Merge branch 'dev'

- Add mock data files for directories, fingerprints, IP addresses, notification settings, nuclei templates, search, system logs, tools, and wordlists
- Update mock index to export new mock data modules
- Increase SCAN_LOAD_CHECK_INTERVAL from 30 to 180 seconds for better system stability
- Improve load check logging message to clarify OOM prevention strategy
- Enhance mock data infrastructure to support frontend development and testing

.github/workflows/docker-build.yml

@@ -19,7 +19,8 @@ permissions:
   contents: write
 
 jobs:
-  build:
+  # AMD64 构建（原生 x64 runner）
+  build-amd64:
     runs-on: ubuntu-latest
     strategy:
       matrix:
@@ -27,39 +28,30 @@ jobs:
           - image: xingrin-server
             dockerfile: docker/server/Dockerfile
             context: .
-            platforms: linux/amd64,linux/arm64
           - image: xingrin-frontend
             dockerfile: docker/frontend/Dockerfile
             context: .
-            platforms: linux/amd64  # ARM64 构建时 Next.js 在 QEMU 下会崩溃
           - image: xingrin-worker
             dockerfile: docker/worker/Dockerfile
             context: .
-            platforms: linux/amd64,linux/arm64
           - image: xingrin-nginx
             dockerfile: docker/nginx/Dockerfile
             context: .
-            platforms: linux/amd64,linux/arm64
           - image: xingrin-agent
             dockerfile: docker/agent/Dockerfile
             context: .
-            platforms: linux/amd64,linux/arm64
+          - image: xingrin-postgres
+            dockerfile: docker/postgres/Dockerfile
+            context: docker/postgres
 
     steps:
       - name: Checkout
         uses: actions/checkout@v4
 
-      - name: Free disk space (for large builds like worker)
+      - name: Free disk space
         run: |
-          echo "=== Before cleanup ==="
-          df -h
-          sudo rm -rf /usr/share/dotnet
-          sudo rm -rf /usr/local/lib/android
-          sudo rm -rf /opt/ghc
-          sudo rm -rf /opt/hostedtoolcache/CodeQL
+          sudo rm -rf /usr/share/dotnet /usr/local/lib/android /opt/ghc /opt/hostedtoolcache/CodeQL
           sudo docker image prune -af
-          echo "=== After cleanup ==="
-          df -h
 
       - name: Generate SSL certificates for nginx build
         if: matrix.image == 'xingrin-nginx'
@@ -69,10 +61,6 @@ jobs:
             -keyout docker/nginx/ssl/privkey.pem \
             -out docker/nginx/ssl/fullchain.pem \
             -subj "/CN=localhost"
-          echo "SSL certificates generated for CI build"
-
-      - name: Set up QEMU
-        uses: docker/setup-qemu-action@v3
 
       - name: Set up Docker Buildx
         uses: docker/setup-buildx-action@v3
@@ -83,7 +71,120 @@ jobs:
           username: ${{ secrets.DOCKERHUB_USERNAME }}
           password: ${{ secrets.DOCKERHUB_TOKEN }}
 
-      - name: Get version from git tag
+      - name: Get version
+        id: version
+        run: |
+          if [[ $GITHUB_REF == refs/tags/* ]]; then
+            echo "VERSION=${GITHUB_REF#refs/tags/}" >> $GITHUB_OUTPUT
+          else
+            echo "VERSION=dev-$(git rev-parse --short HEAD)" >> $GITHUB_OUTPUT
+          fi
+
+      - name: Build and push AMD64
+        uses: docker/build-push-action@v5
+        with:
+          context: ${{ matrix.context }}
+          file: ${{ matrix.dockerfile }}
+          platforms: linux/amd64
+          push: true
+          tags: ${{ env.IMAGE_PREFIX }}/${{ matrix.image }}:${{ steps.version.outputs.VERSION }}-amd64
+          build-args: IMAGE_TAG=${{ steps.version.outputs.VERSION }}
+          cache-from: type=registry,ref=${{ env.IMAGE_PREFIX }}/${{ matrix.image }}:cache-amd64
+          cache-to: type=registry,ref=${{ env.IMAGE_PREFIX }}/${{ matrix.image }}:cache-amd64,mode=max
+          provenance: false
+          sbom: false
+
+  # ARM64 构建（原生 ARM64 runner）
+  build-arm64:
+    runs-on: ubuntu-22.04-arm
+    strategy:
+      matrix:
+        include:
+          - image: xingrin-server
+            dockerfile: docker/server/Dockerfile
+            context: .
+          - image: xingrin-frontend
+            dockerfile: docker/frontend/Dockerfile
+            context: .
+          - image: xingrin-worker
+            dockerfile: docker/worker/Dockerfile
+            context: .
+          - image: xingrin-nginx
+            dockerfile: docker/nginx/Dockerfile
+            context: .
+          - image: xingrin-agent
+            dockerfile: docker/agent/Dockerfile
+            context: .
+          - image: xingrin-postgres
+            dockerfile: docker/postgres/Dockerfile
+            context: docker/postgres
+
+    steps:
+      - name: Checkout
+        uses: actions/checkout@v4
+
+      - name: Generate SSL certificates for nginx build
+        if: matrix.image == 'xingrin-nginx'
+        run: |
+          mkdir -p docker/nginx/ssl
+          openssl req -x509 -nodes -days 365 -newkey rsa:2048 \
+            -keyout docker/nginx/ssl/privkey.pem \
+            -out docker/nginx/ssl/fullchain.pem \
+            -subj "/CN=localhost"
+
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v3
+
+      - name: Login to Docker Hub
+        uses: docker/login-action@v3
+        with:
+          username: ${{ secrets.DOCKERHUB_USERNAME }}
+          password: ${{ secrets.DOCKERHUB_TOKEN }}
+
+      - name: Get version
+        id: version
+        run: |
+          if [[ $GITHUB_REF == refs/tags/* ]]; then
+            echo "VERSION=${GITHUB_REF#refs/tags/}" >> $GITHUB_OUTPUT
+          else
+            echo "VERSION=dev-$(git rev-parse --short HEAD)" >> $GITHUB_OUTPUT
+          fi
+
+      - name: Build and push ARM64
+        uses: docker/build-push-action@v5
+        with:
+          context: ${{ matrix.context }}
+          file: ${{ matrix.dockerfile }}
+          platforms: linux/arm64
+          push: true
+          tags: ${{ env.IMAGE_PREFIX }}/${{ matrix.image }}:${{ steps.version.outputs.VERSION }}-arm64
+          build-args: IMAGE_TAG=${{ steps.version.outputs.VERSION }}
+          cache-from: type=registry,ref=${{ env.IMAGE_PREFIX }}/${{ matrix.image }}:cache-arm64
+          cache-to: type=registry,ref=${{ env.IMAGE_PREFIX }}/${{ matrix.image }}:cache-arm64,mode=max
+          provenance: false
+          sbom: false
+
+  # 合并多架构 manifest
+  merge-manifests:
+    runs-on: ubuntu-latest
+    needs: [build-amd64, build-arm64]
+    strategy:
+      matrix:
+        image:
+          - xingrin-server
+          - xingrin-frontend
+          - xingrin-worker
+          - xingrin-nginx
+          - xingrin-agent
+          - xingrin-postgres
+    steps:
+      - name: Login to Docker Hub
+        uses: docker/login-action@v3
+        with:
+          username: ${{ secrets.DOCKERHUB_USERNAME }}
+          password: ${{ secrets.DOCKERHUB_TOKEN }}
+
+      - name: Get version
         id: version
         run: |
           if [[ $GITHUB_REF == refs/tags/* ]]; then
@@ -94,28 +195,27 @@ jobs:
             echo "IS_RELEASE=false" >> $GITHUB_OUTPUT
           fi
 
-      - name: Build and push
-        uses: docker/build-push-action@v5
-        with:
-          context: ${{ matrix.context }}
-          file: ${{ matrix.dockerfile }}
-          platforms: ${{ matrix.platforms }}
-          push: true
-          tags: |
-            ${{ env.IMAGE_PREFIX }}/${{ matrix.image }}:${{ steps.version.outputs.VERSION }}
-            ${{ steps.version.outputs.IS_RELEASE == 'true' && format('{0}/{1}:latest', env.IMAGE_PREFIX, matrix.image) || '' }}
-          build-args: |
-            IMAGE_TAG=${{ steps.version.outputs.VERSION }}
-          cache-from: type=gha,scope=${{ matrix.image }}
-          cache-to: type=gha,mode=max,scope=${{ matrix.image }}
-          provenance: false
-          sbom: false
+      - name: Create and push multi-arch manifest
+        run: |
+          VERSION=${{ steps.version.outputs.VERSION }}
+          IMAGE=${{ env.IMAGE_PREFIX }}/${{ matrix.image }}
+          
+          docker manifest create ${IMAGE}:${VERSION} \
+            ${IMAGE}:${VERSION}-amd64 \
+            ${IMAGE}:${VERSION}-arm64
+          docker manifest push ${IMAGE}:${VERSION}
+          
+          if [[ "${{ steps.version.outputs.IS_RELEASE }}" == "true" ]]; then
+            docker manifest create ${IMAGE}:latest \
+              ${IMAGE}:${VERSION}-amd64 \
+              ${IMAGE}:${VERSION}-arm64
+            docker manifest push ${IMAGE}:latest
+          fi
 
-  # 所有镜像构建成功后，更新 VERSION 文件
-  # 根据 tag 所在的分支更新对应分支的 VERSION 文件
+  # 更新 VERSION 文件
   update-version:
     runs-on: ubuntu-latest
-    needs: build
+    needs: merge-manifests
     if: startsWith(github.ref, 'refs/tags/v')
     steps:
       - name: Checkout repository

README.md

@@ -13,14 +13,14 @@
 
 <p align="center">
   <a href="#-功能特性">功能特性</a> •
+  <a href="#-全局资产搜索">资产搜索</a> •
   <a href="#-快速开始">快速开始</a> •
   <a href="#-文档">文档</a> •
-  <a href="#-技术栈">技术栈</a> •
   <a href="#-反馈与贡献">反馈与贡献</a>
 </p>
 
 <p align="center">
-  <sub>🔍 关键词: ASM | 攻击面管理 | 漏洞扫描 | 资产发现 | Bug Bounty | 渗透测试 | Nuclei | 子域名枚举 | EASM</sub>
+  <sub>🔍 关键词: ASM | 攻击面管理 | 漏洞扫描 | 资产发现 | 资产搜索 | Bug Bounty | 渗透测试 | Nuclei | 子域名枚举 | EASM</sub>
 </p>
 
 ---
@@ -62,9 +62,14 @@
 - **自定义流程** - YAML 配置扫描流程，灵活编排
 - **定时扫描** - Cron 表达式配置，自动化周期扫描
 
+### 🔖 指纹识别
+- **多源指纹库** - 内置 EHole、Goby、Wappalyzer、Fingers、FingerPrintHub、ARL 等 2.7W+ 指纹规则
+- **自动识别** - 扫描流程自动执行，识别 Web 应用技术栈
+- **指纹管理** - 支持查询、导入、导出指纹规则
+
 #### 扫描流程架构
 
-完整的扫描流程包括：子域名发现、端口扫描、站点发现、URL 收集、目录扫描、漏洞扫描等阶段
+完整的扫描流程包括：子域名发现、端口扫描、站点发现、指纹识别、URL 收集、目录扫描、漏洞扫描等阶段
 
 ```mermaid
 flowchart LR
@@ -75,7 +80,8 @@ flowchart LR
         SUB["子域名发现<br/>subfinder, amass, puredns"]
         PORT["端口扫描<br/>naabu"]
         SITE["站点识别<br/>httpx"]
-        SUB --> PORT --> SITE
+        FINGER["指纹识别<br/>xingfinger"]
+        SUB --> PORT --> SITE --> FINGER
     end
     
     subgraph STAGE2["阶段 2: 深度分析"]
@@ -91,7 +97,7 @@ flowchart LR
     FINISH["扫描完成"]
     
     START --> STAGE1
-    SITE --> STAGE2
+    FINGER --> STAGE2
     STAGE2 --> STAGE3
     STAGE3 --> FINISH
     
@@ -103,6 +109,7 @@ flowchart LR
     style SUB fill:#5dade2,stroke:#3498db,stroke-width:1px,color:#fff
     style PORT fill:#5dade2,stroke:#3498db,stroke-width:1px,color:#fff
     style SITE fill:#5dade2,stroke:#3498db,stroke-width:1px,color:#fff
+    style FINGER fill:#5dade2,stroke:#3498db,stroke-width:1px,color:#fff
     style URL fill:#bb8fce,stroke:#9b59b6,stroke-width:1px,color:#fff
     style DIR fill:#bb8fce,stroke:#9b59b6,stroke-width:1px,color:#fff
     style VULN fill:#f0b27a,stroke:#e67e22,stroke-width:1px,color:#fff
@@ -155,9 +162,34 @@ flowchart TB
     W3 -.心跳上报.-> REDIS
 ```
 
+### 🔎 全局资产搜索
+- **多类型搜索** - 支持 Website 和 Endpoint 两种资产类型
+- **表达式语法** - 支持 `=`（模糊）、`==`（精确）、`!=`（不等于）操作符
+- **逻辑组合** - 支持 `&&` (AND) 和 `||` (OR) 逻辑组合
+- **多字段查询** - 支持 host、url、title、tech、status、body、header 字段
+- **CSV 导出** - 流式导出全部搜索结果，无数量限制
+
+#### 搜索语法示例
+
+```bash
+# 基础搜索
+host="api"                    # host 包含 "api"
+status=="200"                 # 状态码精确等于 200
+tech="nginx"                  # 技术栈包含 nginx
+
+# 组合搜索
+host="api" && status=="200"   # host 包含 api 且状态码为 200
+tech="vue" || tech="react"    # 技术栈包含 vue 或 react
+
+# 复杂查询
+host="admin" && tech="php" && status=="200"
+url="/api/v1" && status!="404"
+```
+
 ### 📊 可视化界面
 - **数据统计** - 资产/漏洞统计仪表盘
 - **实时通知** - WebSocket 消息推送
+- **通知推送** - 实时企业微信，tg，discard消息推送服务
 
 ---
 
@@ -165,7 +197,8 @@ flowchart TB
 
 ### 环境要求
 
-- **操作系统**: Ubuntu 20.04+ / Debian 11+ （推荐）
+- **操作系统**: Ubuntu 20.04+ / Debian 11+ 
+- **系统架构**: AMD64 (x86_64) / ARM64 (aarch64)
 - **硬件**: 2核 4G 内存起步，20GB+ 磁盘空间
 
 ### 一键安装
@@ -190,6 +223,7 @@ sudo ./install.sh --mirror
 ### 访问服务
 
 - **Web 界面**: `https://ip:8083` 
+- **默认账号**: admin / admin（首次登录后请修改密码）
 
 ### 常用命令
 
@@ -209,14 +243,10 @@ sudo ./uninstall.sh
 
 ## 🤝 反馈与贡献
 
-- 🐛 **如果发现 Bug** 可以点击右边链接进行提交 [Issue](https://github.com/yyhuni/xingrin/issues)
-- 💡 **有新想法，比如UI设计，功能设计等** 欢迎点击右边链接进行提交建议 [Issue](https://github.com/yyhuni/xingrin/issues)
+- 💡 **发现 Bug，有新想法，比如UI设计，功能设计等** 欢迎点击右边链接进行提交建议 [Issue](https://github.com/yyhuni/xingrin/issues) 或者公众号私信
 
 ## 📧 联系
-- 目前版本就我个人使用，可能会有很多边界问题
-- 如有问题，建议，其他，优先提交[Issue](https://github.com/yyhuni/xingrin/issues)，也可以直接给我的公众号发消息，我都会回复的
-
-- 微信公众号: **洋洋的小黑屋**
+- 微信公众号: **塔罗安全学苑**
 
 <img src="docs/wechat-qrcode.png" alt="微信公众号" width="200">
 

VERSION

@@ -1 +1 @@
-v1.2.5-dev
+v1.3.13

backend/apps/asset/apps.py

@@ -4,7 +4,3 @@ from django.apps import AppConfig
 class AssetConfig(AppConfig):
     default_auto_field = 'django.db.models.BigAutoField'
     name = 'apps.asset'
-    
-    def ready(self):
-        # 导入所有模型以确保Django发现并注册
-        from . import models

backend/apps/asset/dtos/asset/endpoint_dto.py

@@ -14,12 +14,13 @@ class EndpointDTO:
     status_code: Optional[int] = None
     content_length: Optional[int] = None
     webserver: Optional[str] = None
-    body_preview: Optional[str] = None
+    response_body: Optional[str] = None
     content_type: Optional[str] = None
     tech: Optional[List[str]] = None
     vhost: Optional[bool] = None
     location: Optional[str] = None
     matched_gf_patterns: Optional[List[str]] = None
+    response_headers: Optional[str] = None
     
     def __post_init__(self):
         if self.tech is None:

backend/apps/asset/dtos/asset/website_dto.py

@@ -17,9 +17,10 @@ class WebSiteDTO:
     webserver: str = ''
     content_type: str = ''
     tech: List[str] = None
-    body_preview: str = ''
+    response_body: str = ''
     vhost: Optional[bool] = None
     created_at: str = None
+    response_headers: str = ''
     
     def __post_init__(self):
         if self.tech is None:

backend/apps/asset/dtos/snapshot/endpoint_snapshot_dto.py

@@ -13,6 +13,7 @@ class EndpointSnapshotDTO:
     快照只属于 scan。
     """
     scan_id: int
+    target_id: int  # 必填，用于同步到资产表
     url: str
     host: str = ''  # 主机名（域名或IP地址）
     title: str = ''
@@ -22,10 +23,10 @@ class EndpointSnapshotDTO:
     webserver: str = ''
     content_type: str = ''
     tech: List[str] = None
-    body_preview: str = ''
+    response_body: str = ''
     vhost: Optional[bool] = None
     matched_gf_patterns: List[str] = None
-    target_id: Optional[int] = None  # 冗余字段，用于同步到资产表
+    response_headers: str = ''
     
     def __post_init__(self):
         if self.tech is None:
@@ -42,9 +43,6 @@ class EndpointSnapshotDTO:
         """
         from apps.asset.dtos.asset import EndpointDTO
         
-        if self.target_id is None:
-            raise ValueError("target_id 不能为 None，无法同步到资产表")
-        
         return EndpointDTO(
             target_id=self.target_id,
             url=self.url,
@@ -53,10 +51,11 @@ class EndpointSnapshotDTO:
             status_code=self.status_code,
             content_length=self.content_length,
             webserver=self.webserver,
-            body_preview=self.body_preview,
+            response_body=self.response_body,
             content_type=self.content_type,
             tech=self.tech if self.tech else [],
             vhost=self.vhost,
             location=self.location,
-            matched_gf_patterns=self.matched_gf_patterns if self.matched_gf_patterns else []
+            matched_gf_patterns=self.matched_gf_patterns if self.matched_gf_patterns else [],
+            response_headers=self.response_headers,
         )

backend/apps/asset/dtos/snapshot/website_snapshot_dto.py

@@ -13,18 +13,19 @@ class WebsiteSnapshotDTO:
     快照只属于 scan，target 信息通过 scan.target 获取。
     """
     scan_id: int
-    target_id: int  # 仅用于传递数据，不保存到数据库
+    target_id: int  # 必填，用于同步到资产表
     url: str
     host: str
     title: str = ''
-    status: Optional[int] = None
+    status_code: Optional[int] = None  # 统一命名：status -> status_code
     content_length: Optional[int] = None
     location: str = ''
-    web_server: str = ''
+    webserver: str = ''  # 统一命名：web_server -> webserver
     content_type: str = ''
     tech: List[str] = None
-    body_preview: str = ''
+    response_body: str = ''
     vhost: Optional[bool] = None
+    response_headers: str = ''
     
     def __post_init__(self):
         if self.tech is None:
@@ -44,12 +45,13 @@ class WebsiteSnapshotDTO:
             url=self.url,
             host=self.host,
             title=self.title,
-            status_code=self.status,
+            status_code=self.status_code,
             content_length=self.content_length,
             location=self.location,
-            webserver=self.web_server,
+            webserver=self.webserver,
             content_type=self.content_type,
             tech=self.tech if self.tech else [],
-            body_preview=self.body_preview,
-            vhost=self.vhost
+            response_body=self.response_body,
+            vhost=self.vhost,
+            response_headers=self.response_headers,
         )

