似度: ${(source.similarity * 100).toFixed(1)}%
+                                    </span>
+                                </div>
+                                ${source.snippet ? `<div class="source-snippet">${escapeHtml(source.snippet)}</div>` : ''}
+                            </div>
+                        `).join('')}
+                    </div>
+                `;
+            }
+
+            responseArea.innerHTML = html;
+        }
+
+        // 清空结果
+        function clearResponse() {
+            document.getElementById('responseArea').innerHTML = '<div style="padding: 20px; color: #666;">结果已清空，等待新的问题...</div>';
+        }
+
+        // 显示消息
+        function showMessage(message, type) {
+            const container = document.getElementById('messageContainer');
+            const className = type === 'error' ? 'error-message' : 'success-message';
+            container.innerHTML = `<div class="${className}">${message}</div>`;
+
+            // 3秒后自动隐藏
+            setTimeout(() => {
+                container.innerHTML = '';
+            }, 3000);
+        }
+
+        // 格式化文件大小
+        function formatFileSize(bytes) {
+            if (bytes === 0) return '0 B';
+            const k = 1024;
+            const sizes = ['B', 'KB', 'MB', 'GB'];
+            const i = Math.floor(Math.log(bytes) / Math.log(k));
+            return parseFloat((bytes / Math.pow(k, i)).toFixed(2)) + ' ' + sizes[i];
+        }
+
+        // HTML转义
+        function escapeHtml(text) {
+            const div = document.createElement('div');
+            div.textContent = text || '';
+            return div.innerHTML;
+        }
+
+        // 添加一些预设问题
+        function setPresetQuestion(question) {
+            document.getElementById('query').value = question;
+        }
+
+        // 添加键盘快捷键
+        document.addEventListener('keydown', function(e) {
+            // Ctrl+Enter 发送问题
+            if (e.ctrlKey && e.key === 'Enter') {
+                e.preventDefault();
+                sendQuestion();
+            }
+        });
+    </script>
+</body>
+</html>
diff --git a/src/main/resources/static/view/summary-test.html b/src/main/resources/static/view/summary-test.html
new file mode 100644
index 0000000..655f90a
--- /dev/null
+++ b/src/main/resources/static/view/summary-test.html
@@ -0,0 +1,595 @@
+<!DOCTYPE html>
+<html lang="zh-CN">
+<head>
+    <meta charset="UTF-8">
+    <meta name="viewport" content="width=device-width, initial-scale=1.0">
+    <title>文件目录和章节总结测试</title>
+    <style>
+        body {
+            font-family: 'Microsoft YaHei', Arial, sans-serif;
+            margin: 0;
+            padding: 20px;
+            background-color: #f5f5f5;
+            line-height: 1.6;
+        }
+        .container {
+            max-width: 1400px;
+            margin: 0 auto;
+            background: white;
+            border-radius: 8px;
+            box-shadow: 0 2px 10px rgba(0,0,0,0.1);
+            overflow: hidden;
+        }
+        .header {
+            background: linear-gradient(135deg, #43a047 0%, #2e7d32 100%);
+            color: white;
+            padding: 20px;
+            text-align: center;
+        }
+        .header h1 {
+            margin: 0;
+            font-size: 28px;
+        }
+        .header p {
+            margin: 10px 0 0 0;
+            opacity: 0.9;
+        }
+        .content {
+            padding: 30px;
+        }
+        .section {
+            margin-bottom: 30px;
+            padding: 20px;
+            border: 1px solid #e0e0e0;
+            border-radius: 8px;
+            background: #fafafa;
+        }
+        .section h3 {
+            margin-top: 0;
+            color: #333;
+            border-bottom: 2px solid #43a047;
+            padding-bottom: 10px;
+        }
+        .form-group {
+            margin-bottom: 20px;
+        }
+        .form-group label {
+            display: block;
+            margin-bottom: 5px;
+            font-weight: 600;
+            color: #333;
+        }
+        .form-group input {
+            width: 100%;
+            padding: 12px;
+            border: 1px solid #ddd;
+            border-radius: 6px;
+            font-size: 14px;
+            box-sizing: border-box;
+        }
+        .btn {
+            padding: 10px 20px;
+            border: none;
+            border-radius: 6px;
+            cursor: pointer;
+            font-size: 14px;
+            font-weight: 500;
+            transition: all 0.3s ease;
+            display: inline-flex;
+            align-items: center;
+            gap: 8px;
+        }
+        .btn-primary {
+            background: #43a047;
+            color: white;
+        }
+        .btn-primary:hover {
+            background: #2e7d32;
+            transform: translateY(-1px);
+        }
+        .btn-success {
+            background: #2196f3;
+            color: white;
+        }
+        .btn-success:hover {
+            background: #1976d2;
+        }
+        .btn-secondary {
+            background: #6c757d;
+            color: white;
+        }
+        .btn-secondary:hover {
+            background: #5a6268;
+        }
+        .btn-group {
+            display: flex;
+            gap: 10px;
+            margin-top: 10px;
+        }
+        .directory-tree {
+            max-height: 600px;
+            overflow-y: auto;
+            border: 1px solid #ddd;
+            border-radius: 4px;
+            background: white;
+        }
+        .tree-node {
+            padding: 8px 12px;
+            border-bottom: 1px solid #eee;
+            cursor: pointer;
+            transition: background-color 0.3s ease;
+        }
+        .tree-node:hover {
+            background-color: #f8f9fa;
+        }
+        .tree-node:last-child {
+            border-bottom: none;
+        }
+        .tree-node.selected {
+            background-color: #e3f2fd;
+            border-left: 3px solid #2196f3;
+        }
+        .tree-node-content {
+            display: flex;
+            justify-content: space-between;
+            align-items: center;
+        }
+        .node-info {
+            flex: 1;
+        }
+        .node-title {
+            font-weight: 500;
+            color: #333;
+            margin-bottom: 2px;
+        }
+        .node-meta {
+            font-size: 12px;
+            color: #666;
+        }
+        .tree-indent {
+            padding-left: 20px;
+        }
+        .tree-toggle {
+            margin-right: 8px;
+            cursor: pointer;
+            color: #666;
+            font-size: 12px;
+        }
+        .summary-panel {
+            margin-top: 20px;
+            padding: 20px;
+            border: 1px solid #ddd;
+            border-radius: 6px;
+            background: #f8f9fa;
+            min-height: 200px;
+        }
+        .summary-header {
+            display: flex;
+            justify-content: space-between;
+            align-items: center;
+            margin-bottom: 15px;
+        }
+        .summary-title {
+            font-size: 18px;
+            font-weight: 600;
+            color: #333;
+            margin: 0;
+        }
+        .summary-meta {
+            font-size: 12px;
+            color: #666;
+        }
+        .summary-content {
+            line-height: 1.6;
+            color: #333;
+            white-space: pre-wrap;
+        }
+        .loading {
+            text-align: center;
+            padding: 40px;
+            color: #666;
+        }
+        .loading::after {
+            content: '';
+            display: inline-block;
+            width: 20px;
+            height: 20px;
+            border: 2px solid #f3f3f3;
+            border-top: 2px solid #43a047;
+            border-radius: 50%;
+            animation: spin 1s linear infinite;
+            margin-left: 10px;
+        }
+        @keyframes spin {
+            0% { transform: rotate(0deg); }
+            100% { transform: rotate(360deg); }
+        }
+        .error-message {
+            background: #f8d7da;
+            color: #721c24;
+            padding: 15px;
+            border-radius: 6px;
+            border-left: 4px solid #dc3545;
+            margin-bottom: 20px;
+        }
+        .success-message {
+            background: #d4edda;
+            color: #155724;
+            padding: 15px;
+            border-radius: 6px;
+            border-left: 4px solid #28a745;
+            margin-bottom: 20px;
+        }
+        .empty-state {
+            text-align: center;
+            padding: 60px 20px;
+            color: #666;
+        }
+        .empty-state h4 {
+            margin: 0 0 10px 0;
+            color: #333;
+        }
+    </style>
+</head>
+<body>
+    <div class="container">
+        <div class="header">
+            <h1>📚 文件目录和章节总结测试</h1>
+            <p>测试文件目录抽取和章节总结功能</p>
+        </div>
+
+        <div class="content">
+            <!-- 消息提示区域 -->
+            <div id="messageContainer"></div>
+
+            <!-- 查询条件 -->
+            <div class="section">
+                <h3>🔍 查询条件</h3>
+                <div class="form-group">
+                    <label for="fileId">文件ID *</label>
+                    <input type="text" id="fileId" placeholder="输入文件ID，例如：9e607a37-ab87-4e02-8b42-125270d6bf58" value="9e607a37-ab87-4e02-8b42-125270d6bf58">
+                </div>
+                <div class="btn-group">
+                    <button class="btn btn-primary" onclick="loadDirectory()">
+                        📂 加载目录结构
+                    </button>
+                    <button class="btn btn-success" onclick="checkFileStatus()">
+                        🔍 检查文件状态
+                    </button>
+                    <button class="btn btn-warning" onclick="processDirectory()">
+                        ⚡ 手动触发抽取
+                    </button>