backend/apps/asset/migrations/0001_initial.py

@@ -0,0 +1,345 @@
+# Generated by Django 5.2.7 on 2026-01-02 04:45
+
+import django.contrib.postgres.fields
+import django.contrib.postgres.indexes
+import django.core.validators
+import django.db.models.deletion
+from django.db import migrations, models
+
+
+class Migration(migrations.Migration):
+
+    initial = True
+
+    dependencies = [
+        ('scan', '0001_initial'),
+        ('targets', '0001_initial'),
+    ]
+
+    operations = [
+        migrations.CreateModel(
+            name='AssetStatistics',
+            fields=[
+                ('id', models.AutoField(primary_key=True, serialize=False)),
+                ('total_targets', models.IntegerField(default=0, help_text='目标总数')),
+                ('total_subdomains', models.IntegerField(default=0, help_text='子域名总数')),
+                ('total_ips', models.IntegerField(default=0, help_text='IP地址总数')),
+                ('total_endpoints', models.IntegerField(default=0, help_text='端点总数')),
+                ('total_websites', models.IntegerField(default=0, help_text='网站总数')),
+                ('total_vulns', models.IntegerField(default=0, help_text='漏洞总数')),
+                ('total_assets', models.IntegerField(default=0, help_text='总资产数（子域名+IP+端点+网站）')),
+                ('prev_targets', models.IntegerField(default=0, help_text='上次目标总数')),
+                ('prev_subdomains', models.IntegerField(default=0, help_text='上次子域名总数')),
+                ('prev_ips', models.IntegerField(default=0, help_text='上次IP地址总数')),
+                ('prev_endpoints', models.IntegerField(default=0, help_text='上次端点总数')),
+                ('prev_websites', models.IntegerField(default=0, help_text='上次网站总数')),
+                ('prev_vulns', models.IntegerField(default=0, help_text='上次漏洞总数')),
+                ('prev_assets', models.IntegerField(default=0, help_text='上次总资产数')),
+                ('updated_at', models.DateTimeField(auto_now=True, help_text='最后更新时间')),
+            ],
+            options={
+                'verbose_name': '资产统计',
+                'verbose_name_plural': '资产统计',
+                'db_table': 'asset_statistics',
+            },
+        ),
+        migrations.CreateModel(
+            name='StatisticsHistory',
+            fields=[
+                ('id', models.BigAutoField(auto_created=True, primary_key=True, serialize=False, verbose_name='ID')),
+                ('date', models.DateField(help_text='统计日期', unique=True)),
+                ('total_targets', models.IntegerField(default=0, help_text='目标总数')),
+                ('total_subdomains', models.IntegerField(default=0, help_text='子域名总数')),
+                ('total_ips', models.IntegerField(default=0, help_text='IP地址总数')),
+                ('total_endpoints', models.IntegerField(default=0, help_text='端点总数')),
+                ('total_websites', models.IntegerField(default=0, help_text='网站总数')),
+                ('total_vulns', models.IntegerField(default=0, help_text='漏洞总数')),
+                ('total_assets', models.IntegerField(default=0, help_text='总资产数')),
+                ('created_at', models.DateTimeField(auto_now_add=True, help_text='创建时间')),
+                ('updated_at', models.DateTimeField(auto_now=True, help_text='更新时间')),
+            ],
+            options={
+                'verbose_name': '统计历史',
+                'verbose_name_plural': '统计历史',
+                'db_table': 'statistics_history',
+                'ordering': ['-date'],
+                'indexes': [models.Index(fields=['date'], name='statistics__date_1d29cd_idx')],
+            },
+        ),
+        migrations.CreateModel(
+            name='Directory',
+            fields=[
+                ('id', models.AutoField(primary_key=True, serialize=False)),
+                ('url', models.CharField(help_text='完整请求 URL', max_length=2000)),
+                ('status', models.IntegerField(blank=True, help_text='HTTP 响应状态码', null=True)),
+                ('content_length', models.BigIntegerField(blank=True, help_text='响应体字节大小（Content-Length 或实际长度）', null=True)),
+                ('words', models.IntegerField(blank=True, help_text='响应体中单词数量（按空格分割）', null=True)),
+                ('lines', models.IntegerField(blank=True, help_text='响应体行数（按换行符分割）', null=True)),
+                ('content_type', models.CharField(blank=True, default='', help_text='响应头 Content-Type 值', max_length=200)),
+                ('duration', models.BigIntegerField(blank=True, help_text='请求耗时（单位：纳秒）', null=True)),
+                ('created_at', models.DateTimeField(auto_now_add=True, help_text='创建时间')),
+                ('target', models.ForeignKey(help_text='所属的扫描目标', on_delete=django.db.models.deletion.CASCADE, related_name='directories', to='targets.target')),
+            ],
+            options={
+                'verbose_name': '目录',
+                'verbose_name_plural': '目录',
+                'db_table': 'directory',
+                'ordering': ['-created_at'],
+                'indexes': [models.Index(fields=['-created_at'], name='directory_created_2cef03_idx'), models.Index(fields=['target'], name='directory_target__e310c8_idx'), models.Index(fields=['url'], name='directory_url_ba40cd_idx'), models.Index(fields=['status'], name='directory_status_40bbe6_idx'), django.contrib.postgres.indexes.GinIndex(fields=['url'], name='directory_url_trgm_idx', opclasses=['gin_trgm_ops'])],
+                'constraints': [models.UniqueConstraint(fields=('target', 'url'), name='unique_directory_url_target')],
+            },
+        ),
+        migrations.CreateModel(
+            name='DirectorySnapshot',
+            fields=[
+                ('id', models.AutoField(primary_key=True, serialize=False)),
+                ('url', models.CharField(help_text='目录URL', max_length=2000)),
+                ('status', models.IntegerField(blank=True, help_text='HTTP状态码', null=True)),
+                ('content_length', models.BigIntegerField(blank=True, help_text='内容长度', null=True)),
+                ('words', models.IntegerField(blank=True, help_text='响应体中单词数量（按空格分割）', null=True)),
+                ('lines', models.IntegerField(blank=True, help_text='响应体行数（按换行符分割）', null=True)),
+                ('content_type', models.CharField(blank=True, default='', help_text='响应头 Content-Type 值', max_length=200)),
+                ('duration', models.BigIntegerField(blank=True, help_text='请求耗时（单位：纳秒）', null=True)),
+                ('created_at', models.DateTimeField(auto_now_add=True, help_text='创建时间')),
+                ('scan', models.ForeignKey(help_text='所属的扫描任务', on_delete=django.db.models.deletion.CASCADE, related_name='directory_snapshots', to='scan.scan')),
+            ],
+            options={
+                'verbose_name': '目录快照',
+                'verbose_name_plural': '目录快照',
+                'db_table': 'directory_snapshot',
+                'ordering': ['-created_at'],
+                'indexes': [models.Index(fields=['scan'], name='directory_s_scan_id_c45900_idx'), models.Index(fields=['url'], name='directory_s_url_b4b72b_idx'), models.Index(fields=['status'], name='directory_s_status_e9f57e_idx'), models.Index(fields=['content_type'], name='directory_s_content_45e864_idx'), models.Index(fields=['-created_at'], name='directory_s_created_eb9d27_idx'), django.contrib.postgres.indexes.GinIndex(fields=['url'], name='dir_snap_url_trgm', opclasses=['gin_trgm_ops'])],
+                'constraints': [models.UniqueConstraint(fields=('scan', 'url'), name='unique_directory_per_scan_snapshot')],
+            },
+        ),
+        migrations.CreateModel(
+            name='Endpoint',
+            fields=[
+                ('id', models.AutoField(primary_key=True, serialize=False)),
+                ('url', models.TextField(help_text='最终访问的完整URL')),
+                ('host', models.CharField(blank=True, default='', help_text='主机名（域名或IP地址）', max_length=253)),
+                ('location', models.TextField(blank=True, default='', help_text='重定向地址（HTTP 3xx 响应头 Location）')),
+                ('created_at', models.DateTimeField(auto_now_add=True, help_text='创建时间')),
+                ('title', models.TextField(blank=True, default='', help_text='网页标题（HTML <title> 标签内容）')),
+                ('webserver', models.TextField(blank=True, default='', help_text='服务器类型（HTTP 响应头 Server 值）')),
+                ('response_body', models.TextField(blank=True, default='', help_text='HTTP响应体')),
+                ('content_type', models.TextField(blank=True, default='', help_text='响应类型（HTTP Content-Type 响应头）')),
+                ('tech', django.contrib.postgres.fields.ArrayField(base_field=models.CharField(max_length=100), blank=True, default=list, help_text='技术栈（服务器/框架/语言等）', size=None)),
+                ('status_code', models.IntegerField(blank=True, help_text='HTTP状态码', null=True)),
+                ('content_length', models.IntegerField(blank=True, help_text='响应体大小（单位字节）', null=True)),
+                ('vhost', models.BooleanField(blank=True, help_text='是否支持虚拟主机', null=True)),
+                ('matched_gf_patterns', django.contrib.postgres.fields.ArrayField(base_field=models.CharField(max_length=100), blank=True, default=list, help_text='匹配的GF模式列表，用于识别敏感端点（如api, debug, config等）', size=None)),
+                ('response_headers', models.TextField(blank=True, default='', help_text='原始HTTP响应头')),
+                ('target', models.ForeignKey(help_text='所属的扫描目标（主关联字段，表示所属关系，不能为空）', on_delete=django.db.models.deletion.CASCADE, related_name='endpoints', to='targets.target')),
+            ],
+            options={
+                'verbose_name': '端点',
+                'verbose_name_plural': '端点',
+                'db_table': 'endpoint',
+                'ordering': ['-created_at'],
+                'indexes': [models.Index(fields=['-created_at'], name='endpoint_created_44fe9c_idx'), models.Index(fields=['target'], name='endpoint_target__7f9065_idx'), models.Index(fields=['url'], name='endpoint_url_30f66e_idx'), models.Index(fields=['host'], name='endpoint_host_5b4cc8_idx'), models.Index(fields=['status_code'], name='endpoint_status__5d4fdd_idx'), models.Index(fields=['title'], name='endpoint_title_29e26c_idx'), django.contrib.postgres.indexes.GinIndex(fields=['tech'], name='endpoint_tech_2bfa7c_gin'), django.contrib.postgres.indexes.GinIndex(fields=['response_headers'], name='endpoint_resp_headers_trgm_idx', opclasses=['gin_trgm_ops']), django.contrib.postgres.indexes.GinIndex(fields=['url'], name='endpoint_url_trgm_idx', opclasses=['gin_trgm_ops']), django.contrib.postgres.indexes.GinIndex(fields=['title'], name='endpoint_title_trgm_idx', opclasses=['gin_trgm_ops'])],
+                'constraints': [models.UniqueConstraint(fields=('url', 'target'), name='unique_endpoint_url_target')],
+            },
+        ),
+        migrations.CreateModel(
+            name='EndpointSnapshot',
+            fields=[
+                ('id', models.AutoField(primary_key=True, serialize=False)),
+                ('url', models.TextField(help_text='端点URL')),
+                ('host', models.CharField(blank=True, default='', help_text='主机名（域名或IP地址）', max_length=253)),
+                ('title', models.TextField(blank=True, default='', help_text='页面标题')),
+                ('status_code', models.IntegerField(blank=True, help_text='HTTP状态码', null=True)),
+                ('content_length', models.IntegerField(blank=True, help_text='内容长度', null=True)),
+                ('location', models.TextField(blank=True, default='', help_text='重定向位置')),
+                ('webserver', models.TextField(blank=True, default='', help_text='Web服务器')),
+                ('content_type', models.TextField(blank=True, default='', help_text='内容类型')),
+                ('tech', django.contrib.postgres.fields.ArrayField(base_field=models.CharField(max_length=100), blank=True, default=list, help_text='技术栈', size=None)),
+                ('response_body', models.TextField(blank=True, default='', help_text='HTTP响应体')),
+                ('vhost', models.BooleanField(blank=True, help_text='虚拟主机标志', null=True)),
+                ('matched_gf_patterns', django.contrib.postgres.fields.ArrayField(base_field=models.CharField(max_length=100), blank=True, default=list, help_text='匹配的GF模式列表', size=None)),
+                ('response_headers', models.TextField(blank=True, default='', help_text='原始HTTP响应头')),
+                ('created_at', models.DateTimeField(auto_now_add=True, help_text='创建时间')),
+                ('scan', models.ForeignKey(help_text='所属的扫描任务', on_delete=django.db.models.deletion.CASCADE, related_name='endpoint_snapshots', to='scan.scan')),
+            ],
+            options={
+                'verbose_name': '端点快照',
+                'verbose_name_plural': '端点快照',
+                'db_table': 'endpoint_snapshot',
+                'ordering': ['-created_at'],
+                'indexes': [models.Index(fields=['scan'], name='endpoint_sn_scan_id_6ac9a7_idx'), models.Index(fields=['url'], name='endpoint_sn_url_205160_idx'), models.Index(fields=['host'], name='endpoint_sn_host_577bfd_idx'), models.Index(fields=['title'], name='endpoint_sn_title_516a05_idx'), models.Index(fields=['status_code'], name='endpoint_sn_status__83efb0_idx'), models.Index(fields=['webserver'], name='endpoint_sn_webserv_66be83_idx'), models.Index(fields=['-created_at'], name='endpoint_sn_created_21fb5b_idx'), django.contrib.postgres.indexes.GinIndex(fields=['tech'], name='endpoint_sn_tech_0d0752_gin'), django.contrib.postgres.indexes.GinIndex(fields=['response_headers'], name='ep_snap_resp_hdr_trgm', opclasses=['gin_trgm_ops']), django.contrib.postgres.indexes.GinIndex(fields=['url'], name='ep_snap_url_trgm', opclasses=['gin_trgm_ops']), django.contrib.postgres.indexes.GinIndex(fields=['title'], name='ep_snap_title_trgm', opclasses=['gin_trgm_ops'])],
+                'constraints': [models.UniqueConstraint(fields=('scan', 'url'), name='unique_endpoint_per_scan_snapshot')],
+            },
+        ),
+        migrations.CreateModel(
+            name='HostPortMapping',
+            fields=[
+                ('id', models.AutoField(primary_key=True, serialize=False)),
+                ('host', models.CharField(help_text='主机名（域名或IP）', max_length=1000)),
+                ('ip', models.GenericIPAddressField(help_text='IP地址')),
+                ('port', models.IntegerField(help_text='端口号（1-65535）', validators=[django.core.validators.MinValueValidator(1, message='端口号必须大于等于1'), django.core.validators.MaxValueValidator(65535, message='端口号必须小于等于65535')])),
+                ('created_at', models.DateTimeField(auto_now_add=True, help_text='创建时间')),
+                ('target', models.ForeignKey(help_text='所属的扫描目标', on_delete=django.db.models.deletion.CASCADE, related_name='host_port_mappings', to='targets.target')),
+            ],
+            options={
+                'verbose_name': '主机端口映射',
+                'verbose_name_plural': '主机端口映射',
+                'db_table': 'host_port_mapping',
+                'ordering': ['-created_at'],
+                'indexes': [models.Index(fields=['target'], name='host_port_m_target__943e9b_idx'), models.Index(fields=['host'], name='host_port_m_host_f78363_idx'), models.Index(fields=['ip'], name='host_port_m_ip_2e6f02_idx'), models.Index(fields=['port'], name='host_port_m_port_9fb9ff_idx'), models.Index(fields=['host', 'ip'], name='host_port_m_host_3ce245_idx'), models.Index(fields=['-created_at'], name='host_port_m_created_11cd22_idx')],
+                'constraints': [models.UniqueConstraint(fields=('target', 'host', 'ip', 'port'), name='unique_target_host_ip_port')],
+            },
+        ),
+        migrations.CreateModel(
+            name='HostPortMappingSnapshot',
+            fields=[
+                ('id', models.AutoField(primary_key=True, serialize=False)),
+                ('host', models.CharField(help_text='主机名（域名或IP）', max_length=1000)),
+                ('ip', models.GenericIPAddressField(help_text='IP地址')),
+                ('port', models.IntegerField(help_text='端口号（1-65535）', validators=[django.core.validators.MinValueValidator(1, message='端口号必须大于等于1'), django.core.validators.MaxValueValidator(65535, message='端口号必须小于等于65535')])),
+                ('created_at', models.DateTimeField(auto_now_add=True, help_text='创建时间')),
+                ('scan', models.ForeignKey(help_text='所属的扫描任务（主关联）', on_delete=django.db.models.deletion.CASCADE, related_name='host_port_mapping_snapshots', to='scan.scan')),
+            ],
+            options={
+                'verbose_name': '主机端口映射快照',
+                'verbose_name_plural': '主机端口映射快照',
+                'db_table': 'host_port_mapping_snapshot',
+                'ordering': ['-created_at'],
+                'indexes': [models.Index(fields=['scan'], name='host_port_m_scan_id_50ba0b_idx'), models.Index(fields=['host'], name='host_port_m_host_e99054_idx'), models.Index(fields=['ip'], name='host_port_m_ip_54818c_idx'), models.Index(fields=['port'], name='host_port_m_port_ed7b48_idx'), models.Index(fields=['host', 'ip'], name='host_port_m_host_8a463a_idx'), models.Index(fields=['scan', 'host'], name='host_port_m_scan_id_426fdb_idx'), models.Index(fields=['-created_at'], name='host_port_m_created_fb28b8_idx')],
+                'constraints': [models.UniqueConstraint(fields=('scan', 'host', 'ip', 'port'), name='unique_scan_host_ip_port_snapshot')],
+            },
+        ),
+        migrations.CreateModel(
+            name='Subdomain',
+            fields=[
+                ('id', models.AutoField(primary_key=True, serialize=False)),
+                ('name', models.CharField(help_text='子域名名称', max_length=1000)),
+                ('created_at', models.DateTimeField(auto_now_add=True, help_text='创建时间')),
+                ('target', models.ForeignKey(help_text='所属的扫描目标（主关联字段，表示所属关系，不能为空）', on_delete=django.db.models.deletion.CASCADE, related_name='subdomains', to='targets.target')),
+            ],
+            options={
+                'verbose_name': '子域名',
+                'verbose_name_plural': '子域名',
+                'db_table': 'subdomain',
+                'ordering': ['-created_at'],
+                'indexes': [models.Index(fields=['-created_at'], name='subdomain_created_e187a8_idx'), models.Index(fields=['name', 'target'], name='subdomain_name_60e1d0_idx'), models.Index(fields=['target'], name='subdomain_target__e409f0_idx'), models.Index(fields=['name'], name='subdomain_name_d40ba7_idx'), django.contrib.postgres.indexes.GinIndex(fields=['name'], name='subdomain_name_trgm_idx', opclasses=['gin_trgm_ops'])],
+                'constraints': [models.UniqueConstraint(fields=('name', 'target'), name='unique_subdomain_name_target')],
+            },
+        ),
+        migrations.CreateModel(
+            name='SubdomainSnapshot',
+            fields=[
+                ('id', models.AutoField(primary_key=True, serialize=False)),
+                ('name', models.CharField(help_text='子域名名称', max_length=1000)),
+                ('created_at', models.DateTimeField(auto_now_add=True, help_text='创建时间')),
+                ('scan', models.ForeignKey(help_text='所属的扫描任务', on_delete=django.db.models.deletion.CASCADE, related_name='subdomain_snapshots', to='scan.scan')),
+            ],
+            options={
+                'verbose_name': '子域名快照',
+                'verbose_name_plural': '子域名快照',
+                'db_table': 'subdomain_snapshot',
+                'ordering': ['-created_at'],
+                'indexes': [models.Index(fields=['scan'], name='subdomain_s_scan_id_68c253_idx'), models.Index(fields=['name'], name='subdomain_s_name_2da42b_idx'), models.Index(fields=['-created_at'], name='subdomain_s_created_d2b48e_idx'), django.contrib.postgres.indexes.GinIndex(fields=['name'], name='subdomain_snap_name_trgm', opclasses=['gin_trgm_ops'])],
+                'constraints': [models.UniqueConstraint(fields=('scan', 'name'), name='unique_subdomain_per_scan_snapshot')],
+            },
+        ),
+        migrations.CreateModel(
+            name='Vulnerability',
+            fields=[
+                ('id', models.AutoField(primary_key=True, serialize=False)),
+                ('url', models.CharField(help_text='漏洞所在的URL', max_length=2000)),
+                ('vuln_type', models.CharField(help_text='漏洞类型（如 xss, sqli）', max_length=100)),
+                ('severity', models.CharField(choices=[('unknown', '未知'), ('info', '信息'), ('low', '低'), ('medium', '中'), ('high', '高'), ('critical', '危急')], default='unknown', help_text='严重性（未知/信息/低/中/高/危急）', max_length=20)),
+                ('source', models.CharField(blank=True, default='', help_text='来源工具（如 dalfox, nuclei, crlfuzz）', max_length=50)),
+                ('cvss_score', models.DecimalField(blank=True, decimal_places=1, help_text='CVSS 评分（0.0-10.0）', max_digits=3, null=True)),
+                ('description', models.TextField(blank=True, default='', help_text='漏洞描述')),
+                ('raw_output', models.JSONField(blank=True, default=dict, help_text='工具原始输出')),
+                ('created_at', models.DateTimeField(auto_now_add=True, help_text='创建时间')),
+                ('target', models.ForeignKey(help_text='所属的扫描目标', on_delete=django.db.models.deletion.CASCADE, related_name='vulnerabilities', to='targets.target')),
+            ],
+            options={
+                'verbose_name': '漏洞',
+                'verbose_name_plural': '漏洞',
+                'db_table': 'vulnerability',
+                'ordering': ['-created_at'],
+                'indexes': [models.Index(fields=['target'], name='vulnerabili_target__755a02_idx'), models.Index(fields=['vuln_type'], name='vulnerabili_vuln_ty_3010cd_idx'), models.Index(fields=['severity'], name='vulnerabili_severit_1a798b_idx'), models.Index(fields=['source'], name='vulnerabili_source_7c7552_idx'), models.Index(fields=['url'], name='vulnerabili_url_4dcc4d_idx'), models.Index(fields=['-created_at'], name='vulnerabili_created_e25ff7_idx')],
+            },
+        ),
+        migrations.CreateModel(
+            name='VulnerabilitySnapshot',
+            fields=[
+                ('id', models.AutoField(primary_key=True, serialize=False)),
+                ('url', models.CharField(help_text='漏洞所在的URL', max_length=2000)),
+                ('vuln_type', models.CharField(help_text='漏洞类型（如 xss, sqli）', max_length=100)),
+                ('severity', models.CharField(choices=[('unknown', '未知'), ('info', '信息'), ('low', '低'), ('medium', '中'), ('high', '高'), ('critical', '危急')], default='unknown', help_text='严重性（未知/信息/低/中/高/危急）', max_length=20)),
+                ('source', models.CharField(blank=True, default='', help_text='来源工具（如 dalfox, nuclei, crlfuzz）', max_length=50)),
+                ('cvss_score', models.DecimalField(blank=True, decimal_places=1, help_text='CVSS 评分（0.0-10.0）', max_digits=3, null=True)),
+                ('description', models.TextField(blank=True, default='', help_text='漏洞描述')),
+                ('raw_output', models.JSONField(blank=True, default=dict, help_text='工具原始输出')),
+                ('created_at', models.DateTimeField(auto_now_add=True, help_text='创建时间')),
+                ('scan', models.ForeignKey(help_text='所属的扫描任务', on_delete=django.db.models.deletion.CASCADE, related_name='vulnerability_snapshots', to='scan.scan')),
+            ],
+            options={
+                'verbose_name': '漏洞快照',
+                'verbose_name_plural': '漏洞快照',
+                'db_table': 'vulnerability_snapshot',
+                'ordering': ['-created_at'],
+                'indexes': [models.Index(fields=['scan'], name='vulnerabili_scan_id_7b81c9_idx'), models.Index(fields=['url'], name='vulnerabili_url_11a707_idx'), models.Index(fields=['vuln_type'], name='vulnerabili_vuln_ty_6b90ee_idx'), models.Index(fields=['severity'], name='vulnerabili_severit_4eae0d_idx'), models.Index(fields=['source'], name='vulnerabili_source_968b1f_idx'), models.Index(fields=['-created_at'], name='vulnerabili_created_53a12e_idx')],
+            },
+        ),
+        migrations.CreateModel(
+            name='WebSite',
+            fields=[
+                ('id', models.AutoField(primary_key=True, serialize=False)),
+                ('url', models.TextField(help_text='最终访问的完整URL')),
+                ('host', models.CharField(blank=True, default='', help_text='主机名（域名或IP地址）', max_length=253)),
+                ('location', models.TextField(blank=True, default='', help_text='重定向地址（HTTP 3xx 响应头 Location）')),
+                ('created_at', models.DateTimeField(auto_now_add=True, help_text='创建时间')),
+                ('title', models.TextField(blank=True, default='', help_text='网页标题（HTML <title> 标签内容）')),
+                ('webserver', models.TextField(blank=True, default='', help_text='服务器类型（HTTP 响应头 Server 值）')),
+                ('response_body', models.TextField(blank=True, default='', help_text='HTTP响应体')),
+                ('content_type', models.TextField(blank=True, default='', help_text='响应类型（HTTP Content-Type 响应头）')),
+                ('tech', django.contrib.postgres.fields.ArrayField(base_field=models.CharField(max_length=100), blank=True, default=list, help_text='技术栈（服务器/框架/语言等）', size=None)),
+                ('status_code', models.IntegerField(blank=True, help_text='HTTP状态码', null=True)),
+                ('content_length', models.IntegerField(blank=True, help_text='响应体大小（单位字节）', null=True)),
+                ('vhost', models.BooleanField(blank=True, help_text='是否支持虚拟主机', null=True)),
+                ('response_headers', models.TextField(blank=True, default='', help_text='原始HTTP响应头')),
+                ('target', models.ForeignKey(help_text='所属的扫描目标（主关联字段，表示所属关系，不能为空）', on_delete=django.db.models.deletion.CASCADE, related_name='websites', to='targets.target')),
+            ],
+            options={
+                'verbose_name': '站点',
+                'verbose_name_plural': '站点',
+                'db_table': 'website',
+                'ordering': ['-created_at'],
+                'indexes': [models.Index(fields=['-created_at'], name='website_created_c9cfd2_idx'), models.Index(fields=['url'], name='website_url_b18883_idx'), models.Index(fields=['host'], name='website_host_996b50_idx'), models.Index(fields=['target'], name='website_target__2a353b_idx'), models.Index(fields=['title'], name='website_title_c2775b_idx'), models.Index(fields=['status_code'], name='website_status__51663d_idx'), django.contrib.postgres.indexes.GinIndex(fields=['tech'], name='website_tech_e3f0cb_gin'), django.contrib.postgres.indexes.GinIndex(fields=['response_headers'], name='website_resp_headers_trgm_idx', opclasses=['gin_trgm_ops']), django.contrib.postgres.indexes.GinIndex(fields=['url'], name='website_url_trgm_idx', opclasses=['gin_trgm_ops']), django.contrib.postgres.indexes.GinIndex(fields=['title'], name='website_title_trgm_idx', opclasses=['gin_trgm_ops'])],
+                'constraints': [models.UniqueConstraint(fields=('url', 'target'), name='unique_website_url_target')],
+            },
+        ),
+        migrations.CreateModel(
+            name='WebsiteSnapshot',
+            fields=[
+                ('id', models.AutoField(primary_key=True, serialize=False)),
+                ('url', models.TextField(help_text='站点URL')),
+                ('host', models.CharField(blank=True, default='', help_text='主机名（域名或IP地址）', max_length=253)),
+                ('title', models.TextField(blank=True, default='', help_text='页面标题')),
+                ('status_code', models.IntegerField(blank=True, help_text='HTTP状态码', null=True)),
+                ('content_length', models.BigIntegerField(blank=True, help_text='内容长度', null=True)),
+                ('location', models.TextField(blank=True, default='', help_text='重定向位置')),
+                ('webserver', models.TextField(blank=True, default='', help_text='Web服务器')),
+                ('content_type', models.TextField(blank=True, default='', help_text='内容类型')),
+                ('tech', django.contrib.postgres.fields.ArrayField(base_field=models.CharField(max_length=100), blank=True, default=list, help_text='技术栈', size=None)),
+                ('response_body', models.TextField(blank=True, default='', help_text='HTTP响应体')),
+                ('vhost', models.BooleanField(blank=True, help_text='虚拟主机标志', null=True)),
+                ('response_headers', models.TextField(blank=True, default='', help_text='原始HTTP响应头')),
+                ('created_at', models.DateTimeField(auto_now_add=True, help_text='创建时间')),
+                ('scan', models.ForeignKey(help_text='所属的扫描任务', on_delete=django.db.models.deletion.CASCADE, related_name='website_snapshots', to='scan.scan')),
+            ],
+            options={
+                'verbose_name': '网站快照',
+                'verbose_name_plural': '网站快照',
+                'db_table': 'website_snapshot',
+                'ordering': ['-created_at'],
+                'indexes': [models.Index(fields=['scan'], name='website_sna_scan_id_26b6dc_idx'), models.Index(fields=['url'], name='website_sna_url_801a70_idx'), models.Index(fields=['host'], name='website_sna_host_348fe1_idx'), models.Index(fields=['title'], name='website_sna_title_b1a5ee_idx'), models.Index(fields=['-created_at'], name='website_sna_created_2c149a_idx'), django.contrib.postgres.indexes.GinIndex(fields=['tech'], name='website_sna_tech_3d6d2f_gin'), django.contrib.postgres.indexes.GinIndex(fields=['response_headers'], name='ws_snap_resp_hdr_trgm', opclasses=['gin_trgm_ops']), django.contrib.postgres.indexes.GinIndex(fields=['url'], name='ws_snap_url_trgm', opclasses=['gin_trgm_ops']), django.contrib.postgres.indexes.GinIndex(fields=['title'], name='ws_snap_title_trgm', opclasses=['gin_trgm_ops'])],
+                'constraints': [models.UniqueConstraint(fields=('scan', 'url'), name='unique_website_per_scan_snapshot')],
+            },
+        ),
+    ]