+                    <button class="btn btn-secondary" onclick="clearAll()">
+                        🧹 清空显示
+                    </button>
+                </div>
+            </div>
+
+            <!-- 目录树和总结面板 -->
+            <div class="section">
+                <div style="display: grid; grid-template-columns: 1fr 1fr; gap: 20px;">
+                    <!-- 目录树 -->
+                    <div>
+                        <h3>📋 目录结构</h3>
+                        <div id="directoryContainer" class="directory-tree">
+                            <div class="empty-state">
+                                <h4>请先输入文件ID</h4>
+                                <p>输入文件ID并点击"加载目录结构"查看文件目录</p>
+                            </div>
+                        </div>
+                    </div>
+
+                    <!-- 章节总结 -->
+                    <div>
+                        <h3>📝 章节总结</h3>
+                        <div id="summaryContainer" class="summary-panel">
+                            <div class="empty-state">
+                                <h4>请选择章节</h4>
+                                <p>点击左侧目录中的任意章节，查看对应的总结内容</p>
+                            </div>
+                        </div>
+                    </div>
+                </div>
+            </div>
+        </div>
+    </div>
+
+    <script>
+        // 全局变量
+        let directoryData = [];
+        let selectedChapterId = null;
+
+        // 页面加载完成后的初始化
+        document.addEventListener('DOMContentLoaded', function() {
+            // 可以在这里添加初始化逻辑
+        });
+
+        // 加载目录结构
+        async function loadDirectory() {
+            const fileId = document.getElementById('fileId').value.trim();
+
+            if (!fileId) {
+                showMessage('请输入文件ID', 'error');
+                return;
+            }
+
+            const container = document.getElementById('directoryContainer');
+
+            try {
+                container.innerHTML = '<div class="loading">正在加载目录结构...</div>';
+
+                const response = await fetch(`/api/v1/summary/directory?file_id=${encodeURIComponent(fileId)}`);
+                if (!response.ok) {
+                    throw new Error(`HTTP ${response.status}: ${response.statusText}`);
+                }
+
+                const result = await response.json();
+                if (result.code !== 0) {
+                    throw new Error(result.message || '获取目录结构失败');
+                }
+
+                directoryData = result.data.directory;
+                renderDirectoryTree();
+
+                showMessage(`目录加载成功，共 ${countTotalChapters(directoryData)} 个章节`, 'success');
+
+            } catch (error) {
+                console.error('加载目录结构失败:', error);
+                container.innerHTML = '<div class="error-message">加载失败，请检查文件ID是否正确</div>';
+                showMessage('加载目录结构失败: ' + error.message, 'error');
+            }
+        }
+
+        // 渲染目录树
+        function renderDirectoryTree() {
+            const container = document.getElementById('directoryContainer');
+
+            if (directoryData.length === 0) {
+                container.innerHTML = '<div class="empty-state"><h4>暂无目录数据</h4><p>该文件可能还没有进行目录抽取</p></div>';
+                return;
+            }
+
+            container.innerHTML = '';
+            renderTreeNodes(directoryData, container, 0);
+        }
+
+        // 递归渲染树节点
+        function renderTreeNodes(nodes, container, level) {
+            nodes.forEach(node => {
+                const nodeElement = document.createElement('div');
+                nodeElement.className = `tree-node ${selectedChapterId === node.chapterId ? 'selected' : ''}`;
+                nodeElement.onclick = () => selectChapter(node.chapterId);
+
+                const indent = '  '.repeat(level);
+                nodeElement.innerHTML = `
+                    <div class="tree-node-content">
+                        <div class="node-info">
+                            <div class="node-title">${indent}${node.title}</div>
+                            <div class="node-meta">章节: ${node.chapterId} | 页码: ${node.page || 'N/A'}</div>
+                        </div>
+                        ${node.children && node.children.length > 0 ? '<span class="tree-toggle">▶</span>' : ''}
+                    </div>
+                `;
+
+                container.appendChild(nodeElement);
+
+                // 递归渲染子节点
+                if (node.children && node.children.length > 0) {
+                    renderTreeNodes(node.children, container, level + 1);
+                }
+            });
+        }
+
+        // 选择章节
+        function selectChapter(chapterId) {
+            selectedChapterId = chapterId;
+
+            // 更新选中状态
+            document.querySelectorAll('.tree-node').forEach(node => {
+                node.classList.remove('selected');
+            });
+
+            // 找到对应的节点并添加选中状态
+            const selectedNodes = document.querySelectorAll('.tree-node');
+            selectedNodes.forEach(node => {
+                if (node.onclick && node.onclick.toString().includes(chapterId)) {
+                    node.classList.add('selected');
+                }
+            });
+
+            // 加载章节总结
+            loadChapterSummary(chapterId);
+        }
+
+        // 加载章节总结
+        async function loadChapterSummary(chapterId) {
+            const fileId = document.getElementById('fileId').value.trim();
+            const container = document.getElementById('summaryContainer');
+
+            try {
+                container.innerHTML = '<div class="loading">正在加载章节总结...</div>';
+
+                const response = await fetch(`/api/v1/summary/chapter?file_id=${encodeURIComponent(fileId)}&chapter_id=${encodeURIComponent(chapterId)}`);
+                if (!response.ok) {
+                    if (response.status === 404) {
+                        container.innerHTML = '<div class="empty-state"><h4>章节总结不存在</h4><p>该章节可能还没有生成总结内容</p></div>';
+                        return;
+                    }
+                    throw new Error(`HTTP ${response.status}: ${response.statusText}`);
+                }
+
+                const result = await response.json();
+                if (result.code !== 0) {
+                    throw new Error(result.message || '获取章节总结失败');
+                }
+
+                renderChapterSummary(result.data);
+
+            } catch (error) {
+                console.error('加载章节总结失败:', error);
+                container.innerHTML = `<div class="error-message"><strong>加载失败:</strong> ${error.message}</div>`;
+            }
+        }
+
+        // 渲染章节总结
+        function renderChapterSummary(data) {
+            const container = document.getElementById('summaryContainer');
+
+            container.innerHTML = `
+                <div class="summary-header">
+                    <h4 class="summary-title">章节总结</h4>
+                    <div class="summary-meta">文件ID: ${data.fileId} | 章节ID: ${data.chapterId}</div>
+                </div>
+                <div class="summary-content">${escapeHtml(data.summary || '暂无总结内容')}</div>
+                ${data.sources && data.sources.length > 0 ? `
+                    <div style="margin-top: 15px; padding-top: 15px; border-top: 1px solid #ddd;">
+                        <strong>来源信息:</strong> 共 ${data.sources.length} 个来源
+                    </div>
+                ` : ''}
+            `;
+        }
+
+        // 清空所有显示
+        function clearAll() {
+            document.getElementById('directoryContainer').innerHTML = '<div style="padding: 20px; color: #666;">请点击"加载目录结构"查看文件目录</div>';
+            document.getElementById('summaryContainer').innerHTML = '<div class="empty-state"><h4>请选择章节</h4><p>点击左侧目录中的任意章节，查看对应的总结内容</p></div>';
+            directoryData = [];
+            selectedChapterId = null;
+        }
+
+        // 显示消息
+        function showMessage(message, type) {
+            const container = document.getElementById('messageContainer');
+            const className = type === 'error' ? 'error-message' : 'success-message';
+            container.innerHTML = `<div class="${className}">${message}</div>`;
+
+            // 3秒后自动隐藏
+            setTimeout(() => {
+                container.innerHTML = '';
+            }, 3000);
+        }
+
+        // 统计总章节数
+        function countTotalChapters(nodes) {
+            let count = nodes.length;
+            nodes.forEach(node => {
+                if (node.children && node.children.length > 0) {
+                    count += countTotalChapters(node.children);
+                }
+            });
+            return count;
+        }
+
+        // 检查文件状态
+        async function checkFileStatus() {
+            const fileId = document.getElementById('fileId').value.trim();
+
+            if (!fileId) {
+                showMessage('请输入文件ID', 'error');
+                return;
+            }
+
+            try {
+                const response = await fetch(`/api/v1/summary/directory/status?file_id=${encodeURIComponent(fileId)}`);
+                if (!response.ok) {
+                    throw new Error(`HTTP ${response.status}: ${response.statusText}`);
+                }
+
+                const result = await response.json();
+                if (result.code !== 0) {
+                    throw new Error(result.message || '检查文件状态失败');
+                }
+
+                // 显示文件状态信息
+                let statusMessage = `文件状态检查结果:\n`;
+                statusMessage += `文件ID: ${result.fileId}\n`;
+                statusMessage += `文件存在: ${result.fileExists ? '是' : '否'}\n`;
+                statusMessage += `目录数据存在: ${result.directoryExists ? '是' : '否'}\n`;
+                statusMessage += `目录章节数量: ${result.directoryCount}\n`;
+
+                if (result.fileExists) {