backend/apps/asset/migrations/0002_create_search_materialized_view.py

@@ -0,0 +1,196 @@
+"""
+创建资产搜索 IMMV（增量维护物化视图）
+
+使用 pg_ivm 扩展创建 IMMV，数据变更时自动增量更新，无需手动刷新。
+
+包含：
+1. asset_search_view - Website 搜索视图
+2. endpoint_search_view - Endpoint 搜索视图
+
+重要限制：
+⚠️ pg_ivm 不支持数组类型字段（ArrayField），因为其使用 anyarray 伪类型进行比较时，
+PostgreSQL 无法确定空数组的元素类型，导致错误：
+  "cannot determine element type of \"anyarray\" argument"
+
+因此，所有 ArrayField 字段（tech, matched_gf_patterns 等）已从 IMMV 中移除，
+搜索时通过 JOIN 原表获取。
+
+如需添加新的数组字段，请：
+1. 不要将其包含在 IMMV 视图中
+2. 在搜索服务中通过 JOIN 原表获取
+"""
+
+from django.db import migrations
+
+
+class Migration(migrations.Migration):
+
+    dependencies = [
+        ('asset', '0001_initial'),
+    ]
+
+    operations = [
+        # 1. 确保 pg_trgm 扩展已启用（用于文本模糊搜索索引）
+        migrations.RunSQL(
+            sql="CREATE EXTENSION IF NOT EXISTS pg_trgm;",
+            reverse_sql="-- pg_trgm extension kept for other uses"
+        ),
+        
+        # 2. 确保 pg_ivm 扩展已启用（用于 IMMV 增量维护）
+        migrations.RunSQL(
+            sql="CREATE EXTENSION IF NOT EXISTS pg_ivm;",
+            reverse_sql="-- pg_ivm extension kept for other uses"
+        ),
+        
+        # ==================== Website IMMV ====================
+        
+        # 2. 创建 asset_search_view IMMV
+        # ⚠️ 注意：不包含 w.tech 数组字段，pg_ivm 不支持 ArrayField
+        # 数组字段通过 search_service.py 中 JOIN website 表获取
+        migrations.RunSQL(
+            sql="""
+                SELECT pgivm.create_immv('asset_search_view', $$
+                    SELECT 
+                        w.id,
+                        w.url,
+                        w.host,
+                        w.title,
+                        w.status_code,
+                        w.response_headers,
+                        w.response_body,
+                        w.content_type,
+                        w.content_length,
+                        w.webserver,
+                        w.location,
+                        w.vhost,
+                        w.created_at,
+                        w.target_id
+                    FROM website w
+                $$);
+            """,
+            reverse_sql="SELECT pgivm.drop_immv('asset_search_view');"
+        ),
+        
+        # 3. 创建 asset_search_view 索引
+        migrations.RunSQL(
+            sql="""
+                -- 唯一索引
+                CREATE UNIQUE INDEX IF NOT EXISTS asset_search_view_id_idx 
+                ON asset_search_view (id);
+                
+                -- host 模糊搜索索引
+                CREATE INDEX IF NOT EXISTS asset_search_view_host_trgm_idx 
+                ON asset_search_view USING gin (host gin_trgm_ops);
+                
+                -- title 模糊搜索索引
+                CREATE INDEX IF NOT EXISTS asset_search_view_title_trgm_idx 
+                ON asset_search_view USING gin (title gin_trgm_ops);
+                
+                -- url 模糊搜索索引
+                CREATE INDEX IF NOT EXISTS asset_search_view_url_trgm_idx 
+                ON asset_search_view USING gin (url gin_trgm_ops);
+                
+                -- response_headers 模糊搜索索引
+                CREATE INDEX IF NOT EXISTS asset_search_view_headers_trgm_idx 
+                ON asset_search_view USING gin (response_headers gin_trgm_ops);
+                
+                -- response_body 模糊搜索索引
+                CREATE INDEX IF NOT EXISTS asset_search_view_body_trgm_idx 
+                ON asset_search_view USING gin (response_body gin_trgm_ops);
+                
+                -- status_code 索引
+                CREATE INDEX IF NOT EXISTS asset_search_view_status_idx 
+                ON asset_search_view (status_code);
+                
+                -- created_at 排序索引
+                CREATE INDEX IF NOT EXISTS asset_search_view_created_idx 
+                ON asset_search_view (created_at DESC);
+            """,
+            reverse_sql="""
+                DROP INDEX IF EXISTS asset_search_view_id_idx;
+                DROP INDEX IF EXISTS asset_search_view_host_trgm_idx;
+                DROP INDEX IF EXISTS asset_search_view_title_trgm_idx;
+                DROP INDEX IF EXISTS asset_search_view_url_trgm_idx;
+                DROP INDEX IF EXISTS asset_search_view_headers_trgm_idx;
+                DROP INDEX IF EXISTS asset_search_view_body_trgm_idx;
+                DROP INDEX IF EXISTS asset_search_view_status_idx;
+                DROP INDEX IF EXISTS asset_search_view_created_idx;
+            """
+        ),
+
+        # ==================== Endpoint IMMV ====================
+        
+        # 4. 创建 endpoint_search_view IMMV
+        # ⚠️ 注意：不包含 e.tech 和 e.matched_gf_patterns 数组字段，pg_ivm 不支持 ArrayField
+        # 数组字段通过 search_service.py 中 JOIN endpoint 表获取
+        migrations.RunSQL(
+            sql="""
+                SELECT pgivm.create_immv('endpoint_search_view', $$
+                    SELECT 
+                        e.id,
+                        e.url,
+                        e.host,
+                        e.title,
+                        e.status_code,
+                        e.response_headers,
+                        e.response_body,
+                        e.content_type,
+                        e.content_length,
+                        e.webserver,
+                        e.location,
+                        e.vhost,
+                        e.created_at,
+                        e.target_id
+                    FROM endpoint e
+                $$);
+            """,
+            reverse_sql="SELECT pgivm.drop_immv('endpoint_search_view');"
+        ),
+        
+        # 5. 创建 endpoint_search_view 索引
+        migrations.RunSQL(
+            sql="""
+                -- 唯一索引
+                CREATE UNIQUE INDEX IF NOT EXISTS endpoint_search_view_id_idx 
+                ON endpoint_search_view (id);
+                
+                -- host 模糊搜索索引
+                CREATE INDEX IF NOT EXISTS endpoint_search_view_host_trgm_idx 
+                ON endpoint_search_view USING gin (host gin_trgm_ops);
+                
+                -- title 模糊搜索索引
+                CREATE INDEX IF NOT EXISTS endpoint_search_view_title_trgm_idx 
+                ON endpoint_search_view USING gin (title gin_trgm_ops);
+                
+                -- url 模糊搜索索引
+                CREATE INDEX IF NOT EXISTS endpoint_search_view_url_trgm_idx 
+                ON endpoint_search_view USING gin (url gin_trgm_ops);
+                
+                -- response_headers 模糊搜索索引
+                CREATE INDEX IF NOT EXISTS endpoint_search_view_headers_trgm_idx 
+                ON endpoint_search_view USING gin (response_headers gin_trgm_ops);
+                
+                -- response_body 模糊搜索索引
+                CREATE INDEX IF NOT EXISTS endpoint_search_view_body_trgm_idx 
+                ON endpoint_search_view USING gin (response_body gin_trgm_ops);
+                
+                -- status_code 索引
+                CREATE INDEX IF NOT EXISTS endpoint_search_view_status_idx 
+                ON endpoint_search_view (status_code);
+                
+                -- created_at 排序索引
+                CREATE INDEX IF NOT EXISTS endpoint_search_view_created_idx 
+                ON endpoint_search_view (created_at DESC);
+            """,
+            reverse_sql="""
+                DROP INDEX IF EXISTS endpoint_search_view_id_idx;
+                DROP INDEX IF EXISTS endpoint_search_view_host_trgm_idx;
+                DROP INDEX IF EXISTS endpoint_search_view_title_trgm_idx;
+                DROP INDEX IF EXISTS endpoint_search_view_url_trgm_idx;
+                DROP INDEX IF EXISTS endpoint_search_view_headers_trgm_idx;
+                DROP INDEX IF EXISTS endpoint_search_view_body_trgm_idx;
+                DROP INDEX IF EXISTS endpoint_search_view_status_idx;
+                DROP INDEX IF EXISTS endpoint_search_view_created_idx;
+            """
+        ),
+    ]

backend/apps/asset/models/asset_models.py

@@ -1,6 +1,7 @@
 
 from django.db import models
 from django.contrib.postgres.fields import ArrayField
+from django.contrib.postgres.indexes import GinIndex
 from django.core.validators import MinValueValidator, MaxValueValidator
 
 
@@ -34,6 +35,12 @@ class Subdomain(models.Model):
             models.Index(fields=['name', 'target']),  # 复合索引，优化 get_by_names_and_target_id 批量查询
             models.Index(fields=['target']),     # 优化从target_id快速查找下面的子域名
             models.Index(fields=['name']),            # 优化从name快速查找子域名，搜索场景
+            # pg_trgm GIN 索引，支持 LIKE '%keyword%' 模糊搜索
+            GinIndex(
+                name='subdomain_name_trgm_idx',
+                fields=['name'],
+                opclasses=['gin_trgm_ops']
+            ),
         ]
         constraints = [
             # 普通唯一约束：name + target 组合唯一
@@ -58,40 +65,35 @@ class Endpoint(models.Model):
         help_text='所属的扫描目标（主关联字段，表示所属关系，不能为空）'
     )
     
-    url = models.CharField(max_length=2000, help_text='最终访问的完整URL')
+    url = models.TextField(help_text='最终访问的完整URL')
     host = models.CharField(
         max_length=253,
         blank=True,
         default='',
         help_text='主机名（域名或IP地址）'
     )
-    location = models.CharField(
-        max_length=1000,
+    location = models.TextField(
         blank=True,
         default='',
         help_text='重定向地址（HTTP 3xx 响应头 Location）'
     )
     created_at = models.DateTimeField(auto_now_add=True, help_text='创建时间')
-    title = models.CharField(
-        max_length=1000,
+    title = models.TextField(
         blank=True,
         default='',
         help_text='网页标题（HTML <title> 标签内容）'
     )
-    webserver = models.CharField(
-        max_length=200,
+    webserver = models.TextField(
         blank=True,
         default='',
         help_text='服务器类型（HTTP 响应头 Server 值）'
     )
-    body_preview = models.CharField(
-        max_length=1000,
+    response_body = models.TextField(
         blank=True,
         default='',
-        help_text='响应正文前N个字符（默认100个字符）'
+        help_text='HTTP响应体'
     )
-    content_type = models.CharField(
-        max_length=200,
+    content_type = models.TextField(
         blank=True,
         default='',
         help_text='响应类型（HTTP Content-Type 响应头）'
@@ -123,6 +125,11 @@ class Endpoint(models.Model):
         default=list,
         help_text='匹配的GF模式列表，用于识别敏感端点（如api, debug, config等）'
     )
+    response_headers = models.TextField(
+        blank=True,
+        default='',
+        help_text='原始HTTP响应头'
+    )
 
     class Meta:
         db_table = 'endpoint'
@@ -131,11 +138,28 @@ class Endpoint(models.Model):
         ordering = ['-created_at']
         indexes = [
             models.Index(fields=['-created_at']),
-            models.Index(fields=['target']),       # 优化从target_id快速查找下面的端点（主关联字段）
+            models.Index(fields=['target']),       # 优化从 target_id快速查找下面的端点（主关联字段）
             models.Index(fields=['url']),          # URL索引，优化查询性能
             models.Index(fields=['host']),         # host索引，优化根据主机名查询
             models.Index(fields=['status_code']),  # 状态码索引，优化筛选
             models.Index(fields=['title']),        # title索引，优化智能过滤搜索
+            GinIndex(fields=['tech']),             # GIN索引，优化 tech 数组字段的 __contains 查询
+            # pg_trgm GIN 索引，支持 LIKE '%keyword%' 模糊搜索
+            GinIndex(
+                name='endpoint_resp_headers_trgm_idx',
+                fields=['response_headers'],
+                opclasses=['gin_trgm_ops']
+            ),
+            GinIndex(
+                name='endpoint_url_trgm_idx',
+                fields=['url'],
+                opclasses=['gin_trgm_ops']
+            ),
+            GinIndex(
+                name='endpoint_title_trgm_idx',
+                fields=['title'],
+                opclasses=['gin_trgm_ops']
+            ),
         ]
         constraints = [
             # 普通唯一约束：url + target 组合唯一
@@ -160,40 +184,35 @@ class WebSite(models.Model):
         help_text='所属的扫描目标（主关联字段，表示所属关系，不能为空）'
     )
 
-    url = models.CharField(max_length=2000, help_text='最终访问的完整URL')
+    url = models.TextField(help_text='最终访问的完整URL')
     host = models.CharField(
         max_length=253,
         blank=True,
         default='',
         help_text='主机名（域名或IP地址）'
     )
-    location = models.CharField(
-        max_length=1000,
+    location = models.TextField(
         blank=True,
         default='',
         help_text='重定向地址（HTTP 3xx 响应头 Location）'
     )
     created_at = models.DateTimeField(auto_now_add=True, help_text='创建时间')
-    title = models.CharField(
-        max_length=1000,
+    title = models.TextField(
         blank=True,
         default='',
         help_text='网页标题（HTML <title> 标签内容）'
     )
-    webserver = models.CharField(
-        max_length=200,
+    webserver = models.TextField(
         blank=True,
         default='',
         help_text='服务器类型（HTTP 响应头 Server 值）'
     )
-    body_preview = models.CharField(
-        max_length=1000,
+    response_body = models.TextField(
         blank=True,
         default='',
-        help_text='响应正文前N个字符（默认100个字符）'
+        help_text='HTTP响应体'
     )
-    content_type = models.CharField(
-        max_length=200,
+    content_type = models.TextField(
         blank=True,
         default='',
         help_text='响应类型（HTTP Content-Type 响应头）'
@@ -219,6 +238,11 @@ class WebSite(models.Model):
         blank=True,
         help_text='是否支持虚拟主机'
     )
+    response_headers = models.TextField(
+        blank=True,
+        default='',
+        help_text='原始HTTP响应头'
+    )
 
     class Meta:
         db_table = 'website'
@@ -229,9 +253,26 @@ class WebSite(models.Model):
             models.Index(fields=['-created_at']),
             models.Index(fields=['url']),  # URL索引，优化查询性能
             models.Index(fields=['host']),  # host索引，优化根据主机名查询
-            models.Index(fields=['target']),     # 优化从target_id快速查找下面的站点
+            models.Index(fields=['target']),     # 优化从 target_id快速查找下面的站点
             models.Index(fields=['title']),      # title索引，优化智能过滤搜索
             models.Index(fields=['status_code']),  # 状态码索引，优化智能过滤搜索
+            GinIndex(fields=['tech']),  # GIN索引，优化 tech 数组字段的 __contains 查询
+            # pg_trgm GIN 索引，支持 LIKE '%keyword%' 模糊搜索
+            GinIndex(
+                name='website_resp_headers_trgm_idx',
+                fields=['response_headers'],
+                opclasses=['gin_trgm_ops']
+            ),
+            GinIndex(
+                name='website_url_trgm_idx',
+                fields=['url'],
+                opclasses=['gin_trgm_ops']
+            ),
+            GinIndex(
+                name='website_title_trgm_idx',
+                fields=['title'],
+                opclasses=['gin_trgm_ops']
+            ),
         ]
         constraints = [
             # 普通唯一约束：url + target 组合唯一
@@ -308,6 +349,12 @@ class Directory(models.Model):
             models.Index(fields=['target']),     # 优化从target_id快速查找下面的目录
             models.Index(fields=['url']),        # URL索引，优化搜索和唯一约束
             models.Index(fields=['status']),     # 状态码索引，优化筛选
+            # pg_trgm GIN 索引，支持 LIKE '%keyword%' 模糊搜索
+            GinIndex(
+                name='directory_url_trgm_idx',
+                fields=['url'],
+                opclasses=['gin_trgm_ops']
+            ),
         ]
         constraints = [
             # 普通唯一约束：target + url 组合唯一

backend/apps/asset/models/snapshot_models.py

@@ -1,5 +1,6 @@
 from django.db import models
 from django.contrib.postgres.fields import ArrayField
+from django.contrib.postgres.indexes import GinIndex
 from django.core.validators import MinValueValidator, MaxValueValidator
 
 
@@ -26,6 +27,12 @@ class SubdomainSnapshot(models.Model):
             models.Index(fields=['scan']),
             models.Index(fields=['name']),
             models.Index(fields=['-created_at']),
+            # pg_trgm GIN 索引，支持 LIKE '%keyword%' 模糊搜索
+            GinIndex(
+                name='subdomain_snap_name_trgm',
+                fields=['name'],
+                opclasses=['gin_trgm_ops']
+            ),
         ]
         constraints = [
             # 唯一约束：同一次扫描中，同一个子域名只能记录一次
@@ -54,22 +61,27 @@ class WebsiteSnapshot(models.Model):
     )
     
     # 扫描结果数据
-    url = models.CharField(max_length=2000, help_text='站点URL')
+    url = models.TextField(help_text='站点URL')
     host = models.CharField(max_length=253, blank=True, default='', help_text='主机名（域名或IP地址）')
-    title = models.CharField(max_length=500, blank=True, default='', help_text='页面标题')
-    status = models.IntegerField(null=True, blank=True, help_text='HTTP状态码')
+    title = models.TextField(blank=True, default='', help_text='页面标题')
+    status_code = models.IntegerField(null=True, blank=True, help_text='HTTP状态码')
     content_length = models.BigIntegerField(null=True, blank=True, help_text='内容长度')
-    location = models.CharField(max_length=1000, blank=True, default='', help_text='重定向位置')
-    web_server = models.CharField(max_length=200, blank=True, default='', help_text='Web服务器')
-    content_type = models.CharField(max_length=200, blank=True, default='', help_text='内容类型')
+    location = models.TextField(blank=True, default='', help_text='重定向位置')
+    webserver = models.TextField(blank=True, default='', help_text='Web服务器')
+    content_type = models.TextField(blank=True, default='', help_text='内容类型')
     tech = ArrayField(
         models.CharField(max_length=100),
         blank=True,
         default=list,
         help_text='技术栈'
     )
-    body_preview = models.TextField(blank=True, default='', help_text='响应体预览')
+    response_body = models.TextField(blank=True, default='', help_text='HTTP响应体')
     vhost = models.BooleanField(null=True, blank=True, help_text='虚拟主机标志')
+    response_headers = models.TextField(
+        blank=True,
+        default='',
+        help_text='原始HTTP响应头'
+    )
     created_at = models.DateTimeField(auto_now_add=True, help_text='创建时间')
 
     class Meta:
@@ -83,6 +95,23 @@ class WebsiteSnapshot(models.Model):
             models.Index(fields=['host']),  # host索引，优化根据主机名查询
             models.Index(fields=['title']),  # title索引，优化标题搜索
             models.Index(fields=['-created_at']),
+            GinIndex(fields=['tech']),  # GIN索引，优化数组字段查询
+            # pg_trgm GIN 索引，支持 LIKE '%keyword%' 模糊搜索
+            GinIndex(
+                name='ws_snap_resp_hdr_trgm',
+                fields=['response_headers'],
+                opclasses=['gin_trgm_ops']
+            ),
+            GinIndex(
+                name='ws_snap_url_trgm',
+                fields=['url'],
+                opclasses=['gin_trgm_ops']
+            ),
+            GinIndex(
+                name='ws_snap_title_trgm',
+                fields=['title'],
+                opclasses=['gin_trgm_ops']
+            ),
         ]
         constraints = [
             # 唯一约束：同一次扫描中，同一个URL只能记录一次
@@ -132,6 +161,12 @@ class DirectorySnapshot(models.Model):
             models.Index(fields=['status']),  # 状态码索引，优化筛选
             models.Index(fields=['content_type']),  # content_type索引，优化内容类型搜索
             models.Index(fields=['-created_at']),
+            # pg_trgm GIN 索引，支持 LIKE '%keyword%' 模糊搜索
+            GinIndex(
+                name='dir_snap_url_trgm',
+                fields=['url'],
+                opclasses=['gin_trgm_ops']
+            ),
         ]
         constraints = [
             # 唯一约束：同一次扫描中，同一个目录URL只能记录一次
@@ -232,26 +267,26 @@ class EndpointSnapshot(models.Model):
     )
     