+                    statusMessage += `\n文件信息:\n`;
+                    statusMessage += `文件名: ${result.fileName || '未知'}\n`;
+                    statusMessage += `文件大小: ${result.fileSize ? formatFileSize(result.fileSize) : '未知'}\n`;
+                    statusMessage += `上传时间: ${result.uploadTime ? new Date(result.uploadTime).toLocaleString('zh-CN') : '未知'}\n`;
+                    statusMessage += `向量化状态: ${result.vectorizationStatus || '未知'}\n`;
+                }
+
+                if (!result.fileExists) {
+                    statusMessage += `\n建议:\n`;
+                    statusMessage += `1. 检查文件ID是否正确\n`;
+                    statusMessage += `2. 确认文件是否已上传到系统中\n`;
+                } else if (!result.directoryExists) {
+                    statusMessage += `\n建议:\n`;
+                    statusMessage += `1. 文件存在但没有目录数据\n`;
+                    statusMessage += `2. 可以尝试手动触发目录抽取\n`;
+                    statusMessage += `3. 或等待异步处理完成\n`;
+                }
+
+                alert(statusMessage);
+                showMessage('文件状态检查完成', 'success');
+
+            } catch (error) {
+                console.error('检查文件状态失败:', error);
+                showMessage('检查文件状态失败: ' + error.message, 'error');
+            }
+        }
+
+        // 手动触发目录抽取
+        async function processDirectory() {
+            const fileId = document.getElementById('fileId').value.trim();
+
+            if (!fileId) {
+                showMessage('请输入文件ID', 'error');
+                return;
+            }
+
+            if (!confirm(`确定要为文件 ${fileId} 手动触发目录抽取吗？这可能需要一些时间。`)) {
+                return;
+            }
+
+            try {
+                const response = await fetch(`/api/v1/summary/directory/process?file_id=${encodeURIComponent(fileId)}`, {
+                    method: 'POST'
+                });
+
+                if (!response.ok) {
+                    throw new Error(`HTTP ${response.status}: ${response.statusText}`);
+                }
+
+                const result = await response.json();
+                if (result.code !== 0) {
+                    throw new Error(result.message || '手动触发抽取失败');
+                }
+
+                showMessage(`目录抽取任务已启动: ${result.fileName}，请等待几分钟后重新查询`, 'success');
+
+                // 3分钟后自动重新加载目录
+                setTimeout(() => {
+                    loadDirectory();
+                }, 180000); // 3分钟
+
+            } catch (error) {
+                console.error('手动触发抽取失败:', error);
+                showMessage('手动触发抽取失败: ' + error.message, 'error');
+            }
+        }
+
+        // 格式化文件大小
+        function formatFileSize(bytes) {
+            if (!bytes) return '0 B';
+            const k = 1024;
+            const sizes = ['B', 'KB', 'MB', 'GB'];
+            const i = Math.floor(Math.log(bytes) / Math.log(k));
+            return parseFloat((bytes / Math.pow(k, i)).toFixed(2)) + ' ' + sizes[i];
+        }
+
+        // HTML转义
+        function escapeHtml(text) {
+            const div = document.createElement('div');
+            div.textContent = text || '';
+            return div.innerHTML;
+        }
+    </script>
+</body>
+</html>
diff --git a/src/main/resources/test.pdf b/src/main/resources/test.pdf
new file mode 100644
index 0000000..e69de29
diff --git a/src/test/java/com/zhpb/gdyd_zhpb_zgf/GdydZhpbZgfApplicationTests.java b/src/test/java/com/zhpb/gdyd_zhpb_zgf/GdydZhpbZgfApplicationTests.java
new file mode 100644
index 0000000..58bc9dc
--- /dev/null
+++ b/src/test/java/com/zhpb/gdyd_zhpb_zgf/GdydZhpbZgfApplicationTests.java
@@ -0,0 +1,23 @@
+/**
+ * @ProjectName: 智慧评标-主观分助手
+ * @Package: com.zhpb.gdyd_zhpb_zgf
+ * @ClassName: GdydZhpbZgfApplicationTests
+ * @Description: Spring Boot应用测试类
+ * @Author: 张志锋
+ * @Date: 2025-10-31
+ * @Version: 1.0
+ * @Copyright: © 2025 智慧评标-主观分助手
+ */
+package com.zhpb.gdyd_zhpb_zgf;
+
+import org.junit.jupiter.api.Test;
+import org.springframework.boot.test.context.SpringBootTest;
+
+@SpringBootTest
+class GdydZhpbZgfApplicationTests {
+
+    @Test
+    void contextLoads() {
+    }
+
+}
diff --git a/test.txt b/test.txt
new file mode 100644
index 0000000..954efb9
Binary files /dev/null and b/test.txt differ
diff --git a/test_api_upload.py b/test_api_upload.py
new file mode 100644
index 0000000..6057ad5
--- /dev/null
+++ b/test_api_upload.py
@@ -0,0 +1,114 @@
+import time
+import json
+import base64
+import hashlib
+import uuid
+import requests
+
+# 配置信息
+API_URL = "http://149.88.90.45:8000/api/v1/files/upload"
+# API_URL = "http://127.0.0.1:8000/api/v1/files/upload"
+
+# 注意：这里的 APP_ID 需要与后端校验逻辑匹配（如果后端开启了严格校验）
+APP_ID = "your-app-id"
+# APP_KEY 需要与后端配置文件中的 private-model.appkey 一致
+APP_KEY = "test-app-key" 
+
+# 测试用的文件 URL (这里使用用户提供的示例URL，或者替换为可访问的公共URL)
+# 注意：如果后端无法访问此内网地址，下载会失败。
+# 北京协成致远网络科技有限公司_标段1_标包1_文件.zip
+FILE_URL = "http://45.192.102.202/%E5%8C%97%E4%BA%AC%E5%8D%8F%E6%88%90%E8%87%B4%E8%BF%9C%E7%BD%91%E7%BB%9C%E7%A7%91%E6%8A%80%E6%9C%89%E9%99%90%E5%85%AC%E5%8F%B8_%E6%A0%87%E6%AE%B51_%E6%A0%87%E5%8C%851_%E6%96%87%E4%BB%B6.zip"
+# 北京直真科技股份有限公司_标段1_标包1_文件.zip
+# FILE_URL = "http://45.192.102.202/%E5%8C%97%E4%BA%AC%E7%9B%B4%E7%9C%9F%E7%A7%91%E6%8A%80%E8%82%A1%E4%BB%BD%E6%9C%89%E9%99%90%E5%85%AC%E5%8F%B8_%E6%A0%87%E6%AE%B51_%E6%A0%87%E5%8C%851_%E6%96%87%E4%BB%B6.zip"
+# 博瑞得科技有限公司_标段1_标包1_文件.zip
+# FILE_URL = "http://45.192.102.202/%E5%8D%9A%E7%91%9E%E5%BE%97%E7%A7%91%E6%8A%80%E6%9C%89%E9%99%90%E5%85%AC%E5%8F%B8_%E6%A0%87%E6%AE%B51_%E6%A0%87%E5%8C%851_%E6%96%87%E4%BB%B6.zip"
+JSON_FILE_URL = "http://45.192.102.202/%E9%87%87%E8%B4%AD%E6%96%87%E4%BB%B6json20250717161005.txt"
+
+def generate_headers():
+    """生成符合接口规范的请求头"""
+    cur_time = str(int(time.time()))
+    trace_id = str(uuid.uuid4())
+    
+    # 1. 构造 X-Server-Param (Base64编码的JSON)
+    param_dict = {
+        "appid": APP_ID,
+        "csid": trace_id
+    }
+    param_json = json.dumps(param_dict)
+    x_server_param = base64.b64encode(param_json.encode('utf-8')).decode('utf-8')
+    
+    # 2. 构造 X-CheckSum (MD5签名)
+    # 公式: MD5(appKey + curTime + serverParam)
+    check_sum_str = APP_KEY + cur_time + x_server_param
+    x_check_sum = hashlib.md5(check_sum_str.encode('utf-8')).hexdigest()
+    
+    headers = {
+        "Content-Type": "application/json",
+        "X-CurTime": cur_time,
+        "X-Server-Param": x_server_param,
+        "X-CheckSum": x_check_sum,
+        "traceId": trace_id
+    }
+    return headers
+
+def test_upload(file_type="1"):
+    print(f"=== 开始测试接口: {API_URL} [fileType={file_type}] ===")
+    
+    headers = generate_headers()
+    
+    # 构造请求体 (完全按照用户提供的JSON结构)
+    payload = {
+        "fileIds": [FILE_URL],
+        "fileType": file_type,  # 1商务、2技术
+        "projectId": "project001",
+        "packageId": "project001",
+        "supplierId": "supplier001",
+        "jsonFileId": JSON_FILE_URL,
+        "unitId": "unit001"
+    }
+    
+    print("\n[请求头]:")
+    print(json.dumps(headers, indent=2))
+    
+    print("\n[请求体]:")
+    print(json.dumps(payload, indent=2, ensure_ascii=False))
+    
+    try:
+        print("\n正在发送请求...")
+        response = requests.post(API_URL, headers=headers, json=payload)
+        
+        print(f"\n[响应状态码]: {response.status_code}")
+        print("[响应内容]:")
+        try:
+            print(json.dumps(response.json(), indent=2, ensure_ascii=False))
+        except:
+            print(response.text)
+            
+    except requests.exceptions.ConnectionError:
+        print(f"\n❌ 连接失败: 无法连接到 {API_URL}")
+        print("请确保后端服务已启动并在端口 8000 上监听")
+    except Exception as e:
+        print(f"\n❌ 发生错误: {e}")
+
+if __name__ == "__main__":
+    # 检查是否安装了 requests 库
+    try:
+        import requests
+        
+        # 测试场景1: 商务文件 (fileType="1") - 应该直接返回成功，不下载
+        print("\n>>> 测试场景1: 商务文件 (fileType='1')")
+        test_upload(file_type="1")
+        
+        print("\n" + "="*50 + "\n")
+        
+        # 测试场景2: 技术文件 (fileType="2") - 应该触发下载和处理
+        # 注意：如果后端无法访问内网URL，这里可能会在日志中报错，但接口会立即返回成功（因为是异步处理）
+        print(">>> 测试场景2: 技术文件 (fileType='2')")
+        test_upload(file_type="2")
+        
+    except ImportError:
+        print("错误: 需要安装 'requests' 库才能运行此脚本。")
+        print("请运行: pip install requests")
+
+
+# http://149.88.90.45/#/home?projectName=0717%E5%9B%9E%E5%BD%92%E6%B5%8B%E8%AF%95-%E6%9D%A8%E6%B1%B6%E5%AE%87&projectId=project001
\ No newline at end of file
diff --git a/test_catalog_extract.py b/test_catalog_extract.py
new file mode 100644
index 0000000..8df0291
--- /dev/null
+++ b/test_catalog_extract.py
@@ -0,0 +1,336 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+"""
+PDF目录摘取测试脚本 (与Java代码逻辑一致)
+
+模拟 FileServiceImpl.extractDirectoryFromPDF 的逻辑，用于测试验证。
+
+使用方法:
+    python test_catalog_extract.py <pdf文件路径>
+
+依赖安装:
+    pip install pypdf
+"""
+
+import sys
+import json
+import os
+from datetime import datetime
+
+# 模拟Java的日志输出
+def log_info(msg):
+    print(f"[{datetime.now().strftime('%H:%M:%S')}] INFO  [目录摘取] {msg}")
+
+def log_debug(msg):
+    print(f"[{datetime.now().strftime('%H:%M:%S')}] DEBUG [目录摘取] {msg}")
+
+def log_warn(msg):
+    print(f"[{datetime.now().strftime('%H:%M:%S')}] WARN  [目录摘取] {msg}")
+
+def log_error(msg):
+    print(f"[{datetime.now().strftime('%H:%M:%S')}] ERROR [目录摘取] {msg}")
+
+
+def get_page_number_from_outline_item(reader, item):
+    """
+    对应Java: getPageNumberFromOutlineItem(PDOutlineItem item)
+    尝试从目录项获取页码
+    """
+    try:
+        if hasattr(item, 'page') and item.page:
+            # 直接有page属性
+            page_num = reader.get_destination_page_number(item)
+            return page_num + 1  # 转为1-based
+    except:
+        pass
+    return -1
+
+
+def extract_outline_items(reader, outline_items, chapters, level=1, indent=""):
+    """
+    对应Java: extractOutlineItems(PDOutlineItem item, String indent, List<JSONObject> chapters, int level)