     # 扫描结果数据
-    url = models.CharField(max_length=2000, help_text='端点URL')
+    url = models.TextField(help_text='端点URL')
     host = models.CharField(
         max_length=253,
         blank=True,
         default='',
         help_text='主机名（域名或IP地址）'
     )
-    title = models.CharField(max_length=1000, blank=True, default='', help_text='页面标题')
+    title = models.TextField(blank=True, default='', help_text='页面标题')
     status_code = models.IntegerField(null=True, blank=True, help_text='HTTP状态码')
     content_length = models.IntegerField(null=True, blank=True, help_text='内容长度')
-    location = models.CharField(max_length=1000, blank=True, default='', help_text='重定向位置')
-    webserver = models.CharField(max_length=200, blank=True, default='', help_text='Web服务器')
-    content_type = models.CharField(max_length=200, blank=True, default='', help_text='内容类型')
+    location = models.TextField(blank=True, default='', help_text='重定向位置')
+    webserver = models.TextField(blank=True, default='', help_text='Web服务器')
+    content_type = models.TextField(blank=True, default='', help_text='内容类型')
     tech = ArrayField(
         models.CharField(max_length=100),
         blank=True,
         default=list,
         help_text='技术栈'
     )
-    body_preview = models.CharField(max_length=1000, blank=True, default='', help_text='响应体预览')
+    response_body = models.TextField(blank=True, default='', help_text='HTTP响应体')
     vhost = models.BooleanField(null=True, blank=True, help_text='虚拟主机标志')
     matched_gf_patterns = ArrayField(
         models.CharField(max_length=100),
@@ -259,6 +294,11 @@ class EndpointSnapshot(models.Model):
         default=list,
         help_text='匹配的GF模式列表'
     )
+    response_headers = models.TextField(
+        blank=True,
+        default='',
+        help_text='原始HTTP响应头'
+    )
     created_at = models.DateTimeField(auto_now_add=True, help_text='创建时间')
 
     class Meta:
@@ -274,6 +314,23 @@ class EndpointSnapshot(models.Model):
             models.Index(fields=['status_code']),  # 状态码索引，优化筛选
             models.Index(fields=['webserver']),  # webserver索引，优化服务器搜索
             models.Index(fields=['-created_at']),
+            GinIndex(fields=['tech']),  # GIN索引，优化数组字段查询
+            # pg_trgm GIN 索引，支持 LIKE '%keyword%' 模糊搜索
+            GinIndex(
+                name='ep_snap_resp_hdr_trgm',
+                fields=['response_headers'],
+                opclasses=['gin_trgm_ops']
+            ),
+            GinIndex(
+                name='ep_snap_url_trgm',
+                fields=['url'],
+                opclasses=['gin_trgm_ops']
+            ),
+            GinIndex(
+                name='ep_snap_title_trgm',
+                fields=['title'],
+                opclasses=['gin_trgm_ops']
+            ),
         ]
         constraints = [
             # 唯一约束：同一次扫描中，同一个URL只能记录一次

backend/apps/asset/repositories/asset/endpoint_repository.py

@@ -48,12 +48,13 @@ class DjangoEndpointRepository:
                     status_code=item.status_code,
                     content_length=item.content_length,
                     webserver=item.webserver or '',
-                    body_preview=item.body_preview or '',
+                    response_body=item.response_body or '',
                     content_type=item.content_type or '',
                     tech=item.tech if item.tech else [],
                     vhost=item.vhost,
                     location=item.location or '',
-                    matched_gf_patterns=item.matched_gf_patterns if item.matched_gf_patterns else []
+                    matched_gf_patterns=item.matched_gf_patterns if item.matched_gf_patterns else [],
+                    response_headers=item.response_headers if item.response_headers else ''
                 )
                 for item in unique_items
             ]
@@ -65,8 +66,8 @@ class DjangoEndpointRepository:
                     unique_fields=['url', 'target'],
                     update_fields=[
                         'host', 'title', 'status_code', 'content_length',
-                        'webserver', 'body_preview', 'content_type', 'tech',
-                        'vhost', 'location', 'matched_gf_patterns'
+                        'webserver', 'response_body', 'content_type', 'tech',
+                        'vhost', 'location', 'matched_gf_patterns', 'response_headers'
                     ],
                     batch_size=1000
                 )
@@ -138,12 +139,13 @@ class DjangoEndpointRepository:
                     status_code=item.status_code,
                     content_length=item.content_length,
                     webserver=item.webserver or '',
-                    body_preview=item.body_preview or '',
+                    response_body=item.response_body or '',
                     content_type=item.content_type or '',
                     tech=item.tech if item.tech else [],
                     vhost=item.vhost,
                     location=item.location or '',
-                    matched_gf_patterns=item.matched_gf_patterns if item.matched_gf_patterns else []
+                    matched_gf_patterns=item.matched_gf_patterns if item.matched_gf_patterns else [],
+                    response_headers=item.response_headers if item.response_headers else ''
                 )
                 for item in unique_items
             ]
@@ -183,7 +185,7 @@ class DjangoEndpointRepository:
             .values(
                 'url', 'host', 'location', 'title', 'status_code',
                 'content_length', 'content_type', 'webserver', 'tech',
-                'body_preview', 'vhost', 'matched_gf_patterns', 'created_at'
+                'response_body', 'response_headers', 'vhost', 'matched_gf_patterns', 'created_at'
             )
             .order_by('url')
         )

backend/apps/asset/repositories/asset/website_repository.py

@@ -49,12 +49,13 @@ class DjangoWebSiteRepository:
                     location=item.location or '',
                     title=item.title or '',
                     webserver=item.webserver or '',
-                    body_preview=item.body_preview or '',
+                    response_body=item.response_body or '',
                     content_type=item.content_type or '',
                     tech=item.tech if item.tech else [],
                     status_code=item.status_code,
                     content_length=item.content_length,
-                    vhost=item.vhost
+                    vhost=item.vhost,
+                    response_headers=item.response_headers if item.response_headers else ''
                 )
                 for item in unique_items
             ]
@@ -66,8 +67,8 @@ class DjangoWebSiteRepository:
                     unique_fields=['url', 'target'],
                     update_fields=[
                         'host', 'location', 'title', 'webserver',
-                        'body_preview', 'content_type', 'tech',
-                        'status_code', 'content_length', 'vhost'
+                        'response_body', 'content_type', 'tech',
+                        'status_code', 'content_length', 'vhost', 'response_headers'
                     ],
                     batch_size=1000
                 )
@@ -132,12 +133,13 @@ class DjangoWebSiteRepository:
                     location=item.location or '',
                     title=item.title or '',
                     webserver=item.webserver or '',
-                    body_preview=item.body_preview or '',
+                    response_body=item.response_body or '',
                     content_type=item.content_type or '',
                     tech=item.tech if item.tech else [],
                     status_code=item.status_code,
                     content_length=item.content_length,
-                    vhost=item.vhost
+                    vhost=item.vhost,
+                    response_headers=item.response_headers if item.response_headers else ''
                 )
                 for item in unique_items
             ]
@@ -177,7 +179,7 @@ class DjangoWebSiteRepository:
             .values(
                 'url', 'host', 'location', 'title', 'status_code',
                 'content_length', 'content_type', 'webserver', 'tech',
-                'body_preview', 'vhost', 'created_at'
+                'response_body', 'response_headers', 'vhost', 'created_at'
             )
             .order_by('url')
         )

backend/apps/asset/repositories/snapshot/endpoint_snapshot_repository.py

@@ -44,6 +44,7 @@ class DjangoEndpointSnapshotRepository:
                 snapshots.append(EndpointSnapshot(
                     scan_id=item.scan_id,
                     url=item.url,
+                    host=item.host if item.host else '',
                     title=item.title,
                     status_code=item.status_code,
                     content_length=item.content_length,
@@ -51,9 +52,10 @@ class DjangoEndpointSnapshotRepository:
                     webserver=item.webserver,
                     content_type=item.content_type,
                     tech=item.tech if item.tech else [],
-                    body_preview=item.body_preview,
+                    response_body=item.response_body,
                     vhost=item.vhost,
-                    matched_gf_patterns=item.matched_gf_patterns if item.matched_gf_patterns else []
+                    matched_gf_patterns=item.matched_gf_patterns if item.matched_gf_patterns else [],
+                    response_headers=item.response_headers if item.response_headers else ''
                 ))
             
             # 批量创建（忽略冲突，基于唯一约束去重）
@@ -100,7 +102,7 @@ class DjangoEndpointSnapshotRepository:
             .values(
                 'url', 'host', 'location', 'title', 'status_code',
                 'content_length', 'content_type', 'webserver', 'tech',
-                'body_preview', 'vhost', 'matched_gf_patterns', 'created_at'
+                'response_body', 'response_headers', 'vhost', 'matched_gf_patterns', 'created_at'
             )
             .order_by('url')
         )

backend/apps/asset/repositories/snapshot/website_snapshot_repository.py

@@ -46,14 +46,15 @@ class DjangoWebsiteSnapshotRepository:
                     url=item.url,
                     host=item.host,
                     title=item.title,
-                    status=item.status,
+                    status_code=item.status_code,
                     content_length=item.content_length,
                     location=item.location,
-                    web_server=item.web_server,
+                    webserver=item.webserver,
                     content_type=item.content_type,
                     tech=item.tech if item.tech else [],
-                    body_preview=item.body_preview,
-                    vhost=item.vhost
+                    response_body=item.response_body,
+                    vhost=item.vhost,
+                    response_headers=item.response_headers if item.response_headers else ''
                 ))
             
             # 批量创建（忽略冲突，基于唯一约束去重）
@@ -98,26 +99,12 @@ class DjangoWebsiteSnapshotRepository:
             WebsiteSnapshot.objects
             .filter(scan_id=scan_id)
             .values(
-                'url', 'host', 'location', 'title', 'status',
-                'content_length', 'content_type', 'web_server', 'tech',
-                'body_preview', 'vhost', 'created_at'
+                'url', 'host', 'location', 'title', 'status_code',
+                'content_length', 'content_type', 'webserver', 'tech',
+                'response_body', 'response_headers', 'vhost', 'created_at'
             )
             .order_by('url')
         )
         
         for row in qs.iterator(chunk_size=batch_size):
-            # 重命名字段以匹配 CSV 表头
-            yield {
-                'url': row['url'],
-                'host': row['host'],
-                'location': row['location'],
-                'title': row['title'],
-                'status_code': row['status'],
-                'content_length': row['content_length'],
-                'content_type': row['content_type'],
-                'webserver': row['web_server'],
-                'tech': row['tech'],
-                'body_preview': row['body_preview'],
-                'vhost': row['vhost'],
-                'created_at': row['created_at'],
-            }
+            yield row

backend/apps/asset/serializers.py

@@ -67,9 +67,10 @@ class SubdomainListSerializer(serializers.ModelSerializer):
 
 
 class WebSiteSerializer(serializers.ModelSerializer):
-    """站点序列化器"""
+    """站点序列化器（目标详情页）"""
     
     subdomain = serializers.CharField(source='subdomain.name', allow_blank=True, default='')
+    responseHeaders = serializers.CharField(source='response_headers', read_only=True)  # 原始HTTP响应头
     
     class Meta:
         model = WebSite
@@ -83,9 +84,10 @@ class WebSiteSerializer(serializers.ModelSerializer):
             'content_type',
             'status_code',
             'content_length',
-            'body_preview',
+            'response_body',
             'tech',
             'vhost',
+            'responseHeaders',  # HTTP响应头
             'subdomain',
             'created_at',
         ]
@@ -140,6 +142,7 @@ class EndpointListSerializer(serializers.ModelSerializer):
         source='matched_gf_patterns',
         read_only=True,
     )
+    responseHeaders = serializers.CharField(source='response_headers', read_only=True)  # 原始HTTP响应头
 
     class Meta:
         model = Endpoint
@@ -152,9 +155,10 @@ class EndpointListSerializer(serializers.ModelSerializer):
             'content_length',
             'content_type',
             'webserver',
-            'body_preview',
+            'response_body',
             'tech',
             'vhost',
+            'responseHeaders',  # HTTP响应头
             'gfPatterns',
             'created_at',
         ]
@@ -213,8 +217,7 @@ class WebsiteSnapshotSerializer(serializers.ModelSerializer):
     """网站快照序列化器（用于扫描历史）"""
     
     subdomain_name = serializers.CharField(source='subdomain.name', read_only=True)
-    webserver = serializers.CharField(source='web_server', read_only=True)  # 映射字段名
-    status_code = serializers.IntegerField(source='status', read_only=True)  # 映射字段名
+    responseHeaders = serializers.CharField(source='response_headers', read_only=True)  # 原始HTTP响应头
     
     class Meta:
         model = WebsiteSnapshot
@@ -223,13 +226,14 @@ class WebsiteSnapshotSerializer(serializers.ModelSerializer):
             'url',
             'location',
             'title',
-            'webserver',  # 使用映射后的字段名
+            'webserver',
             'content_type',
-            'status_code',  # 使用映射后的字段名
+            'status_code',
             'content_length',
-            'body_preview',
+            'response_body',
             'tech',
             'vhost',
+            'responseHeaders',  # HTTP响应头
             'subdomain_name',
             'created_at',
         ]
@@ -264,6 +268,7 @@ class EndpointSnapshotSerializer(serializers.ModelSerializer):
         source='matched_gf_patterns',
         read_only=True,
     )
+    responseHeaders = serializers.CharField(source='response_headers', read_only=True)  # 原始HTTP响应头
 
     class Meta:
         model = EndpointSnapshot
@@ -277,9 +282,10 @@ class EndpointSnapshotSerializer(serializers.ModelSerializer):
             'content_type',
             'status_code',
             'content_length',
-            'body_preview',
+            'response_body',
             'tech',
             'vhost',
+            'responseHeaders',  # HTTP响应头
             'gfPatterns',
             'created_at',
         ]

backend/apps/asset/services/asset/endpoint_service.py

@@ -27,7 +27,8 @@ class EndpointService:
         'url': 'url',
         'host': 'host',
         'title': 'title',
-        'status': 'status_code',
+        'status_code': 'status_code',
+        'tech': 'tech',
     }
     
     def __init__(self):
@@ -115,7 +116,7 @@ class EndpointService:
         """获取目标下的所有端点"""
         queryset = self.repo.get_by_target(target_id)
         if filter_query:
-            queryset = apply_filters(queryset, filter_query, self.FILTER_FIELD_MAPPING)
+            queryset = apply_filters(queryset, filter_query, self.FILTER_FIELD_MAPPING, json_array_fields=['tech'])
         return queryset
     
     def count_endpoints_by_target(self, target_id: int) -> int:
@@ -134,7 +135,7 @@ class EndpointService:
         """获取所有端点（全局查询）"""
         queryset = self.repo.get_all()
         if filter_query:
-            queryset = apply_filters(queryset, filter_query, self.FILTER_FIELD_MAPPING)
+            queryset = apply_filters(queryset, filter_query, self.FILTER_FIELD_MAPPING, json_array_fields=['tech'])
         return queryset
     
     def iter_endpoint_urls_by_target(self, target_id: int, chunk_size: int = 1000) -> Iterator[str]:

backend/apps/asset/services/asset/website_service.py

@@ -19,7 +19,8 @@ class WebSiteService:
         'url': 'url',
         'host': 'host',
         'title': 'title',
-        'status': 'status_code',
+        'status_code': 'status_code',
+        'tech': 'tech',
     }
     
     def __init__(self, repository=None):
@@ -107,14 +108,14 @@ class WebSiteService:
         """获取目标下的所有网站"""
         queryset = self.repo.get_by_target(target_id)
         if filter_query:
-            queryset = apply_filters(queryset, filter_query, self.FILTER_FIELD_MAPPING)
+            queryset = apply_filters(queryset, filter_query, self.FILTER_FIELD_MAPPING, json_array_fields=['tech'])
         return queryset
     
     def get_all(self, filter_query: Optional[str] = None):
         """获取所有网站"""
         queryset = self.repo.get_all()
         if filter_query:
-            queryset = apply_filters(queryset, filter_query, self.FILTER_FIELD_MAPPING)
+            queryset = apply_filters(queryset, filter_query, self.FILTER_FIELD_MAPPING, json_array_fields=['tech'])
         return queryset
     
     def get_by_url(self, url: str, target_id: int) -> int:

backend/apps/asset/services/search_service.py

@@ -0,0 +1,477 @@
+"""
+资产搜索服务
+
+提供资产搜索的核心业务逻辑：
+- 从物化视图查询数据
+- 支持表达式语法解析
+- 支持 =（模糊）、==（精确）、!=（不等于）操作符
+- 支持 && (AND) 和 || (OR) 逻辑组合
+- 支持 Website 和 Endpoint 两种资产类型
+"""
+
+import logging
+import re
+from typing import Optional, List, Dict, Any, Tuple, Literal, Iterator
+
+from django.db import connection
+
+logger = logging.getLogger(__name__)
+
+# 支持的字段映射（前端字段名 -> 数据库字段名）
+FIELD_MAPPING = {
+    'host': 'host',
+    'url': 'url',
+    'title': 'title',
+    'tech': 'tech',
+    'status': 'status_code',
+    'body': 'response_body',
+    'header': 'response_headers',
+}
+
+# 数组类型字段
+ARRAY_FIELDS = {'tech'}
+
+# 资产类型到视图名的映射
+VIEW_MAPPING = {
+    'website': 'asset_search_view',
+    'endpoint': 'endpoint_search_view',
+}
+
+# 资产类型到原表名的映射（用于 JOIN 获取数组字段）
+# ⚠️ 重要：pg_ivm 不支持 ArrayField，所有数组字段必须从原表 JOIN 获取
+TABLE_MAPPING = {
+    'website': 'website',
+    'endpoint': 'endpoint',
+}
+
+# 有效的资产类型
+VALID_ASSET_TYPES = {'website', 'endpoint'}
+
+# Website 查询字段（v=视图，t=原表）
+# ⚠️ 注意：t.tech 从原表获取，因为 pg_ivm 不支持 ArrayField
+WEBSITE_SELECT_FIELDS = """
+    v.id,
+    v.url,
+    v.host,
+    v.title,
+    t.tech,  -- ArrayField，从 website 表 JOIN 获取
+    v.status_code,
+    v.response_headers,
+    v.response_body,
+    v.content_type,
+    v.content_length,
+    v.webserver,
+    v.location,
+    v.vhost,
+    v.created_at,
+    v.target_id
+"""
+
+# Endpoint 查询字段
+# ⚠️ 注意：t.tech 和 t.matched_gf_patterns 从原表获取，因为 pg_ivm 不支持 ArrayField
+ENDPOINT_SELECT_FIELDS = """
+    v.id,
+    v.url,
+    v.host,
+    v.title,
+    t.tech,  -- ArrayField，从 endpoint 表 JOIN 获取
+    v.status_code,
+    v.response_headers,
+    v.response_body,
+    v.content_type,
+    v.content_length,
+    v.webserver,
+    v.location,
+    v.vhost,
+    t.matched_gf_patterns,  -- ArrayField，从 endpoint 表 JOIN 获取
+    v.created_at,
+    v.target_id
+"""
+
+
+class SearchQueryParser:
+    """
+    搜索查询解析器
+    
+    支持语法：
+    - field="value"     模糊匹配（ILIKE %value%）
+    - field=="value"    精确匹配
+    - field!="value"    不等于
+    - &&                AND 连接
+    - ||                OR 连接
+    - ()                分组（暂不支持嵌套）
+    
+    示例：
+    - host="api" && tech="nginx"
+    - tech="vue" || tech="react"
+    - status=="200" && host!="test"
+    """
+    
+    # 匹配单个条件: field="value" 或 field=="value" 或 field!="value"
+    CONDITION_PATTERN = re.compile(r'(\w+)\s*(==|!=|=)\s*"([^"]*)"')
+    
+    @classmethod
+    def parse(cls, query: str) -> Tuple[str, List[Any]]:
+        """
+        解析查询字符串，返回 SQL WHERE 子句和参数
+        
+        Args:
+            query: 搜索查询字符串
+        
+        Returns:
+            (where_clause, params) 元组
+        """
+        if not query or not query.strip():
+            return "1=1", []
+        
+        query = query.strip()
+        
+        # 检查是否包含操作符语法，如果不包含则作为 host 模糊搜索
+        if not cls.CONDITION_PATTERN.search(query):
+            # 裸文本，默认作为 host 模糊搜索（v 是视图别名）
+            return "v.host ILIKE %s", [f"%{query}%"]
+        
+        # 按 || 分割为 OR 组
+        or_groups = cls._split_by_or(query)
+        
+        if len(or_groups) == 1:
+            # 没有 OR，直接解析 AND 条件
+            return cls._parse_and_group(or_groups[0])
+        
+        # 多个 OR 组
+        or_clauses = []
+        all_params = []
+        
+        for group in or_groups:
+            clause, params = cls._parse_and_group(group)
+            if clause and clause != "1=1":
+                or_clauses.append(f"({clause})")
+                all_params.extend(params)
+        
+        if not or_clauses:
+            return "1=1", []
+        
+        return " OR ".join(or_clauses), all_params
+    
+    @classmethod
+    def _split_by_or(cls, query: str) -> List[str]:
+        """按 || 分割查询，但忽略引号内的 ||"""
+        parts = []
+        current = ""
+        in_quotes = False
+        i = 0
+        
+        while i < len(query):
+            char = query[i]
+            
+            if char == '"':
+                in_quotes = not in_quotes
+                current += char
+            elif not in_quotes and i + 1 < len(query) and query[i:i+2] == '||':
+                if current.strip():
+                    parts.append(current.strip())
+                current = ""
+                i += 1  # 跳过第二个 |
+            else:
+                current += char
+            
+            i += 1
+        
+        if current.strip():
+            parts.append(current.strip())
+        
+        return parts if parts else [query]
+    
+    @classmethod
+    def _parse_and_group(cls, group: str) -> Tuple[str, List[Any]]:
+        """解析 AND 组（用 && 连接的条件）"""
+        # 移除外层括号
+        group = group.strip()
+        if group.startswith('(') and group.endswith(')'):
+            group = group[1:-1].strip()
+        
+        # 按 && 分割
+        parts = cls._split_by_and(group)
+        
+        and_clauses = []
+        all_params = []
+        
+        for part in parts:
+            clause, params = cls._parse_condition(part.strip())
+            if clause:
+                and_clauses.append(clause)
+                all_params.extend(params)
+        
+        if not and_clauses:
+            return "1=1", []
+        
+        return " AND ".join(and_clauses), all_params
+    
+    @classmethod
+    def _split_by_and(cls, query: str) -> List[str]:
+        """按 && 分割查询，但忽略引号内的 &&"""
+        parts = []
+        current = ""
+        in_quotes = False
+        i = 0
+        
+        while i < len(query):
+            char = query[i]
+            
+            if char == '"':
+                in_quotes = not in_quotes
+                current += char
+            elif not in_quotes and i + 1 < len(query) and query[i:i+2] == '&&':
+                if current.strip():
+                    parts.append(current.strip())
+                current = ""
+                i += 1  # 跳过第二个 &
+            else:
+                current += char
+            
+            i += 1
+        
+        if current.strip():
+            parts.append(current.strip())
+        
+        return parts if parts else [query]
+    
+    @classmethod
+    def _parse_condition(cls, condition: str) -> Tuple[Optional[str], List[Any]]:
+        """
+        解析单个条件
+        
+        Returns:
+            (sql_clause, params) 或 (None, []) 如果解析失败
+        """
+        # 移除括号
+        condition = condition.strip()
+        if condition.startswith('(') and condition.endswith(')'):
+            condition = condition[1:-1].strip()
+        
+        match = cls.CONDITION_PATTERN.match(condition)
+        if not match:
+            logger.warning(f"无法解析条件: {condition}")
+            return None, []
+        
+        field, operator, value = match.groups()
+        field = field.lower()
+        
+        # 验证字段
+        if field not in FIELD_MAPPING:
+            logger.warning(f"未知字段: {field}")
+            return None, []
+        
+        db_field = FIELD_MAPPING[field]
+        is_array = field in ARRAY_FIELDS
+        
+        # 根据操作符生成 SQL
+        if operator == '=':
+            # 模糊匹配
+            return cls._build_like_condition(db_field, value, is_array)
+        elif operator == '==':
+            # 精确匹配
+            return cls._build_exact_condition(db_field, value, is_array)
+        elif operator == '!=':
+            # 不等于
+            return cls._build_not_equal_condition(db_field, value, is_array)
+        
+        return None, []
+    
+    @classmethod
+    def _build_like_condition(cls, field: str, value: str, is_array: bool) -> Tuple[str, List[Any]]:
+        """构建模糊匹配条件"""
+        if is_array:
+            # 数组字段：检查数组中是否有元素包含该值（从原表 t 获取）
+            return f"EXISTS (SELECT 1 FROM unnest(t.{field}) AS elem WHERE elem ILIKE %s)", [f"%{value}%"]
+        elif field == 'status_code':
+            # 状态码是整数，模糊匹配转为精确匹配
+            try:
+                return f"v.{field} = %s", [int(value)]
+            except ValueError:
+                return f"v.{field}::text ILIKE %s", [f"%{value}%"]
+        else:
+            return f"v.{field} ILIKE %s", [f"%{value}%"]
+    
+    @classmethod
+    def _build_exact_condition(cls, field: str, value: str, is_array: bool) -> Tuple[str, List[Any]]:
+        """构建精确匹配条件"""
+        if is_array:
+            # 数组字段：检查数组中是否包含该精确值（从原表 t 获取）
+            return f"%s = ANY(t.{field})", [value]
+        elif field == 'status_code':
+            # 状态码是整数
+            try:
+                return f"v.{field} = %s", [int(value)]
+            except ValueError:
+                return f"v.{field}::text = %s", [value]
+        else:
+            return f"v.{field} = %s", [value]
+    
+    @classmethod
+    def _build_not_equal_condition(cls, field: str, value: str, is_array: bool) -> Tuple[str, List[Any]]:
+        """构建不等于条件"""
+        if is_array:
+            # 数组字段：检查数组中不包含该值（从原表 t 获取）
+            return f"NOT (%s = ANY(t.{field}))", [value]
+        elif field == 'status_code':
+            try:
+                return f"(v.{field} IS NULL OR v.{field} != %s)", [int(value)]
+            except ValueError:
+                return f"(v.{field} IS NULL OR v.{field}::text != %s)", [value]
+        else:
+            return f"(v.{field} IS NULL OR v.{field} != %s)", [value]
+
+
+AssetType = Literal['website', 'endpoint']
+
+
+class AssetSearchService:
+    """资产搜索服务"""
+    
+    def search(
+        self, 
+        query: str, 
+        asset_type: AssetType = 'website',
+        limit: Optional[int] = None
+    ) -> List[Dict[str, Any]]:
+        """
+        搜索资产
+        
+        Args:
+            query: 搜索查询字符串
+            asset_type: 资产类型 ('website' 或 'endpoint')
+            limit: 最大返回数量（可选）
+        
+        Returns:
+            List[Dict]: 搜索结果列表
+        """
+        where_clause, params = SearchQueryParser.parse(query)
+        
+        # 根据资产类型选择视图、原表和字段
+        view_name = VIEW_MAPPING.get(asset_type, 'asset_search_view')
+        table_name = TABLE_MAPPING.get(asset_type, 'website')
+        select_fields = ENDPOINT_SELECT_FIELDS if asset_type == 'endpoint' else WEBSITE_SELECT_FIELDS
+        
+        # JOIN 原表获取数组字段（tech, matched_gf_patterns）
+        sql = f"""
+            SELECT {select_fields}
+            FROM {view_name} v
+            JOIN {table_name} t ON v.id = t.id
+            WHERE {where_clause}
+            ORDER BY v.created_at DESC
+        """
+        
+        # 添加 LIMIT
+        if limit is not None and limit > 0:
+            sql += f" LIMIT {int(limit)}"
+        
+        try:
+            with connection.cursor() as cursor:
+                cursor.execute(sql, params)
+                columns = [col[0] for col in cursor.description]
+                results = []
+                
+                for row in cursor.fetchall():
+                    result = dict(zip(columns, row))
+                    results.append(result)
+                
+                return results
+        except Exception as e:
+            logger.error(f"搜索查询失败: {e}, SQL: {sql}, params: {params}")
+            raise
+    
+    def count(self, query: str, asset_type: AssetType = 'website', statement_timeout_ms: int = 300000) -> int:
+        """
+        统计搜索结果数量
+        
+        Args:
+            query: 搜索查询字符串
+            asset_type: 资产类型 ('website' 或 'endpoint')
+            statement_timeout_ms: SQL 语句超时时间（毫秒），默认 5 分钟
+        
+        Returns:
+            int: 结果总数
+        """
+        where_clause, params = SearchQueryParser.parse(query)
+        
+        # 根据资产类型选择视图和原表
+        view_name = VIEW_MAPPING.get(asset_type, 'asset_search_view')
+        table_name = TABLE_MAPPING.get(asset_type, 'website')
+        
+        # JOIN 原表以支持数组字段查询
+        sql = f"SELECT COUNT(*) FROM {view_name} v JOIN {table_name} t ON v.id = t.id WHERE {where_clause}"
+        
+        try:
+            with connection.cursor() as cursor:
+                # 为导出设置更长的超时时间（仅影响当前会话）
+                cursor.execute(f"SET LOCAL statement_timeout = {statement_timeout_ms}")
+                cursor.execute(sql, params)
+                return cursor.fetchone()[0]
+        except Exception as e:
+            logger.error(f"统计查询失败: {e}")
+            raise
+    
+    def search_iter(
+        self, 
+        query: str, 
+        asset_type: AssetType = 'website',
+        batch_size: int = 1000,
+        statement_timeout_ms: int = 300000
+    ) -> Iterator[Dict[str, Any]]:
+        """
+        流式搜索资产（使用分批查询，内存友好）
+        
+        Args:
+            query: 搜索查询字符串
+            asset_type: 资产类型 ('website' 或 'endpoint')
+            batch_size: 每批获取的数量
+            statement_timeout_ms: SQL 语句超时时间（毫秒），默认 5 分钟
+        
+        Yields:
+            Dict: 单条搜索结果
+        """
+        where_clause, params = SearchQueryParser.parse(query)
+        
+        # 根据资产类型选择视图、原表和字段
+        view_name = VIEW_MAPPING.get(asset_type, 'asset_search_view')
+        table_name = TABLE_MAPPING.get(asset_type, 'website')
+        select_fields = ENDPOINT_SELECT_FIELDS if asset_type == 'endpoint' else WEBSITE_SELECT_FIELDS
+        
+        # 使用 OFFSET/LIMIT 分批查询（Django 不支持命名游标）
+        offset = 0
+        
+        try:
+            while True:
+                # JOIN 原表获取数组字段
+                sql = f"""
+                    SELECT {select_fields}
+                    FROM {view_name} v
+                    JOIN {table_name} t ON v.id = t.id
+                    WHERE {where_clause}
+                    ORDER BY v.created_at DESC
+                    LIMIT {batch_size} OFFSET {offset}
+                """
+                
+                with connection.cursor() as cursor:
+                    # 为导出设置更长的超时时间（仅影响当前会话）
+                    cursor.execute(f"SET LOCAL statement_timeout = {statement_timeout_ms}")
+                    cursor.execute(sql, params)
+                    columns = [col[0] for col in cursor.description]
+                    rows = cursor.fetchall()
+                
+                if not rows:
+                    break
+                
+                for row in rows:
+                    yield dict(zip(columns, row))
+                
+                # 如果返回的行数少于 batch_size，说明已经是最后一批
+                if len(rows) < batch_size:
+                    break
+                
+                offset += batch_size
+                
+        except Exception as e:
+            logger.error(f"流式搜索查询失败: {e}, SQL: {sql}, params: {params}")
+            raise

backend/apps/asset/services/snapshot/endpoint_snapshots_service.py

@@ -72,7 +72,7 @@ class EndpointSnapshotsService:
         'url': 'url',
         'host': 'host',
         'title': 'title',
-        'status': 'status_code',
+        'status_code': 'status_code',
         'webserver': 'webserver',
         'tech': 'tech',
     }

backend/apps/asset/services/snapshot/website_snapshots_service.py

@@ -73,8 +73,8 @@ class WebsiteSnapshotsService:
         'url': 'url',
         'host': 'host',
         'title': 'title',
-        'status': 'status',
-        'webserver': 'web_server',
+        'status_code': 'status_code',
+        'webserver': 'webserver',
         'tech': 'tech',
     }
     

backend/apps/asset/urls.py

@@ -10,6 +10,8 @@ from .views import (
     DirectoryViewSet,
     VulnerabilityViewSet,
     AssetStatisticsViewSet,
+    AssetSearchView,
+    AssetSearchExportView,
 )
 
 # 创建 DRF 路由器
@@ -25,4 +27,6 @@ router.register(r'statistics', AssetStatisticsViewSet, basename='asset-statistic
 
 urlpatterns = [
     path('assets/', include(router.urls)),
+    path('assets/search/', AssetSearchView.as_view(), name='asset-search'),
+    path('assets/search/export/', AssetSearchExportView.as_view(), name='asset-search-export'),
 ]

backend/apps/asset/views/__init__.py

@@ -0,0 +1,40 @@
+"""
+Asset 应用视图模块
+
+重新导出所有视图类以保持向后兼容
+"""
+
+from .asset_views import (
+    AssetStatisticsViewSet,
+    SubdomainViewSet,
+    WebSiteViewSet,
+    DirectoryViewSet,
+    EndpointViewSet,
+    HostPortMappingViewSet,
+    VulnerabilityViewSet,
+    SubdomainSnapshotViewSet,
+    WebsiteSnapshotViewSet,
+    DirectorySnapshotViewSet,
+    EndpointSnapshotViewSet,
+    HostPortMappingSnapshotViewSet,
+    VulnerabilitySnapshotViewSet,
+)
+from .search_views import AssetSearchView, AssetSearchExportView
+
+__all__ = [
+    'AssetStatisticsViewSet',
+    'SubdomainViewSet',
+    'WebSiteViewSet',
+    'DirectoryViewSet',
+    'EndpointViewSet',
+    'HostPortMappingViewSet',
+    'VulnerabilityViewSet',
+    'SubdomainSnapshotViewSet',
+    'WebsiteSnapshotViewSet',
+    'DirectorySnapshotViewSet',
+    'EndpointSnapshotViewSet',
+    'HostPortMappingSnapshotViewSet',
+    'VulnerabilitySnapshotViewSet',
+    'AssetSearchView',
+    'AssetSearchExportView',
+]

backend/apps/asset/views/asset_views.py

@@ -8,19 +8,18 @@ from rest_framework.request import Request
 from rest_framework.exceptions import NotFound, ValidationError as DRFValidationError
 from django.core.exceptions import ValidationError, ObjectDoesNotExist
 from django.db import DatabaseError, IntegrityError, OperationalError
-from django.http import StreamingHttpResponse
 
-from .serializers import (
+from ..serializers import (
     SubdomainListSerializer, WebSiteSerializer, DirectorySerializer, 
     VulnerabilitySerializer, EndpointListSerializer, IPAddressAggregatedSerializer,
     SubdomainSnapshotSerializer, WebsiteSnapshotSerializer, DirectorySnapshotSerializer,
     EndpointSnapshotSerializer, VulnerabilitySnapshotSerializer
 )
-from .services import (
+from ..services import (
     SubdomainService, WebSiteService, DirectoryService, 
     VulnerabilityService, AssetStatisticsService, EndpointService, HostPortMappingService
 )
-from .services.snapshot import (
+from ..services.snapshot import (
     SubdomainSnapshotsService, WebsiteSnapshotsService, DirectorySnapshotsService,
     EndpointSnapshotsService, HostPortMappingSnapshotsService, VulnerabilitySnapshotsService
 )
@@ -243,7 +242,7 @@ class SubdomainViewSet(viewsets.ModelViewSet):
         
         CSV 列：name, created_at
         """
-        from apps.common.utils import generate_csv_rows, format_datetime
+        from apps.common.utils import create_csv_export_response, format_datetime
         
         target_pk = self.kwargs.get('target_pk')
         if not target_pk:
@@ -254,12 +253,12 @@ class SubdomainViewSet(viewsets.ModelViewSet):
         headers = ['name', 'created_at']
         formatters = {'created_at': format_datetime}
         
-        response = StreamingHttpResponse(
-            generate_csv_rows(data_iterator, headers, formatters),
-            content_type='text/csv; charset=utf-8'
+        return create_csv_export_response(
+            data_iterator=data_iterator,
+            headers=headers,
+            filename=f"target-{target_pk}-subdomains.csv",
+            field_formatters=formatters
         )
-        response['Content-Disposition'] = f'attachment; filename="target-{target_pk}-subdomains.csv"'
-        return response
 
 
 class WebSiteViewSet(viewsets.ModelViewSet):
@@ -274,6 +273,7 @@ class WebSiteViewSet(viewsets.ModelViewSet):
     - host="example"     主机名模糊匹配
     - title="login"      标题模糊匹配
     - status="200,301"   状态码多值匹配
+    - tech="nginx"       技术栈匹配（数组字段）
     - 多条件空格分隔     AND 关系
     """
     
@@ -366,9 +366,9 @@ class WebSiteViewSet(viewsets.ModelViewSet):
     def export(self, request, **kwargs):
         """导出网站为 CSV 格式
         
-        CSV 列：url, host, location, title, status_code, content_length, content_type, webserver, tech, body_preview, vhost, created_at
+        CSV 列：url, host, location, title, status_code, content_length, content_type, webserver, tech, response_body, response_headers, vhost, created_at
         """
-        from apps.common.utils import generate_csv_rows, format_datetime, format_list_field
+        from apps.common.utils import create_csv_export_response, format_datetime, format_list_field
         
         target_pk = self.kwargs.get('target_pk')
         if not target_pk:
@@ -379,19 +379,19 @@ class WebSiteViewSet(viewsets.ModelViewSet):
         headers = [
             'url', 'host', 'location', 'title', 'status_code',
             'content_length', 'content_type', 'webserver', 'tech',
-            'body_preview', 'vhost', 'created_at'
+            'response_body', 'response_headers', 'vhost', 'created_at'
         ]
         formatters = {
             'created_at': format_datetime,
             'tech': lambda x: format_list_field(x, separator=','),
         }
         
-        response = StreamingHttpResponse(
-            generate_csv_rows(data_iterator, headers, formatters),
-            content_type='text/csv; charset=utf-8'
+        return create_csv_export_response(
+            data_iterator=data_iterator,
+            headers=headers,
+            filename=f"target-{target_pk}-websites.csv",
+            field_formatters=formatters
         )
-        response['Content-Disposition'] = f'attachment; filename="target-{target_pk}-websites.csv"'
-        return response
 
 
 class DirectoryViewSet(viewsets.ModelViewSet):
@@ -498,7 +498,7 @@ class DirectoryViewSet(viewsets.ModelViewSet):
         
         CSV 列：url, status, content_length, words, lines, content_type, duration, created_at
         """
-        from apps.common.utils import generate_csv_rows, format_datetime
+        from apps.common.utils import create_csv_export_response, format_datetime
         
         target_pk = self.kwargs.get('target_pk')
         if not target_pk:
@@ -514,12 +514,12 @@ class DirectoryViewSet(viewsets.ModelViewSet):
             'created_at': format_datetime,
         }
         
-        response = StreamingHttpResponse(
-            generate_csv_rows(data_iterator, headers, formatters),
-            content_type='text/csv; charset=utf-8'
+        return create_csv_export_response(
+            data_iterator=data_iterator,
+            headers=headers,
+            filename=f"target-{target_pk}-directories.csv",
+            field_formatters=formatters
         )
-        response['Content-Disposition'] = f'attachment; filename="target-{target_pk}-directories.csv"'
-        return response
 
 
 class EndpointViewSet(viewsets.ModelViewSet):
@@ -534,6 +534,7 @@ class EndpointViewSet(viewsets.ModelViewSet):
     - host="example"     主机名模糊匹配
     - title="login"      标题模糊匹配
     - status="200,301"   状态码多值匹配
+    - tech="nginx"       技术栈匹配（数组字段）
     - 多条件空格分隔     AND 关系
     """
     
@@ -626,9 +627,9 @@ class EndpointViewSet(viewsets.ModelViewSet):
     def export(self, request, **kwargs):
         """导出端点为 CSV 格式
         
-        CSV 列：url, host, location, title, status_code, content_length, content_type, webserver, tech, body_preview, vhost, matched_gf_patterns, created_at
+        CSV 列：url, host, location, title, status_code, content_length, content_type, webserver, tech, response_body, response_headers, vhost, matched_gf_patterns, created_at
         """
-        from apps.common.utils import generate_csv_rows, format_datetime, format_list_field
+        from apps.common.utils import create_csv_export_response, format_datetime, format_list_field
         
         target_pk = self.kwargs.get('target_pk')
         if not target_pk:
@@ -639,7 +640,7 @@ class EndpointViewSet(viewsets.ModelViewSet):
         headers = [
             'url', 'host', 'location', 'title', 'status_code',
             'content_length', 'content_type', 'webserver', 'tech',
-            'body_preview', 'vhost', 'matched_gf_patterns', 'created_at'
+            'response_body', 'response_headers', 'vhost', 'matched_gf_patterns', 'created_at'
         ]
         formatters = {
             'created_at': format_datetime,
@@ -647,12 +648,12 @@ class EndpointViewSet(viewsets.ModelViewSet):
             'matched_gf_patterns': lambda x: format_list_field(x, separator=','),
         }
         
-        response = StreamingHttpResponse(
-            generate_csv_rows(data_iterator, headers, formatters),
-            content_type='text/csv; charset=utf-8'
+        return create_csv_export_response(
+            data_iterator=data_iterator,
+            headers=headers,
+            filename=f"target-{target_pk}-endpoints.csv",
+            field_formatters=formatters
         )
-        response['Content-Disposition'] = f'attachment; filename="target-{target_pk}-endpoints.csv"'
-        return response
 
 
 class HostPortMappingViewSet(viewsets.ModelViewSet):
@@ -705,7 +706,7 @@ class HostPortMappingViewSet(viewsets.ModelViewSet):
         
         CSV 列：ip, host, port, created_at
         """
-        from apps.common.utils import generate_csv_rows, format_datetime
+        from apps.common.utils import create_csv_export_response, format_datetime
         
         target_pk = self.kwargs.get('target_pk')
         if not target_pk:
@@ -720,14 +721,12 @@ class HostPortMappingViewSet(viewsets.ModelViewSet):
             'created_at': format_datetime
         }
         
-        # 生成流式响应
-        response = StreamingHttpResponse(
-            generate_csv_rows(data_iterator, headers, formatters),
-            content_type='text/csv; charset=utf-8'
+        return create_csv_export_response(
+            data_iterator=data_iterator,
+            headers=headers,
+            filename=f"target-{target_pk}-ip-addresses.csv",
+            field_formatters=formatters
         )
-        response['Content-Disposition'] = f'attachment; filename="target-{target_pk}-ip-addresses.csv"'
-        
-        return response
 
 
 class VulnerabilityViewSet(viewsets.ModelViewSet):
@@ -799,7 +798,7 @@ class SubdomainSnapshotViewSet(viewsets.ModelViewSet):
         
         CSV 列：name, created_at
         """
-        from apps.common.utils import generate_csv_rows, format_datetime
+        from apps.common.utils import create_csv_export_response, format_datetime
         
         scan_pk = self.kwargs.get('scan_pk')
         if not scan_pk:
@@ -810,12 +809,12 @@ class SubdomainSnapshotViewSet(viewsets.ModelViewSet):
         headers = ['name', 'created_at']
         formatters = {'created_at': format_datetime}
         
-        response = StreamingHttpResponse(
-            generate_csv_rows(data_iterator, headers, formatters),
-            content_type='text/csv; charset=utf-8'
+        return create_csv_export_response(
+            data_iterator=data_iterator,
+            headers=headers,
+            filename=f"scan-{scan_pk}-subdomains.csv",
+            field_formatters=formatters
         )
-        response['Content-Disposition'] = f'attachment; filename="scan-{scan_pk}-subdomains.csv"'
-        return response
 
 
 class WebsiteSnapshotViewSet(viewsets.ModelViewSet):
@@ -851,9 +850,9 @@ class WebsiteSnapshotViewSet(viewsets.ModelViewSet):
     def export(self, request, **kwargs):
         """导出网站快照为 CSV 格式
         
-        CSV 列：url, host, location, title, status_code, content_length, content_type, webserver, tech, body_preview, vhost, created_at
+        CSV 列：url, host, location, title, status_code, content_length, content_type, webserver, tech, response_body, response_headers, vhost, created_at
         """
-        from apps.common.utils import generate_csv_rows, format_datetime, format_list_field
+        from apps.common.utils import create_csv_export_response, format_datetime, format_list_field
         
         scan_pk = self.kwargs.get('scan_pk')
         if not scan_pk:
@@ -864,19 +863,19 @@ class WebsiteSnapshotViewSet(viewsets.ModelViewSet):
         headers = [
             'url', 'host', 'location', 'title', 'status_code',
             'content_length', 'content_type', 'webserver', 'tech',
-            'body_preview', 'vhost', 'created_at'
+            'response_body', 'response_headers', 'vhost', 'created_at'
         ]
         formatters = {
             'created_at': format_datetime,
             'tech': lambda x: format_list_field(x, separator=','),
         }
         
-        response = StreamingHttpResponse(
-            generate_csv_rows(data_iterator, headers, formatters),
-            content_type='text/csv; charset=utf-8'
+        return create_csv_export_response(
+            data_iterator=data_iterator,
+            headers=headers,
+            filename=f"scan-{scan_pk}-websites.csv",
+            field_formatters=formatters
         )
-        response['Content-Disposition'] = f'attachment; filename="scan-{scan_pk}-websites.csv"'
-        return response
 
 
 class DirectorySnapshotViewSet(viewsets.ModelViewSet):
@@ -911,7 +910,7 @@ class DirectorySnapshotViewSet(viewsets.ModelViewSet):
         