+    递归提取PDF目录项
+    """
+    if not outline_items:
+        return
+    
+    for i, item in enumerate(outline_items):
+        try:
+            # 如果是列表（子项），递归处理
+            if isinstance(item, list):
+                extract_outline_items(reader, item, chapters, level + 1, indent + "  ")
+                continue
+            
+            # 获取标题
+            title = item.title if hasattr(item, 'title') else None
+            
+            if title and title.strip():
+                title = title.strip()
+                
+                chapter = {
+                    "chapterId": f"chap-{len(chapters)}",
+                    "title": title,
+                    "level": level,
+                    "page": "1",
+                    "startPage": 1,
+                    "endPage": 1,
+                    "children": []
+                }
+                
+                # 尝试获取页码信息
+                start_page = -1
+                try:
+                    page_num = reader.get_destination_page_number(item)
+                    if page_num is not None and page_num >= 0:
+                        start_page = page_num + 1  # PyPDF页码从0开始，转为1-based
+                except Exception as e:
+                    log_debug(f"无法获取章节页码信息: {title} - {e}")
+                
+                # 如果获取到了起始页码
+                if start_page != -1:
+                    chapter["page"] = str(start_page)
+                    chapter["startPage"] = start_page
+                    chapter["endPage"] = start_page  # 暂时设为起始页码
+                else:
+                    chapter["page"] = "1"
+                    chapter["startPage"] = 1
+                    chapter["endPage"] = 1
+                
+                chapters.append(chapter)
+                log_debug(f"{indent}• {title} (第{chapter['startPage']}页)")
+                
+        except Exception as e:
+            log_warn(f"处理目录项失败: {e}")
+
+
+def build_hierarchical_structure(chapters, max_depth=2):
+    """
+    对应Java: buildHierarchicalStructure(List<JSONObject> chapters)
+    构建层级目录结构
+    """
+    root_array = []
+    if not chapters:
+        return root_array
+    
+    level_parents = []  # 存储每个层级的当前父节点
+    
+    for chapter in chapters:
+        level = max(0, chapter.get("level", 1))
+        if chapter.get("children") is None:
+            chapter["children"] = []
+        
+        if level == 0 or not level_parents or level_parents[0] is None:
+            root_array.append(chapter)
+            level_parents.clear()
+            level_parents.append(chapter)
+            continue
+        
+        parent_index = min(level - 1, len(level_parents) - 1)
+        parent = level_parents[parent_index] if parent_index >= 0 else None
+        
+        if parent is None:
+            # 找不到父节点时降级为根节点
+            root_array.append(chapter)
+            level_parents.clear()
+            level_parents.append(chapter)
+            continue
+        
+        parent_children = parent.get("children", [])
+        parent_children.append(chapter)
+        parent["children"] = parent_children
+        
+        # 更新level_parents
+        while len(level_parents) < level:
+            level_parents.append(None)
+        if level >= len(level_parents):
+            level_parents.append(chapter)
+        else:
+            level_parents[level] = chapter
+        
+        # 清空更深层级的父节点
+        for idx in range(level + 1, len(level_parents)):
+            level_parents[idx] = None
+    
+    # 限制目录深度
+    limit_directory_depth(root_array, max_depth)
+    
+    return root_array
+
+
+def limit_directory_depth(chapters, max_depth, current_depth=1):
+    """限制目录深度"""
+    if current_depth >= max_depth:
+        for chapter in chapters:
+            chapter["children"] = []
+    else:
+        for chapter in chapters:
+            if chapter.get("children"):
+                limit_directory_depth(chapter["children"], max_depth, current_depth + 1)
+
+
+def extract_directory_from_pdf(file_path, file_id="test-file", file_name=None):
+    """
+    对应Java: extractDirectoryFromPDF(String filePath, String fileId, String fileName)
+    从PDF中提取目录结构
+    """
+    try:
+        from pypdf import PdfReader
+    except ImportError:
+        try:
+            from PyPDF2 import PdfReader
+        except ImportError:
+            log_error("请先安装 pypdf: pip install pypdf")
+            return None
+    
+    if file_name is None:
+        file_name = os.path.basename(file_path)
+    
+    log_info(f"开始处理PDF文件: {file_name}")
+    
+    try:
+        total_pages = 1
+        chapters = []
+        
+        # 加载PDF文档
+        reader = PdfReader(file_path)
+        total_pages = len(reader.pages)
+        log_info(f"PDF总页数: {total_pages}")
+        
+        # 尝试从PDF大纲中提取目录
+        outline = reader.outline
+        
+        if outline:
+            log_info("发现PDF大纲，开始提取...")