         CSV 列：url, status, content_length, words, lines, content_type, duration, created_at
         """
-        from apps.common.utils import generate_csv_rows, format_datetime
+        from apps.common.utils import create_csv_export_response, format_datetime
         
         scan_pk = self.kwargs.get('scan_pk')
         if not scan_pk:
@@ -927,12 +926,12 @@ class DirectorySnapshotViewSet(viewsets.ModelViewSet):
             'created_at': format_datetime,
         }
         
-        response = StreamingHttpResponse(
-            generate_csv_rows(data_iterator, headers, formatters),
-            content_type='text/csv; charset=utf-8'
+        return create_csv_export_response(
+            data_iterator=data_iterator,
+            headers=headers,
+            filename=f"scan-{scan_pk}-directories.csv",
+            field_formatters=formatters
         )
-        response['Content-Disposition'] = f'attachment; filename="scan-{scan_pk}-directories.csv"'
-        return response
 
 
 class EndpointSnapshotViewSet(viewsets.ModelViewSet):
@@ -968,9 +967,9 @@ class EndpointSnapshotViewSet(viewsets.ModelViewSet):
     def export(self, request, **kwargs):
         """导出端点快照为 CSV 格式
         
-        CSV 列：url, host, location, title, status_code, content_length, content_type, webserver, tech, body_preview, vhost, matched_gf_patterns, created_at
+        CSV 列：url, host, location, title, status_code, content_length, content_type, webserver, tech, response_body, response_headers, vhost, matched_gf_patterns, created_at
         """
-        from apps.common.utils import generate_csv_rows, format_datetime, format_list_field
+        from apps.common.utils import create_csv_export_response, format_datetime, format_list_field
         
         scan_pk = self.kwargs.get('scan_pk')
         if not scan_pk:
@@ -981,7 +980,7 @@ class EndpointSnapshotViewSet(viewsets.ModelViewSet):
         headers = [
             'url', 'host', 'location', 'title', 'status_code',
             'content_length', 'content_type', 'webserver', 'tech',
-            'body_preview', 'vhost', 'matched_gf_patterns', 'created_at'
+            'response_body', 'response_headers', 'vhost', 'matched_gf_patterns', 'created_at'
         ]
         formatters = {
             'created_at': format_datetime,
@@ -989,12 +988,12 @@ class EndpointSnapshotViewSet(viewsets.ModelViewSet):
             'matched_gf_patterns': lambda x: format_list_field(x, separator=','),
         }
         
-        response = StreamingHttpResponse(
-            generate_csv_rows(data_iterator, headers, formatters),
-            content_type='text/csv; charset=utf-8'
+        return create_csv_export_response(
+            data_iterator=data_iterator,
+            headers=headers,
+            filename=f"scan-{scan_pk}-endpoints.csv",
+            field_formatters=formatters
         )
-        response['Content-Disposition'] = f'attachment; filename="scan-{scan_pk}-endpoints.csv"'
-        return response
 
 
 class HostPortMappingSnapshotViewSet(viewsets.ModelViewSet):
@@ -1029,7 +1028,7 @@ class HostPortMappingSnapshotViewSet(viewsets.ModelViewSet):
         
         CSV 列：ip, host, port, created_at
         """
-        from apps.common.utils import generate_csv_rows, format_datetime
+        from apps.common.utils import create_csv_export_response, format_datetime
         
         scan_pk = self.kwargs.get('scan_pk')
         if not scan_pk:
@@ -1044,14 +1043,12 @@ class HostPortMappingSnapshotViewSet(viewsets.ModelViewSet):
             'created_at': format_datetime
         }
         
-        # 生成流式响应
-        response = StreamingHttpResponse(
-            generate_csv_rows(data_iterator, headers, formatters),
-            content_type='text/csv; charset=utf-8'
+        return create_csv_export_response(
+            data_iterator=data_iterator,
+            headers=headers,
+            filename=f"scan-{scan_pk}-ip-addresses.csv",
+            field_formatters=formatters
         )
-        response['Content-Disposition'] = f'attachment; filename="scan-{scan_pk}-ip-addresses.csv"'
-        
-        return response
 
 
 class VulnerabilitySnapshotViewSet(viewsets.ModelViewSet):

backend/apps/asset/views/search_views.py

@@ -0,0 +1,361 @@
+"""
+资产搜索 API 视图
+
+提供资产搜索的 REST API 接口：
+- GET /api/assets/search/ - 搜索资产
+- GET /api/assets/search/export/ - 导出搜索结果为 CSV
+
+搜索语法：
+- field="value"     模糊匹配（ILIKE %value%）
+- field=="value"    精确匹配
+- field!="value"    不等于
+- &&                AND 连接
+- ||                OR 连接
+
+支持的字段：
+- host: 主机名
+- url: URL
+- title: 标题
+- tech: 技术栈
+- status: 状态码
+- body: 响应体
+- header: 响应头
+
+支持的资产类型：
+- website: 站点（默认）
+- endpoint: 端点
+"""
+
+import logging
+import json
+from datetime import datetime
+from urllib.parse import urlparse, urlunparse
+from rest_framework import status
+from rest_framework.views import APIView
+from rest_framework.request import Request
+from django.db import connection
+
+from apps.common.response_helpers import success_response, error_response
+from apps.common.error_codes import ErrorCodes
+from apps.asset.services.search_service import AssetSearchService, VALID_ASSET_TYPES
+
+logger = logging.getLogger(__name__)
+
+
+class AssetSearchView(APIView):
+    """
+    资产搜索 API
+    
+    GET /api/assets/search/
+    
+    Query Parameters:
+        q: 搜索查询表达式
+        asset_type: 资产类型 ('website' 或 'endpoint'，默认 'website')
+        page: 页码（从 1 开始，默认 1）
+        pageSize: 每页数量（默认 10，最大 100）
+    
+    示例查询：
+        ?q=host="api" && tech="nginx"
+        ?q=tech="vue" || tech="react"&asset_type=endpoint
+        ?q=status=="200" && host!="test"
+    
+    Response:
+        {
+            "results": [...],
+            "total": 100,
+            "page": 1,
+            "pageSize": 10,
+            "totalPages": 10,
+            "assetType": "website"
+        }
+    """
+    
+    def __init__(self, **kwargs):
+        super().__init__(**kwargs)
+        self.service = AssetSearchService()
+    
+    def _parse_headers(self, headers_data) -> dict:
+        """解析响应头为字典"""
+        if not headers_data:
+            return {}
+        try:
+            return json.loads(headers_data)
+        except (json.JSONDecodeError, TypeError):
+            result = {}
+            for line in str(headers_data).split('\n'):
+                if ':' in line:
+                    key, value = line.split(':', 1)
+                    result[key.strip()] = value.strip()
+            return result
+    
+    def _format_result(self, result: dict, vulnerabilities_by_url: dict, asset_type: str) -> dict:
+        """格式化单个搜索结果"""
+        url = result.get('url', '')
+        vulns = vulnerabilities_by_url.get(url, [])
+        
+        # 基础字段（Website 和 Endpoint 共有）
+        formatted = {
+            'id': result.get('id'),
+            'url': url,
+            'host': result.get('host', ''),
+            'title': result.get('title', ''),
+            'technologies': result.get('tech', []) or [],
+            'statusCode': result.get('status_code'),
+            'contentLength': result.get('content_length'),
+            'contentType': result.get('content_type', ''),
+            'webserver': result.get('webserver', ''),
+            'location': result.get('location', ''),
+            'vhost': result.get('vhost'),
+            'responseHeaders': self._parse_headers(result.get('response_headers')),
+            'responseBody': result.get('response_body', ''),
+            'createdAt': result.get('created_at').isoformat() if result.get('created_at') else None,
+            'targetId': result.get('target_id'),
+        }
+        
+        # Website 特有字段：漏洞关联
+        if asset_type == 'website':
+            formatted['vulnerabilities'] = [
+                {
+                    'id': v.get('id'),
+                    'name': v.get('vuln_type', ''),
+                    'vulnType': v.get('vuln_type', ''),
+                    'severity': v.get('severity', 'info'),
+                }
+                for v in vulns
+            ]
+        
+        # Endpoint 特有字段
+        if asset_type == 'endpoint':
+            formatted['matchedGfPatterns'] = result.get('matched_gf_patterns', []) or []
+        
+        return formatted
+    
+    def _get_vulnerabilities_by_url_prefix(self, website_urls: list) -> dict:
+        """
+        根据 URL 前缀批量查询漏洞数据
+        
+        漏洞 URL 是 website URL 的子路径，使用前缀匹配：
+        - website.url: https://example.com/path?query=1
+        - vulnerability.url: https://example.com/path/api/users
+        
+        Args:
+            website_urls: website URL 列表，格式为 [(url, target_id), ...]
+        
+        Returns:
+            dict: {website_url: [vulnerability_list]}
+        """
+        if not website_urls:
+            return {}
+        
+        try:
+            with connection.cursor() as cursor:
+                # 构建 OR 条件：每个 website URL（去掉查询参数）作为前缀匹配
+                conditions = []
+                params = []
+                url_mapping = {}  # base_url -> original_url
+                
+                for url, target_id in website_urls:
+                    if not url or target_id is None:
+                        continue
+                    # 使用 urlparse 去掉查询参数和片段，只保留 scheme://netloc/path
+                    parsed = urlparse(url)
+                    base_url = urlunparse((parsed.scheme, parsed.netloc, parsed.path, '', '', ''))
+                    url_mapping[base_url] = url
+                    conditions.append("(v.url LIKE %s AND v.target_id = %s)")
+                    params.extend([base_url + '%', target_id])
+                
+                if not conditions:
+                    return {}
+                
+                where_clause = " OR ".join(conditions)
+                
+                sql = f"""
+                    SELECT v.id, v.vuln_type, v.severity, v.url, v.target_id
+                    FROM vulnerability v
+                    WHERE {where_clause}
+                    ORDER BY 
+                        CASE v.severity 
+                            WHEN 'critical' THEN 1 
+                            WHEN 'high' THEN 2 
+                            WHEN 'medium' THEN 3 
+                            WHEN 'low' THEN 4 
+                            ELSE 5 
+                        END
+                """
+                cursor.execute(sql, params)
+                
+                # 获取所有漏洞
+                all_vulns = []
+                for row in cursor.fetchall():
+                    all_vulns.append({
+                        'id': row[0],
+                        'vuln_type': row[1],
+                        'name': row[1],
+                        'severity': row[2],
+                        'url': row[3],
+                        'target_id': row[4],
+                    })
+                
+                # 按原始 website URL 分组（用于返回结果）
+                result = {url: [] for url, _ in website_urls}
+                for vuln in all_vulns:
+                    vuln_url = vuln['url']
+                    # 找到匹配的 website URL（最长前缀匹配）
+                    for website_url, target_id in website_urls:
+                        parsed = urlparse(website_url)
+                        base_url = urlunparse((parsed.scheme, parsed.netloc, parsed.path, '', '', ''))
+                        if vuln_url.startswith(base_url) and vuln['target_id'] == target_id:
+                            result[website_url].append(vuln)
+                            break
+                
+                return result
+        except Exception as e:
+            logger.error(f"批量查询漏洞失败: {e}")
+            return {}
+    
+    def get(self, request: Request):
+        """搜索资产"""
+        # 获取搜索查询
+        query = request.query_params.get('q', '').strip()
+        
+        if not query:
+            return error_response(
+                code=ErrorCodes.VALIDATION_ERROR,
+                message='Search query (q) is required',
+                status_code=status.HTTP_400_BAD_REQUEST
+            )
+        
+        # 获取并验证资产类型
+        asset_type = request.query_params.get('asset_type', 'website').strip().lower()
+        if asset_type not in VALID_ASSET_TYPES:
+            return error_response(
+                code=ErrorCodes.VALIDATION_ERROR,
+                message=f'Invalid asset_type. Must be one of: {", ".join(VALID_ASSET_TYPES)}',
+                status_code=status.HTTP_400_BAD_REQUEST
+            )
+        
+        # 获取分页参数
+        try:
+            page = int(request.query_params.get('page', 1))
+            page_size = int(request.query_params.get('pageSize', 10))
+        except (ValueError, TypeError):
+            page = 1
+            page_size = 10
+        
+        # 限制分页参数
+        page = max(1, page)
+        page_size = min(max(1, page_size), 100)
+        
+        # 获取总数和搜索结果
+        total = self.service.count(query, asset_type)
+        total_pages = (total + page_size - 1) // page_size if total > 0 else 1
+        offset = (page - 1) * page_size
+        
+        all_results = self.service.search(query, asset_type)
+        results = all_results[offset:offset + page_size]
+        
+        # 批量查询漏洞数据（仅 Website 类型需要）
+        vulnerabilities_by_url = {}
+        if asset_type == 'website':
+            website_urls = [(r.get('url'), r.get('target_id')) for r in results if r.get('url') and r.get('target_id')]
+            vulnerabilities_by_url = self._get_vulnerabilities_by_url_prefix(website_urls) if website_urls else {}
+        
+        # 格式化结果
+        formatted_results = [self._format_result(r, vulnerabilities_by_url, asset_type) for r in results]
+        
+        return success_response(data={
+            'results': formatted_results,
+            'total': total,
+            'page': page,
+            'pageSize': page_size,
+            'totalPages': total_pages,
+            'assetType': asset_type,
+        })
+
+
+class AssetSearchExportView(APIView):
+    """
+    资产搜索导出 API
+    
+    GET /api/assets/search/export/
+    
+    Query Parameters:
+        q: 搜索查询表达式
+        asset_type: 资产类型 ('website' 或 'endpoint'，默认 'website')
+    
+    Response:
+        CSV 文件（带 Content-Length，支持浏览器显示下载进度）
+    """
+    
+    def __init__(self, **kwargs):
+        super().__init__(**kwargs)
+        self.service = AssetSearchService()
+    
+    def _get_headers_and_formatters(self, asset_type: str):
+        """获取 CSV 表头和格式化器"""
+        from apps.common.utils import format_datetime, format_list_field
+        
+        if asset_type == 'website':
+            headers = ['url', 'host', 'title', 'status_code', 'content_type', 'content_length', 
+                      'webserver', 'location', 'tech', 'vhost', 'created_at']
+        else:
+            headers = ['url', 'host', 'title', 'status_code', 'content_type', 'content_length',
+                      'webserver', 'location', 'tech', 'matched_gf_patterns', 'vhost', 'created_at']
+        
+        formatters = {
+            'created_at': format_datetime,
+            'tech': lambda x: format_list_field(x, separator='; '),
+            'matched_gf_patterns': lambda x: format_list_field(x, separator='; '),
+            'vhost': lambda x: 'true' if x else ('false' if x is False else ''),
+        }
+        
+        return headers, formatters
+    
+    def get(self, request: Request):
+        """导出搜索结果为 CSV（带 Content-Length，支持下载进度显示）"""
+        from apps.common.utils import create_csv_export_response
+        
+        # 获取搜索查询
+        query = request.query_params.get('q', '').strip()
+        
+        if not query:
+            return error_response(
+                code=ErrorCodes.VALIDATION_ERROR,
+                message='Search query (q) is required',
+                status_code=status.HTTP_400_BAD_REQUEST
+            )
+        
+        # 获取并验证资产类型
+        asset_type = request.query_params.get('asset_type', 'website').strip().lower()
+        if asset_type not in VALID_ASSET_TYPES:
+            return error_response(
+                code=ErrorCodes.VALIDATION_ERROR,
+                message=f'Invalid asset_type. Must be one of: {", ".join(VALID_ASSET_TYPES)}',
+                status_code=status.HTTP_400_BAD_REQUEST
+            )
+        
+        # 检查是否有结果（快速检查，避免空导出）
+        total = self.service.count(query, asset_type)
+        if total == 0:
+            return error_response(
+                code=ErrorCodes.NOT_FOUND,
+                message='No results to export',
+                status_code=status.HTTP_404_NOT_FOUND
+            )
+        
+        # 获取表头和格式化器
+        headers, formatters = self._get_headers_and_formatters(asset_type)
+        
+        # 生成文件名
+        timestamp = datetime.now().strftime('%Y%m%d_%H%M%S')
+        filename = f'search_{asset_type}_{timestamp}.csv'
+        
+        # 使用通用导出工具
+        data_iterator = self.service.search_iter(query, asset_type)
+        return create_csv_export_response(
+            data_iterator=data_iterator,
+            headers=headers,
+            filename=filename,
+            field_formatters=formatters,
+            show_progress=True  # 显示下载进度
+        )

backend/apps/common/container_bootstrap.py

@@ -40,8 +40,14 @@ def fetch_config_and_setup_django():
     print(f"[CONFIG] 正在从配置中心获取配置: {config_url}")
     print(f"[CONFIG] IS_LOCAL={is_local}")
     try:
+        # 构建请求头（包含 Worker API Key）
+        headers = {}
+        worker_api_key = os.environ.get("WORKER_API_KEY", "")
+        if worker_api_key:
+            headers["X-Worker-API-Key"] = worker_api_key
+        
         # verify=False: 远程 Worker 通过 HTTPS 访问时可能使用自签名证书
-        resp = requests.get(config_url, timeout=10, verify=False)
+        resp = requests.get(config_url, headers=headers, timeout=10, verify=False)
         resp.raise_for_status()
         config = resp.json()
         
@@ -57,9 +63,6 @@ def fetch_config_and_setup_django():
         os.environ.setdefault("DB_USER", db_user)
         os.environ.setdefault("DB_PASSWORD", config['db']['password'])
         
-        # Redis 配置
-        os.environ.setdefault("REDIS_URL", config['redisUrl'])
-        
         # 日志配置
         os.environ.setdefault("LOG_DIR", config['paths']['logs'])
         os.environ.setdefault("LOG_LEVEL", config['logging']['level'])
@@ -71,7 +74,6 @@ def fetch_config_and_setup_django():
         print(f"[CONFIG]   DB_PORT: {db_port}")
         print(f"[CONFIG]   DB_NAME: {db_name}")
         print(f"[CONFIG]   DB_USER: {db_user}")
-        print(f"[CONFIG]   REDIS_URL: {config['redisUrl']}")
         
     except Exception as e:
         print(f"[ERROR] 获取配置失败: {config_url} - {e}", file=sys.stderr)

backend/apps/common/exception_handlers.py

@@ -0,0 +1,49 @@
+"""
+自定义异常处理器
+
+统一处理 DRF 异常，确保错误响应格式一致
+"""
+
+from rest_framework.views import exception_handler
+from rest_framework import status
+from rest_framework.exceptions import AuthenticationFailed, NotAuthenticated
+
+from apps.common.response_helpers import error_response
+from apps.common.error_codes import ErrorCodes
+
+
+def custom_exception_handler(exc, context):
+    """
+    自定义异常处理器
+    
+    处理认证相关异常，返回统一格式的错误响应
+    """
+    # 先调用 DRF 默认的异常处理器
+    response = exception_handler(exc, context)
+    
+    if response is not None:
+        # 处理 401 未认证错误
+        if response.status_code == status.HTTP_401_UNAUTHORIZED:
+            return error_response(
+                code=ErrorCodes.UNAUTHORIZED,
+                message='Authentication required',
+                status_code=status.HTTP_401_UNAUTHORIZED
+            )
+        
+        # 处理 403 权限不足错误
+        if response.status_code == status.HTTP_403_FORBIDDEN:
+            return error_response(
+                code=ErrorCodes.PERMISSION_DENIED,
+                message='Permission denied',
+                status_code=status.HTTP_403_FORBIDDEN
+            )
+    
+    # 处理 NotAuthenticated 和 AuthenticationFailed 异常
+    if isinstance(exc, (NotAuthenticated, AuthenticationFailed)):
+        return error_response(
+            code=ErrorCodes.UNAUTHORIZED,
+            message='Authentication required',
+            status_code=status.HTTP_401_UNAUTHORIZED
+        )
+    
+    return response

backend/apps/common/permissions.py

@@ -0,0 +1,80 @@
+"""
+集中式权限管理
+
+实现三类端点的认证逻辑：
+1. 公开端点（无需认证）：登录、登出、获取当前用户状态
+2. Worker 端点（API Key 认证）：注册、配置、心跳、回调、资源同步
+3. 业务端点（Session 认证）：其他所有 API
+"""
+
+import re
+import logging
+from django.conf import settings
+from rest_framework.permissions import BasePermission
+
+logger = logging.getLogger(__name__)
+
+# 公开端点白名单（无需任何认证）
+PUBLIC_ENDPOINTS = [
+    r'^/api/auth/login/$',
+    r'^/api/auth/logout/$',
+    r'^/api/auth/me/$',
+]
+
+# Worker API 端点（需要 API Key 认证）
+# 包括：注册、配置、心跳、回调、资源同步（字典下载）
+WORKER_ENDPOINTS = [
+    r'^/api/workers/register/$',
+    r'^/api/workers/config/$',
+    r'^/api/workers/\d+/heartbeat/$',
+    r'^/api/callbacks/',
+    # 资源同步端点（Worker 需要下载字典文件）
+    r'^/api/wordlists/download/$',
+    # 注意：指纹导出 API 使用 Session 认证（前端用户导出用）
+    # Worker 通过数据库直接获取指纹数据，不需要 HTTP API
+]
+
+
+class IsAuthenticatedOrPublic(BasePermission):
+    """
+    自定义权限类：
+    - 白名单内的端点公开访问
+    - Worker 端点需要 API Key 认证
+    - 其他端点需要 Session 认证
+    """
+    
+    def has_permission(self, request, view):
+        path = request.path
+        
+        # 检查是否在公开白名单内
+        for pattern in PUBLIC_ENDPOINTS:
+            if re.match(pattern, path):
+                return True
+        
+        # 检查是否是 Worker 端点
+        for pattern in WORKER_ENDPOINTS:
+            if re.match(pattern, path):
+                return self._check_worker_api_key(request)
+        
+        # 其他路径需要 Session 认证
+        return request.user and request.user.is_authenticated
+    
+    def _check_worker_api_key(self, request):
+        """验证 Worker API Key"""
+        api_key = request.headers.get('X-Worker-API-Key')
+        expected_key = getattr(settings, 'WORKER_API_KEY', None)
+        
+        if not expected_key:
+            # 未配置 API Key 时，拒绝所有 Worker 请求
+            logger.warning("WORKER_API_KEY 未配置，拒绝 Worker 请求")
+            return False
+        
+        if not api_key:
+            logger.warning(f"Worker 请求缺少 X-Worker-API-Key Header: {request.path}")
+            return False
+        
+        if api_key != expected_key:
+            logger.warning(f"Worker API Key 无效: {request.path}")
+            return False
+        
+        return True

backend/apps/common/urls.py

@@ -2,14 +2,18 @@
 通用模块 URL 配置
 
 路由说明：
+- /api/health/    健康检查接口（无需认证）
 - /api/auth/*     认证相关接口（登录、登出、用户信息）
 - /api/system/*   系统管理接口（日志查看等）
 """
 
 from django.urls import path
-from .views import LoginView, LogoutView, MeView, ChangePasswordView, SystemLogsView, SystemLogFilesView
+from .views import LoginView, LogoutView, MeView, ChangePasswordView, SystemLogsView, SystemLogFilesView, HealthCheckView
 
 urlpatterns = [
+    # 健康检查（无需认证）
+    path('health/', HealthCheckView.as_view(), name='health-check'),
+    
     # 认证相关
     path('auth/login/', LoginView.as_view(), name='auth-login'),
     path('auth/logout/', LogoutView.as_view(), name='auth-logout'),

backend/apps/common/utils/__init__.py

@@ -11,6 +11,7 @@ from .csv_utils import (
     generate_csv_rows,
     format_list_field,
     format_datetime,
+    create_csv_export_response,
     UTF8_BOM,
 )
 
@@ -24,5 +25,6 @@ __all__ = [
     'generate_csv_rows',
     'format_list_field',
     'format_datetime',
+    'create_csv_export_response',
     'UTF8_BOM',
 ]

backend/apps/common/utils/csv_utils.py

@@ -4,13 +4,21 @@
 - UTF-8 BOM（Excel 兼容）
 - RFC 4180 规范转义
 - 流式生成（内存友好）
+- 带 Content-Length 的文件响应（支持浏览器下载进度显示）
 """
 
 import csv
 import io
+import os
+import tempfile
+import logging
 from datetime import datetime
 from typing import Iterator, Dict, Any, List, Callable, Optional
 
+from django.http import FileResponse, StreamingHttpResponse
+
+logger = logging.getLogger(__name__)
+
 # UTF-8 BOM，确保 Excel 正确识别编码
 UTF8_BOM = '\ufeff'
 
@@ -114,3 +122,123 @@ def format_datetime(dt: Optional[datetime]) -> str:
         dt = timezone.localtime(dt)
     
     return dt.strftime('%Y-%m-%d %H:%M:%S')
+
+
+def create_csv_export_response(
+    data_iterator: Iterator[Dict[str, Any]],
+    headers: List[str],
+    filename: str,
+    field_formatters: Optional[Dict[str, Callable]] = None,
+    show_progress: bool = True
+) -> FileResponse | StreamingHttpResponse:
+    """
+    创建 CSV 导出响应
+    
+    根据 show_progress 参数选择响应类型：
+    - True: 使用临时文件 + FileResponse，带 Content-Length（浏览器显示下载进度）
+    - False: 使用 StreamingHttpResponse（内存更友好，但无下载进度）
+    
+    Args:
+        data_iterator: 数据迭代器，每个元素是一个字典
+        headers: CSV 表头列表
+        filename: 下载文件名（如 "export_2024.csv"）
+        field_formatters: 字段格式化函数字典
+        show_progress: 是否显示下载进度（默认 True）
+    
+    Returns:
+        FileResponse 或 StreamingHttpResponse
+    
+    Example:
+        >>> data_iter = service.iter_data()
+        >>> headers = ['url', 'host', 'created_at']
+        >>> formatters = {'created_at': format_datetime}
+        >>> response = create_csv_export_response(
+        ...     data_iter, headers, 'websites.csv', formatters
+        ... )
+        >>> return response
+    """
+    if show_progress:
+        return _create_file_response(data_iterator, headers, filename, field_formatters)
+    else:
+        return _create_streaming_response(data_iterator, headers, filename, field_formatters)
+
+
+def _create_file_response(
+    data_iterator: Iterator[Dict[str, Any]],
+    headers: List[str],
+    filename: str,
+    field_formatters: Optional[Dict[str, Callable]] = None
+) -> FileResponse:
+    """
+    创建带 Content-Length 的文件响应（支持浏览器下载进度）
+    
+    实现方式：先写入临时文件，再返回 FileResponse
+    """
+    # 创建临时文件
+    temp_file = tempfile.NamedTemporaryFile(
+        mode='w', 
+        suffix='.csv', 
+        delete=False, 
+        encoding='utf-8'
+    )
+    temp_path = temp_file.name
+    
+    try:
+        # 流式写入 CSV 数据到临时文件
+        for row in generate_csv_rows(data_iterator, headers, field_formatters):
+            temp_file.write(row)
+        temp_file.close()
+        
+        # 获取文件大小
+        file_size = os.path.getsize(temp_path)
+        
+        # 创建文件响应
+        response = FileResponse(
+            open(temp_path, 'rb'),
+            content_type='text/csv; charset=utf-8',
+            as_attachment=True,
+            filename=filename
+        )
+        response['Content-Length'] = file_size
+        
+        # 设置清理回调：响应完成后删除临时文件
+        original_close = response.file_to_stream.close
+        def close_and_cleanup():
+            original_close()
+            try:
+                os.unlink(temp_path)
+            except OSError:
+                pass
+        response.file_to_stream.close = close_and_cleanup
+        
+        return response
+        
+    except Exception as e:
+        # 清理临时文件
+        try:
+            temp_file.close()
+        except:
+            pass
+        try:
+            os.unlink(temp_path)
+        except OSError:
+            pass
+        logger.error(f"创建 CSV 导出响应失败: {e}")
+        raise
+
+
+def _create_streaming_response(
+    data_iterator: Iterator[Dict[str, Any]],
+    headers: List[str],
+    filename: str,
+    field_formatters: Optional[Dict[str, Callable]] = None
+) -> StreamingHttpResponse:
+    """
+    创建流式响应（无 Content-Length，内存更友好）
+    """
+    response = StreamingHttpResponse(
+        generate_csv_rows(data_iterator, headers, field_formatters),
+        content_type='text/csv; charset=utf-8'
+    )
+    response['Content-Disposition'] = f'attachment; filename="{filename}"'
+    return response

backend/apps/common/utils/filter_utils.py

@@ -29,11 +29,19 @@ from dataclasses import dataclass
 from typing import List, Dict, Optional, Union
 from enum import Enum
 
-from django.db.models import QuerySet, Q
+from django.db.models import QuerySet, Q, F, Func, CharField
+from django.db.models.functions import Cast
 
 logger = logging.getLogger(__name__)
 
 
+class ArrayToString(Func):
+    """PostgreSQL array_to_string 函数"""
+    function = 'array_to_string'
+    template = "%(function)s(%(expressions)s, ',')"
+    output_field = CharField()
+
+
 class LogicalOp(Enum):
     """逻辑运算符"""
     AND = 'AND'
@@ -86,9 +94,21 @@ class QueryParser:
         if not query_string or not query_string.strip():
             return []
         
+        # 第一步：提取所有过滤条件并用占位符替换，保护引号内的空格
+        filters_found = []
+        placeholder_pattern = '__FILTER_{}__'
+        
+        def replace_filter(match):
+            idx = len(filters_found)
+            filters_found.append(match.group(0))
+            return placeholder_pattern.format(idx)
+        
+        # 先用正则提取所有 field="value" 形式的条件
+        protected = cls.FILTER_PATTERN.sub(replace_filter, query_string)
+        
         # 标准化逻辑运算符
         # 先处理 || 和 or -> __OR__
-        normalized = cls.OR_PATTERN.sub(' __OR__ ', query_string)
+        normalized = cls.OR_PATTERN.sub(' __OR__ ', protected)
         # 再处理 && 和 and -> __AND__
         normalized = cls.AND_PATTERN.sub(' __AND__ ', normalized)
         
@@ -103,20 +123,26 @@ class QueryParser:
                 pending_op = LogicalOp.OR
             elif token == '__AND__':
                 pending_op = LogicalOp.AND
-            else:
-                # 尝试解析为过滤条件
-                match = cls.FILTER_PATTERN.match(token)
-                if match:
-                    field, operator, value = match.groups()
-                    groups.append(FilterGroup(
-                        filter=ParsedFilter(
-                            field=field.lower(),
-                            operator=operator,
-                            value=value
-                        ),
-                        logical_op=pending_op if groups else LogicalOp.AND  # 第一个条件默认 AND
-                    ))
-                    pending_op = LogicalOp.AND  # 重置为默认 AND
+            elif token.startswith('__FILTER_') and token.endswith('__'):
+                # 还原占位符为原始过滤条件
+                try:
+                    idx = int(token[9:-2])  # 提取索引
+                    original_filter = filters_found[idx]
+                    match = cls.FILTER_PATTERN.match(original_filter)
+                    if match:
+                        field, operator, value = match.groups()
+                        groups.append(FilterGroup(
+                            filter=ParsedFilter(
+                                field=field.lower(),
+                                operator=operator,
+                                value=value
+                            ),
+                            logical_op=pending_op if groups else LogicalOp.AND
+                        ))
+                        pending_op = LogicalOp.AND  # 重置为默认 AND
+                except (ValueError, IndexError):
+                    pass
+            # 其他 token 忽略（无效输入）
         
         return groups
 
@@ -151,6 +177,21 @@ class QueryBuilder:
         
         json_array_fields = json_array_fields or []
         
+        # 收集需要 annotate 的数组模糊搜索字段
+        array_fuzzy_fields = set()
+        
+        # 第一遍：检查是否有数组模糊匹配
+        for group in filter_groups:
+            f = group.filter
+            db_field = field_mapping.get(f.field)
+            if db_field and db_field in json_array_fields and f.operator == '=':
+                array_fuzzy_fields.add(db_field)
+        
+        # 对数组模糊搜索字段做 annotate
+        for field in array_fuzzy_fields:
+            annotate_name = f'{field}_text'
+            queryset = queryset.annotate(**{annotate_name: ArrayToString(F(field))})
+        
         # 构建 Q 对象
         combined_q = None
         
@@ -187,8 +228,17 @@ class QueryBuilder:
     def _build_single_q(cls, field: str, operator: str, value: str, is_json_array: bool = False) -> Optional[Q]:
         """构建单个条件的 Q 对象"""
         if is_json_array:
-            # JSON 数组字段使用 __contains 查询
-            return Q(**{f'{field}__contains': [value]})
+            if operator == '==':
+                # 精确匹配：数组中包含完全等于 value 的元素
+                return Q(**{f'{field}__contains': [value]})
+            elif operator == '!=':
+                # 不包含：数组中不包含完全等于 value 的元素
+                return ~Q(**{f'{field}__contains': [value]})
+            else:  # '=' 模糊匹配
+                # 使用 annotate 后的字段进行模糊搜索
+                # 字段已在 build_query 中通过 ArrayToString 转换为文本
+                annotate_name = f'{field}_text'
+                return Q(**{f'{annotate_name}__icontains': value})
         
         if operator == '!=':
             return cls._build_not_equal_q(field, value)

backend/apps/common/views/__init__.py

@@ -2,11 +2,17 @@
 通用模块视图导出
 
 包含：
+- 健康检查视图：Docker 健康检查
 - 认证相关视图：登录、登出、用户信息、修改密码
 - 系统日志视图：实时日志查看
 """
 
+from .health_views import HealthCheckView
 from .auth_views import LoginView, LogoutView, MeView, ChangePasswordView
 from .system_log_views import SystemLogsView, SystemLogFilesView
 
-__all__ = ['LoginView', 'LogoutView', 'MeView', 'ChangePasswordView', 'SystemLogsView', 'SystemLogFilesView']
+__all__ = [
+    'HealthCheckView',
+    'LoginView', 'LogoutView', 'MeView', 'ChangePasswordView',
+    'SystemLogsView', 'SystemLogFilesView',
+]

backend/apps/common/views/auth_views.py

@@ -9,7 +9,7 @@ from django.utils.decorators import method_decorator
 from rest_framework import status
 from rest_framework.views import APIView
 from rest_framework.response import Response
-from rest_framework.permissions import AllowAny, IsAuthenticated
+from rest_framework.permissions import AllowAny
 
 from apps.common.response_helpers import success_response, error_response
 from apps.common.error_codes import ErrorCodes
@@ -134,30 +134,10 @@ class ChangePasswordView(APIView):
     修改密码
     POST /api/auth/change-password/
     """
-    authentication_classes = []  # 禁用认证（绕过 CSRF）
-    permission_classes = [AllowAny]  # 手动检查登录状态
     
     def post(self, request):
-        # 手动检查登录状态（从 session 获取用户）
-        from django.contrib.auth import get_user_model
-        User = get_user_model()
-        
-        user_id = request.session.get('_auth_user_id')
-        if not user_id:
-            return error_response(
-                code=ErrorCodes.UNAUTHORIZED,
-                message='Please login first',
-                status_code=status.HTTP_401_UNAUTHORIZED
-            )
-        
-        try:
-            user = User.objects.get(pk=user_id)
-        except User.DoesNotExist:
-            return error_response(
-                code=ErrorCodes.UNAUTHORIZED,
-                message='User does not exist',
-                status_code=status.HTTP_401_UNAUTHORIZED
-            )
+        # 使用全局权限类验证，request.user 已经是认证用户
+        user = request.user
         
         # CamelCaseParser 将 oldPassword -> old_password
         old_password = request.data.get('old_password')

backend/apps/common/views/health_views.py

@@ -0,0 +1,24 @@
+"""
+健康检查视图
+
+提供 Docker 健康检查端点，无需认证。
+"""
+
+from rest_framework.views import APIView
+from rest_framework.response import Response
+from rest_framework.permissions import AllowAny
+
+
+class HealthCheckView(APIView):
+    """
+    健康检查端点
+    
+    GET /api/health/
+    
+    返回服务状态，用于 Docker 健康检查。
+    此端点无需认证。
+    """
+    permission_classes = [AllowAny]
+    
+    def get(self, request):
+        return Response({'status': 'ok'})

backend/apps/common/views/system_log_views.py

@@ -9,7 +9,6 @@ import logging
 from django.utils.decorators import method_decorator
 from django.views.decorators.csrf import csrf_exempt
 from rest_framework import status
-from rest_framework.permissions import AllowAny
 from rest_framework.response import Response
 from rest_framework.views import APIView
 
@@ -42,9 +41,6 @@ class SystemLogFilesView(APIView):
             ]
         }
     """
-    
-    authentication_classes = []
-    permission_classes = [AllowAny]
 
     def __init__(self, **kwargs):
         super().__init__(**kwargs)
@@ -80,15 +76,7 @@ class SystemLogsView(APIView):
         {
             "content": "日志内容字符串..."
         }
-        
-    Note:
-        - 当前为开发阶段，暂时允许匿名访问
-        - 生产环境应添加管理员权限验证
     """
-    
-    # TODO: 生产环境应改为 IsAdminUser 权限
-    authentication_classes = []
-    permission_classes = [AllowAny]
 
     def __init__(self, **kwargs):
         super().__init__(**kwargs)

backend/apps/common/websocket_auth.py

@@ -0,0 +1,44 @@
+"""
+WebSocket 认证基类
+
+提供需要认证的 WebSocket Consumer 基类
+"""
+
+import logging
+from channels.generic.websocket import AsyncWebsocketConsumer
+
+logger = logging.getLogger(__name__)
+
+
+class AuthenticatedWebsocketConsumer(AsyncWebsocketConsumer):
+    """
+    需要认证的 WebSocket Consumer 基类
+    
+    子类应该重写 on_connect() 方法实现具体的连接逻辑
+    """
+    
+    async def connect(self):
+        """
+        连接时验证用户认证状态
+        
+        未认证时使用 close(code=4001) 拒绝连接
+        """
+        user = self.scope.get('user')
+        
+        if not user or not user.is_authenticated:
+            logger.warning(
+                f"WebSocket 连接被拒绝：用户未认证 - Path: {self.scope.get('path')}"
+            )
+            await self.close(code=4001)
+            return
+        
+        # 调用子类的连接逻辑
+        await self.on_connect()
+    
+    async def on_connect(self):
+        """
+        子类实现具体的连接逻辑
+        
+        默认实现：接受连接
+        """
+        await self.accept()

backend/apps/engine/consumers/worker_deploy_consumer.py

@@ -6,17 +6,17 @@ import json
 import logging
 import asyncio
 import os
-from channels.generic.websocket import AsyncWebsocketConsumer
 from asgiref.sync import sync_to_async
 
 from django.conf import settings
 
+from apps.common.websocket_auth import AuthenticatedWebsocketConsumer
 from apps.engine.services import WorkerService
 
 logger = logging.getLogger(__name__)
 
 
-class WorkerDeployConsumer(AsyncWebsocketConsumer):
+class WorkerDeployConsumer(AuthenticatedWebsocketConsumer):
     """
     Worker 交互式终端 WebSocket Consumer
     
@@ -31,8 +31,8 @@ class WorkerDeployConsumer(AsyncWebsocketConsumer):
         self.read_task = None
         self.worker_service = WorkerService()
     
-    async def connect(self):
-        """连接时加入对应 Worker 的组并自动建立 SSH 连接"""
+    async def on_connect(self):
+        """连接时加入对应 Worker 的组并自动建立 SSH 连接（已通过认证）"""
         self.worker_id = self.scope['url_route']['kwargs']['worker_id']
         self.group_name = f'worker_deploy_{self.worker_id}'
         

backend/apps/engine/management/commands/init_default_engine.py

@@ -15,9 +15,10 @@
 """
 
 from django.core.management.base import BaseCommand
+from io import StringIO
 from pathlib import Path
 
-import yaml
+from ruamel.yaml import YAML
 
 from apps.engine.models import ScanEngine
 
@@ -44,10 +45,12 @@ class Command(BaseCommand):
         with open(config_path, 'r', encoding='utf-8') as f:
             default_config = f.read()
 
-        # 解析 YAML 为字典，后续用于生成子引擎配置
+        # 使用 ruamel.yaml 解析，保留注释
+        yaml_parser = YAML()
+        yaml_parser.preserve_quotes = True
         try:
-            config_dict = yaml.safe_load(default_config) or {}
-        except yaml.YAMLError as e:
+            config_dict = yaml_parser.load(default_config) or {}
+        except Exception as e:
             self.stdout.write(self.style.ERROR(f'引擎配置 YAML 解析失败: {e}'))
             return
 
@@ -83,15 +86,13 @@ class Command(BaseCommand):
             if scan_type != 'subdomain_discovery' and 'tools' not in scan_cfg:
                 continue
 
-            # 构造只包含当前扫描类型配置的 YAML
+            # 构造只包含当前扫描类型配置的 YAML（保留注释）
             single_config = {scan_type: scan_cfg}
             try:
-                single_yaml = yaml.safe_dump(
-                    single_config,
-                    sort_keys=False,
-                    allow_unicode=True,
-                )
-            except yaml.YAMLError as e:
+                stream = StringIO()
+                yaml_parser.dump(single_config, stream)
+                single_yaml = stream.getvalue()
+            except Exception as e:
                 self.stdout.write(self.style.ERROR(f'生成子引擎 {scan_type} 配置失败: {e}'))
                 continue
 