+            extract_outline_items(reader, outline, chapters, level=1, indent="")
+            log_info(f"从大纲提取到 {len(chapters)} 个章节")
+        
+        if chapters:
+            # 构建层级结构
+            hierarchical_array = build_hierarchical_structure(chapters)
+            log_info(f"成功构建层级目录结构，根节点数: {len(hierarchical_array)}")
+            return {
+                "source": "pdf_outline",
+                "total_pages": total_pages,
+                "flat_chapters": chapters,
+                "hierarchical_chapters": hierarchical_array
+            }
+        
+        # 回退方案：使用文件名作为单个章节
+        title = file_name
+        if title and title.lower().endswith(".pdf"):
+            title = title[:-4]
+        if not title or not title.strip():
+            title = "文档内容"
+        
+        directory_array = [{
+            "chapterId": "chap-0",
+            "title": title,
+            "level": 1,
+            "page": "1",
+            "startPage": 1,
+            "endPage": total_pages,
+            "children": []
+        }]
+        
+        log_info(f"使用文件名作为默认章节: '{title}', 页数范围: 1-{total_pages}")
+        
+        return {
+            "source": "filename_fallback",
+            "total_pages": total_pages,
+            "flat_chapters": directory_array,
+            "hierarchical_chapters": directory_array
+        }
+        
+    except Exception as e:
+        log_error(f"处理PDF失败: {e}")
+        import traceback
+        traceback.print_exc()
+        return None
+
+
+def print_catalog_tree(chapters, indent=0):
+    """打印目录树形结构"""
+    for chapter in chapters:
+        prefix = "  " * indent + ("├─ " if indent > 0 else "")
+        title = chapter.get("title", "未命名")
+        page = chapter.get("startPage", "?")
+        level = chapter.get("level", 1)
+        print(f"{prefix}[L{level}] {title} (第{page}页)")
+        
+        children = chapter.get("children", [])
+        if children:
+            print_catalog_tree(children, indent + 1)
+
+
+def main():
+    print("=" * 70)
+    print("PDF目录摘取测试脚本 (与Java extractDirectoryFromPDF 逻辑一致)")
+    print("=" * 70)
+    
+    if len(sys.argv) < 2:
+        print("\n用法:")
+        print("  python test_catalog_extract.py <pdf文件路径>")
+        print("\n示例:")
+        print("  python test_catalog_extract.py ./document.pdf")
+        print("  python test_catalog_extract.py /path/to/技术方案.pdf")
+        print("\n依赖:")
+        print("  pip install pypdf")
+        print("=" * 70)
+        sys.exit(1)
+    
+    pdf_path = sys.argv[1]
+    
+    # 检查文件是否存在
+    if not os.path.exists(pdf_path):
+        log_error(f"文件不存在: {pdf_path}")
+        sys.exit(1)
+    
+    print(f"\n📄 PDF文件: {pdf_path}")
+    print(f"⏰ 开始时间: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}")
+    print("-" * 70)
+    
+    start_time = datetime.now()
+    
+    # 提取目录
+    result = extract_directory_from_pdf(pdf_path)
+    
+    duration = (datetime.now() - start_time).total_seconds()
+    
+    if result is None:
+        log_error("目录提取失败")
+        sys.exit(1)
+    
+    print("-" * 70)
+    print(f"\n📊 提取结果:")
+    print(f"  来源: {result.get('source', 'unknown')}")
+    print(f"  总页数: {result.get('total_pages', '?')}")
+    print(f"  扁平目录项数量: {len(result.get('flat_chapters', []))}")
+    print(f"  层级根节点数量: {len(result.get('hierarchical_chapters', []))}")
+    print(f"  处理耗时: {duration:.3f}秒")
+    
+    # 打印扁平目录
+    print(f"\n📋 扁平目录 (flat_chapters):")
+    print("-" * 40)
+    for chapter in result.get("flat_chapters", []):
+        level = chapter.get("level", 1)
+        indent = "  " * (level - 1)
+        print(f"{indent}[{chapter.get('chapterId')}] {chapter.get('title')} (L{level}, 第{chapter.get('startPage')}页)")
+    
+    # 打印层级目录
+    print(f"\n🌳 层级目录 (hierarchical_chapters):")
+    print("-" * 40)
+    print_catalog_tree(result.get("hierarchical_chapters", []))
+    
+    # 保存JSON结果
+    output_file = os.path.splitext(pdf_path)[0] + "_catalog_test.json"
+    with open(output_file, 'w', encoding='utf-8') as f:
+        json.dump(result, f, ensure_ascii=False, indent=2)
+    print(f"\n✅ JSON结果已保存到: {output_file}")
+    
+    print("\n" + "=" * 70)
+    print(f"⏰ 完成时间: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}")
+    print("=" * 70)
+
+
+if __name__ == "__main__":
+    main()