backend/apps/engine/migrations/0001_initial.py

@@ -0,0 +1,213 @@
+# Generated by Django 5.2.7 on 2026-01-02 04:45
+
+from django.db import migrations, models
+
+
+class Migration(migrations.Migration):
+
+    initial = True
+
+    dependencies = [
+    ]
+
+    operations = [
+        migrations.CreateModel(
+            name='NucleiTemplateRepo',
+            fields=[
+                ('id', models.BigAutoField(auto_created=True, primary_key=True, serialize=False, verbose_name='ID')),
+                ('name', models.CharField(help_text='仓库名称，用于前端展示和配置引用', max_length=200, unique=True)),
+                ('repo_url', models.CharField(help_text='Git 仓库地址', max_length=500)),
+                ('local_path', models.CharField(blank=True, default='', help_text='本地工作目录绝对路径', max_length=500)),
+                ('commit_hash', models.CharField(blank=True, default='', help_text='最后同步的 Git commit hash，用于 Worker 版本校验', max_length=40)),
+                ('last_synced_at', models.DateTimeField(blank=True, help_text='最后一次成功同步时间', null=True)),
+                ('created_at', models.DateTimeField(auto_now_add=True, help_text='创建时间')),
+                ('updated_at', models.DateTimeField(auto_now=True, help_text='更新时间')),
+            ],
+            options={
+                'verbose_name': 'Nuclei 模板仓库',
+                'verbose_name_plural': 'Nuclei 模板仓库',
+                'db_table': 'nuclei_template_repo',
+            },
+        ),
+        migrations.CreateModel(
+            name='ARLFingerprint',
+            fields=[
+                ('id', models.BigAutoField(auto_created=True, primary_key=True, serialize=False, verbose_name='ID')),
+                ('name', models.CharField(help_text='指纹名称', max_length=300, unique=True)),
+                ('rule', models.TextField(help_text='匹配规则表达式')),
+                ('created_at', models.DateTimeField(auto_now_add=True)),
+            ],
+            options={
+                'verbose_name': 'ARL 指纹',
+                'verbose_name_plural': 'ARL 指纹',
+                'db_table': 'arl_fingerprint',
+                'ordering': ['-created_at'],
+                'indexes': [models.Index(fields=['name'], name='arl_fingerp_name_c3a305_idx'), models.Index(fields=['-created_at'], name='arl_fingerp_created_ed1060_idx')],
+            },
+        ),
+        migrations.CreateModel(
+            name='EholeFingerprint',
+            fields=[
+                ('id', models.BigAutoField(auto_created=True, primary_key=True, serialize=False, verbose_name='ID')),
+                ('cms', models.CharField(help_text='产品/CMS名称', max_length=200)),
+                ('method', models.CharField(default='keyword', help_text='匹配方式', max_length=200)),
+                ('location', models.CharField(default='body', help_text='匹配位置', max_length=200)),
+                ('keyword', models.JSONField(default=list, help_text='关键词列表')),
+                ('is_important', models.BooleanField(default=False, help_text='是否重点资产')),
+                ('type', models.CharField(blank=True, default='-', help_text='分类', max_length=100)),
+                ('created_at', models.DateTimeField(auto_now_add=True)),
+            ],
+            options={
+                'verbose_name': 'EHole 指纹',
+                'verbose_name_plural': 'EHole 指纹',
+                'db_table': 'ehole_fingerprint',
+                'ordering': ['-created_at'],
+                'indexes': [models.Index(fields=['cms'], name='ehole_finge_cms_72ca2c_idx'), models.Index(fields=['method'], name='ehole_finge_method_17f0db_idx'), models.Index(fields=['location'], name='ehole_finge_locatio_7bb82b_idx'), models.Index(fields=['type'], name='ehole_finge_type_ca2bce_idx'), models.Index(fields=['is_important'], name='ehole_finge_is_impo_d56e64_idx'), models.Index(fields=['-created_at'], name='ehole_finge_created_d862b0_idx')],
+                'constraints': [models.UniqueConstraint(fields=('cms', 'method', 'location'), name='unique_ehole_fingerprint')],
+            },
+        ),
+        migrations.CreateModel(
+            name='FingerPrintHubFingerprint',
+            fields=[
+                ('id', models.BigAutoField(auto_created=True, primary_key=True, serialize=False, verbose_name='ID')),
+                ('fp_id', models.CharField(help_text='指纹ID', max_length=200, unique=True)),
+                ('name', models.CharField(help_text='指纹名称', max_length=300)),
+                ('author', models.CharField(blank=True, default='', help_text='作者', max_length=200)),
+                ('tags', models.CharField(blank=True, default='', help_text='标签', max_length=500)),
+                ('severity', models.CharField(blank=True, default='info', help_text='严重程度', max_length=50)),
+                ('metadata', models.JSONField(blank=True, default=dict, help_text='元数据')),
+                ('http', models.JSONField(default=list, help_text='HTTP 匹配规则')),
+                ('source_file', models.CharField(blank=True, default='', help_text='来源文件', max_length=500)),
+                ('created_at', models.DateTimeField(auto_now_add=True)),
+            ],
+            options={
+                'verbose_name': 'FingerPrintHub 指纹',
+                'verbose_name_plural': 'FingerPrintHub 指纹',
+                'db_table': 'fingerprinthub_fingerprint',
+                'ordering': ['-created_at'],
+                'indexes': [models.Index(fields=['fp_id'], name='fingerprint_fp_id_df467f_idx'), models.Index(fields=['name'], name='fingerprint_name_95b6fb_idx'), models.Index(fields=['author'], name='fingerprint_author_80f54b_idx'), models.Index(fields=['severity'], name='fingerprint_severit_f70422_idx'), models.Index(fields=['-created_at'], name='fingerprint_created_bec16c_idx')],
+            },
+        ),
+        migrations.CreateModel(
+            name='FingersFingerprint',
+            fields=[
+                ('id', models.BigAutoField(auto_created=True, primary_key=True, serialize=False, verbose_name='ID')),
+                ('name', models.CharField(help_text='指纹名称', max_length=300, unique=True)),
+                ('link', models.URLField(blank=True, default='', help_text='相关链接', max_length=500)),
+                ('rule', models.JSONField(default=list, help_text='匹配规则数组')),
+                ('tag', models.JSONField(default=list, help_text='标签数组')),
+                ('focus', models.BooleanField(default=False, help_text='是否重点关注')),
+                ('default_port', models.JSONField(blank=True, default=list, help_text='默认端口数组')),
+                ('created_at', models.DateTimeField(auto_now_add=True)),
+            ],
+            options={
+                'verbose_name': 'Fingers 指纹',
+                'verbose_name_plural': 'Fingers 指纹',
+                'db_table': 'fingers_fingerprint',
+                'ordering': ['-created_at'],
+                'indexes': [models.Index(fields=['name'], name='fingers_fin_name_952de0_idx'), models.Index(fields=['link'], name='fingers_fin_link_4c6b7f_idx'), models.Index(fields=['focus'], name='fingers_fin_focus_568c7f_idx'), models.Index(fields=['-created_at'], name='fingers_fin_created_46fc91_idx')],
+            },
+        ),
+        migrations.CreateModel(
+            name='GobyFingerprint',
+            fields=[
+                ('id', models.BigAutoField(auto_created=True, primary_key=True, serialize=False, verbose_name='ID')),
+                ('name', models.CharField(help_text='产品名称', max_length=300, unique=True)),
+                ('logic', models.CharField(help_text='逻辑表达式', max_length=500)),
+                ('rule', models.JSONField(default=list, help_text='规则数组')),
+                ('created_at', models.DateTimeField(auto_now_add=True)),
+            ],
+            options={
+                'verbose_name': 'Goby 指纹',
+                'verbose_name_plural': 'Goby 指纹',
+                'db_table': 'goby_fingerprint',
+                'ordering': ['-created_at'],
+                'indexes': [models.Index(fields=['name'], name='goby_finger_name_82084c_idx'), models.Index(fields=['logic'], name='goby_finger_logic_a63226_idx'), models.Index(fields=['-created_at'], name='goby_finger_created_50e000_idx')],
+            },
+        ),
+        migrations.CreateModel(
+            name='ScanEngine',
+            fields=[
+                ('id', models.AutoField(primary_key=True, serialize=False)),
+                ('name', models.CharField(help_text='引擎名称', max_length=200, unique=True)),
+                ('configuration', models.CharField(blank=True, default='', help_text='引擎配置，yaml 格式', max_length=10000)),
+                ('created_at', models.DateTimeField(auto_now_add=True, help_text='创建时间')),
+                ('updated_at', models.DateTimeField(auto_now=True, help_text='更新时间')),
+            ],
+            options={
+                'verbose_name': '扫描引擎',
+                'verbose_name_plural': '扫描引擎',
+                'db_table': 'scan_engine',
+                'ordering': ['-created_at'],
+                'indexes': [models.Index(fields=['-created_at'], name='scan_engine_created_da4870_idx')],
+            },
+        ),
+        migrations.CreateModel(
+            name='WappalyzerFingerprint',
+            fields=[
+                ('id', models.BigAutoField(auto_created=True, primary_key=True, serialize=False, verbose_name='ID')),
+                ('name', models.CharField(help_text='应用名称', max_length=300, unique=True)),
+                ('cats', models.JSONField(default=list, help_text='分类 ID 数组')),
+                ('cookies', models.JSONField(blank=True, default=dict, help_text='Cookie 检测规则')),
+                ('headers', models.JSONField(blank=True, default=dict, help_text='HTTP Header 检测规则')),
+                ('script_src', models.JSONField(blank=True, default=list, help_text='脚本 URL 正则数组')),
+                ('js', models.JSONField(blank=True, default=list, help_text='JavaScript 变量检测规则')),
+                ('implies', models.JSONField(blank=True, default=list, help_text='依赖关系数组')),
+                ('meta', models.JSONField(blank=True, default=dict, help_text='HTML meta 标签检测规则')),
+                ('html', models.JSONField(blank=True, default=list, help_text='HTML 内容正则数组')),
+                ('description', models.TextField(blank=True, default='', help_text='应用描述')),
+                ('website', models.URLField(blank=True, default='', help_text='官网链接', max_length=500)),
+                ('cpe', models.CharField(blank=True, default='', help_text='CPE 标识符', max_length=300)),
+                ('created_at', models.DateTimeField(auto_now_add=True)),
+            ],
+            options={
+                'verbose_name': 'Wappalyzer 指纹',
+                'verbose_name_plural': 'Wappalyzer 指纹',
+                'db_table': 'wappalyzer_fingerprint',
+                'ordering': ['-created_at'],
+                'indexes': [models.Index(fields=['name'], name='wappalyzer__name_63c669_idx'), models.Index(fields=['website'], name='wappalyzer__website_88de1c_idx'), models.Index(fields=['cpe'], name='wappalyzer__cpe_30c761_idx'), models.Index(fields=['-created_at'], name='wappalyzer__created_8e6c21_idx')],
+            },
+        ),
+        migrations.CreateModel(
+            name='Wordlist',
+            fields=[
+                ('id', models.AutoField(primary_key=True, serialize=False)),
+                ('name', models.CharField(help_text='字典名称，唯一', max_length=200, unique=True)),
+                ('description', models.CharField(blank=True, default='', help_text='字典描述', max_length=200)),
+                ('file_path', models.CharField(help_text='后端保存的字典文件绝对路径', max_length=500)),
+                ('file_size', models.BigIntegerField(default=0, help_text='文件大小（字节）')),
+                ('line_count', models.IntegerField(default=0, help_text='字典行数')),
+                ('file_hash', models.CharField(blank=True, default='', help_text='文件 SHA-256 哈希，用于缓存校验', max_length=64)),
+                ('created_at', models.DateTimeField(auto_now_add=True, help_text='创建时间')),
+                ('updated_at', models.DateTimeField(auto_now=True, help_text='更新时间')),
+            ],
+            options={
+                'verbose_name': '字典文件',
+                'verbose_name_plural': '字典文件',
+                'db_table': 'wordlist',
+                'ordering': ['-created_at'],
+                'indexes': [models.Index(fields=['-created_at'], name='wordlist_created_4afb02_idx')],
+            },
+        ),
+        migrations.CreateModel(
+            name='WorkerNode',
+            fields=[
+                ('id', models.BigAutoField(auto_created=True, primary_key=True, serialize=False, verbose_name='ID')),
+                ('name', models.CharField(help_text='节点名称', max_length=100)),
+                ('ip_address', models.GenericIPAddressField(help_text='IP 地址（本地节点为 127.0.0.1）')),
+                ('ssh_port', models.IntegerField(default=22, help_text='SSH 端口')),
+                ('username', models.CharField(default='root', help_text='SSH 用户名', max_length=50)),
+                ('password', models.CharField(blank=True, default='', help_text='SSH 密码', max_length=200)),
+                ('is_local', models.BooleanField(default=False, help_text='是否为本地节点（Docker 容器内）')),
+                ('status', models.CharField(choices=[('pending', '待部署'), ('deploying', '部署中'), ('online', '在线'), ('offline', '离线'), ('updating', '更新中'), ('outdated', '版本过低')], default='pending', help_text='状态: pending/deploying/online/offline', max_length=20)),
+                ('created_at', models.DateTimeField(auto_now_add=True)),
+                ('updated_at', models.DateTimeField(auto_now=True)),
+            ],
+            options={
+                'verbose_name': 'Worker 节点',
+                'db_table': 'worker_node',
+                'ordering': ['-created_at'],
+                'constraints': [models.UniqueConstraint(condition=models.Q(('is_local', False)), fields=('ip_address',), name='unique_remote_worker_ip'), models.UniqueConstraint(fields=('name',), name='unique_worker_name')],
+            },
+        ),
+    ]

backend/apps/engine/scheduler.py

@@ -88,6 +88,8 @@ def _register_scheduled_jobs(scheduler: BackgroundScheduler):
         replace_existing=True,
     )
     logger.info("  - 已注册: 扫描结果清理（每天 03:00）")
+    
+    # 注意：搜索物化视图刷新已迁移到 pg_ivm 增量维护，无需定时任务
 
 
 def _trigger_scheduled_scans():

backend/apps/engine/services/deploy_service.py

@@ -66,6 +66,7 @@ def get_start_agent_script(
     # 替换变量
     script = script.replace("{{HEARTBEAT_API_URL}}", heartbeat_api_url or '')
     script = script.replace("{{WORKER_ID}}", str(worker_id) if worker_id else '')
+    script = script.replace("{{WORKER_API_KEY}}", getattr(settings, 'WORKER_API_KEY', ''))
     
     # 注入镜像版本配置（确保远程节点使用相同版本）
     docker_user = getattr(settings, 'DOCKER_USER', 'yyhuni')

backend/apps/engine/services/task_distributor.py

@@ -284,6 +284,7 @@ class TaskDistributor:
         env_vars = [
             f"-e SERVER_URL={shlex.quote(server_url)}",
             f"-e IS_LOCAL={is_local_str}",
+            f"-e WORKER_API_KEY={shlex.quote(settings.WORKER_API_KEY)}",  # Worker API 认证密钥
             "-e PREFECT_HOME=/tmp/.prefect",  # 设置 Prefect 数据目录到可写位置
             "-e PREFECT_SERVER_EPHEMERAL_ENABLED=true",  # 启用 ephemeral server（本地临时服务器）
             "-e PREFECT_SERVER_EPHEMERAL_STARTUP_TIMEOUT_SECONDS=120",  # 增加启动超时时间

backend/apps/engine/views/worker_views.py

@@ -340,13 +340,12 @@ class WorkerNodeViewSet(viewsets.ModelViewSet):
         返回:
         {
             "db": {"host": "...", "port": "...", ...},
-            "redisUrl": "...",
             "paths": {"results": "...", "logs": "..."}
         }
         
         配置逻辑:
-            - 本地 Worker (is_local=true): db_host=postgres, redis=redis:6379
-            - 远程 Worker (is_local=false): db_host=PUBLIC_HOST, redis=PUBLIC_HOST:6379
+            - 本地 Worker (is_local=true): db_host=postgres
+            - 远程 Worker (is_local=false): db_host=PUBLIC_HOST
         """
         from django.conf import settings
         import logging
@@ -371,20 +370,17 @@ class WorkerNodeViewSet(viewsets.ModelViewSet):
             if is_local_worker:
                 # 本地 Worker：直接用 Docker 内部服务名
                 worker_db_host = 'postgres'
-                worker_redis_url = 'redis://redis:6379/0'
             else:
                 # 远程 Worker：通过公网 IP 访问
                 public_host = settings.PUBLIC_HOST
                 if public_host in ('server', 'localhost', '127.0.0.1'):
                     logger.warning("远程 Worker 请求配置，但 PUBLIC_HOST=%s 不是有效的公网地址", public_host)
                 worker_db_host = public_host
-                worker_redis_url = f'redis://{public_host}:6379/0'
         else:
             # 远程数据库场景：所有 Worker 都用 DB_HOST
             worker_db_host = db_host
-            worker_redis_url = getattr(settings, 'WORKER_REDIS_URL', 'redis://redis:6379/0')
         
-        logger.info("返回 Worker 配置 - db_host: %s, redis_url: %s", worker_db_host, worker_redis_url)
+        logger.info("返回 Worker 配置 - db_host: %s", worker_db_host)
         
         return success_response(
             data={
@@ -395,7 +391,6 @@ class WorkerNodeViewSet(viewsets.ModelViewSet):
                     'user': settings.DATABASES['default']['USER'],
                     'password': settings.DATABASES['default']['PASSWORD'],
                 },
-                'redisUrl': worker_redis_url,
                 'paths': {
                     'results': getattr(settings, 'CONTAINER_RESULTS_MOUNT', '/opt/xingrin/results'),
                     'logs': getattr(settings, 'CONTAINER_LOGS_MOUNT', '/opt/xingrin/logs'),

backend/apps/scan/configs/command_templates.py

@@ -97,9 +97,11 @@ SITE_SCAN_COMMANDS = {
         'base': (
             "'{scan_tools_base}/httpx' -l '{url_file}' "
             '-status-code -content-type -content-length '
-            '-location -title -server -body-preview '
+            '-location -title -server '
             '-tech-detect -cdn -vhost '
-            '-random-agent -no-color -json'
+            '-include-response '
+            '-rstr 2000 '
+            '-random-agent -no-color -json -silent'
         ),
         'optional': {
             'threads': '-threads {threads}',
@@ -169,9 +171,11 @@ URL_FETCH_COMMANDS = {
         'base': (
             "'{scan_tools_base}/httpx' -l '{url_file}' "
             '-status-code -content-type -content-length '
-            '-location -title -server -body-preview '
+            '-location -title -server '
             '-tech-detect -cdn -vhost '
-            '-random-agent -no-color -json'
+            '-include-response '
+            '-rstr 2000 '
+            '-random-agent -no-color -json -silent'
         ),
         'optional': {
             'threads': '-threads {threads}',

backend/apps/scan/configs/engine_config_example.yaml

@@ -4,14 +4,12 @@
 # 必需参数：enabled（是否启用）
 # 可选参数：timeout（超时秒数，默认 auto 自动计算）
 
-# ==================== 子域名发现 ====================
-# 
-# Stage 1: 被动收集（并行） - 必选，至少启用一个工具
-# Stage 2: 字典爆破（可选） - 使用字典暴力枚举子域名
-# Stage 3: 变异生成 + 验证（可选） - 基于已发现域名生成变异，流式验证存活
-# Stage 4: DNS 存活验证（可选） - 验证所有候选域名是否能解析
-#
 subdomain_discovery:
+  # ==================== 子域名发现 ====================
+  # Stage 1: 被动收集（并行） - 必选，至少启用一个工具
+  # Stage 2: 字典爆破（可选） - 使用字典暴力枚举子域名
+  # Stage 3: 变异生成 + 验证（可选） - 基于已发现域名生成变异，流式验证存活
+  # Stage 4: DNS 存活验证（可选） - 验证所有候选域名是否能解析
   # === Stage 1: 被动收集工具（并行执行）===
   passive_tools:
     subfinder:
@@ -55,8 +53,8 @@ subdomain_discovery:
     subdomain_resolve:
       timeout: auto    # 自动根据候选子域数量计算
 
-# ==================== 端口扫描 ====================
 port_scan:
+  # ==================== 端口扫描 ====================
   tools:
     naabu_active:
       enabled: true
@@ -70,8 +68,8 @@ port_scan:
       enabled: true
       # timeout: auto    # 被动扫描通常较快
 
-# ==================== 站点扫描 ====================
 site_scan:
+  # ==================== 站点扫描 ====================
   tools:
     httpx:
       enabled: true
@@ -81,16 +79,16 @@ site_scan:
       # request-timeout: 10  # 单个请求超时秒数（默认 10）
       # retries: 2           # 请求失败重试次数
 
-# ==================== 指纹识别 ====================
-# 在 site_scan 后串行执行，识别 WebSite 的技术栈
 fingerprint_detect:
+  # ==================== 指纹识别 ====================
+  # 在 站点扫描 后串行执行，识别 WebSite 的技术栈
   tools:
     xingfinger:
       enabled: true
-      fingerprint-libs: [ehole, goby, wappalyzer]  # 启用的指纹库：ehole, goby, wappalyzer, fingers, fingerprinthub, arl
+      fingerprint-libs: [ehole, goby, wappalyzer, fingers, fingerprinthub, arl]  # 默认启动全部指纹库
 
-# ==================== 目录扫描 ====================
 directory_scan:
+  # ==================== 目录扫描 ====================
   tools:
     ffuf:
       enabled: true
@@ -103,8 +101,8 @@ directory_scan:
       match-codes: 200,201,301,302,401,403  # 匹配的 HTTP 状态码
       # rate: 0                           # 每秒请求数（默认 0 不限制）
 
-# ==================== URL 获取 ====================
 url_fetch:
+  # ==================== URL 获取 ====================
   tools:
     waymore:
       enabled: true
@@ -142,8 +140,8 @@ url_fetch:
       # request-timeout: 10  # 单个请求超时秒数（默认 10）
       # retries: 2           # 请求失败重试次数
 
-# ==================== 漏洞扫描 ====================
 vuln_scan:
+  # ==================== 漏洞扫描 ====================
   tools:
     dalfox_xss:
       enabled: true

backend/apps/scan/flows/directory_scan_flow.py

@@ -33,7 +33,7 @@ from apps.scan.handlers.scan_flow_handlers import (
     on_scan_flow_completed,
     on_scan_flow_failed,
 )
-from apps.scan.utils import config_parser, build_scan_command, ensure_wordlist_local
+from apps.scan.utils import config_parser, build_scan_command, ensure_wordlist_local, user_log
 
 logger = logging.getLogger(__name__)
 
@@ -413,6 +413,7 @@ def _run_scans_concurrently(
         logger.info("="*60)
         logger.info("使用工具: %s (并发模式, max_workers=%d)", tool_name, max_workers)
         logger.info("="*60)
+        user_log(scan_id, "directory_scan", f"Running {tool_name}")
 
         # 如果配置了 wordlist_name，则先确保本地存在对应的字典文件（含 hash 校验）
         wordlist_name = tool_config.get('wordlist_name')
@@ -467,6 +468,11 @@ def _run_scans_concurrently(
         total_tasks = len(scan_params_list)
         logger.info("开始分批执行 %d 个扫描任务（每批 %d 个）...", total_tasks, max_workers)
         
+        # 进度里程碑跟踪
+        last_progress_percent = 0
+        tool_directories = 0
+        tool_processed = 0
+        
         batch_num = 0
         for batch_start in range(0, total_tasks, max_workers):
             batch_end = min(batch_start + max_workers, total_tasks)
@@ -498,7 +504,9 @@ def _run_scans_concurrently(
                     result = future.result()  # 阻塞等待单个任务完成
                     directories_found = result.get('created_directories', 0)
                     total_directories += directories_found
+                    tool_directories += directories_found
                     processed_sites_count += 1
+                    tool_processed += 1
                     
                     logger.info(
                         "✓ [%d/%d] 站点扫描完成: %s - 发现 %d 个目录",
@@ -517,6 +525,19 @@ def _run_scans_concurrently(
                             "✗ [%d/%d] 站点扫描失败: %s - 错误: %s",
                             idx, len(sites), site_url, exc
                         )
+            
+            # 进度里程碑：每 20% 输出一次
+            current_progress = int((batch_end / total_tasks) * 100)
+            if current_progress >= last_progress_percent + 20:
+                user_log(scan_id, "directory_scan", f"Progress: {batch_end}/{total_tasks} sites scanned")
+                last_progress_percent = (current_progress // 20) * 20
+    
+        # 工具完成日志（开发者日志 + 用户日志）
+        logger.info(
+            "✓ 工具 %s 执行完成 - 已处理站点: %d/%d, 发现目录: %d",
+            tool_name, tool_processed, total_tasks, tool_directories
+        )
+        user_log(scan_id, "directory_scan", f"{tool_name} completed: found {tool_directories} directories")
     
     # 输出汇总信息
     if failed_sites:
@@ -605,6 +626,8 @@ def directory_scan_flow(
             "="*60
         )
         
+        user_log(scan_id, "directory_scan", "Starting directory scan")
+        
         # 参数验证
         if scan_id is None:
             raise ValueError("scan_id 不能为空")
@@ -625,7 +648,8 @@ def directory_scan_flow(
         sites_file, site_count = _export_site_urls(target_id, target_name, directory_scan_dir)
         
         if site_count == 0:
-            logger.warning("目标下没有站点，跳过目录扫描")
+            logger.warning("跳过目录扫描：没有站点可扫描 - Scan ID: %s", scan_id)
+            user_log(scan_id, "directory_scan", "Skipped: no sites to scan", "warning")
             return {
                 'success': True,
                 'scan_id': scan_id,
@@ -664,7 +688,9 @@ def directory_scan_flow(
             logger.warning("所有站点扫描均失败 - 总站点数: %d, 失败数: %d", site_count, len(failed_sites))
             # 不抛出异常，让扫描继续
         
-        logger.info("="*60 + "\n✓ 目录扫描完成\n" + "="*60)
+        # 记录 Flow 完成
+        logger.info("✓ 目录扫描完成 - 发现目录: %d", total_directories)
+        user_log(scan_id, "directory_scan", f"directory_scan completed: found {total_directories} directories")
         
         return {
             'success': True,

backend/apps/scan/flows/fingerprint_detect_flow.py

@@ -29,7 +29,7 @@ from apps.scan.tasks.fingerprint_detect import (
     export_urls_for_fingerprint_task,
     run_xingfinger_and_stream_update_tech_task,
 )
-from apps.scan.utils import build_scan_command
+from apps.scan.utils import build_scan_command, user_log
 from apps.scan.utils.fingerprint_helpers import get_fingerprint_paths
 
 logger = logging.getLogger(__name__)
@@ -37,28 +37,24 @@ logger = logging.getLogger(__name__)
 
 def calculate_fingerprint_detect_timeout(
     url_count: int,
-    base_per_url: float = 3.0,
-    min_timeout: int = 60
+    base_per_url: float = 10.0,
+    min_timeout: int = 300
 ) -> int:
     """
     根据 URL 数量计算超时时间
     
     公式：超时时间 = URL 数量 × 每 URL 基础时间
-    最小值：60秒
+    最小值：300秒
     无上限
     
     Args:
         url_count: URL 数量
-        base_per_url: 每 URL 基础时间（秒），默认 3秒
-        min_timeout: 最小超时时间（秒），默认 60秒
+        base_per_url: 每 URL 基础时间（秒），默认 10秒
+        min_timeout: 最小超时时间（秒），默认 300秒
         
     Returns:
         int: 计算出的超时时间（秒）
         
-    示例：
-        100 URL × 3秒 = 300秒
-        1000 URL × 3秒 = 3000秒（50分钟）
-        10000 URL × 3秒 = 30000秒（8.3小时）
     """
     timeout = int(url_count * base_per_url)
     return max(min_timeout, timeout)
@@ -172,6 +168,7 @@ def _run_fingerprint_detect(
             "开始执行 %s 指纹识别 - URL数: %d, 超时: %ds, 指纹库: %s",
             tool_name, url_count, timeout, list(fingerprint_paths.keys())
         )
+        user_log(scan_id, "fingerprint_detect", f"Running {tool_name}: {command}")
         
         # 6. 执行扫描任务
         try:
@@ -194,17 +191,21 @@ def _run_fingerprint_detect(
                 'fingerprint_libs': list(fingerprint_paths.keys())
             }
             
+            tool_updated = result.get('updated_count', 0)
             logger.info(
                 "✓ 工具 %s 执行完成 - 处理记录: %d, 更新: %d, 未找到: %d",
                 tool_name,
                 result.get('processed_records', 0),
-                result.get('updated_count', 0),
+                tool_updated,
                 result.get('not_found_count', 0)
             )
+            user_log(scan_id, "fingerprint_detect", f"{tool_name} completed: identified {tool_updated} fingerprints")
             
         except Exception as exc:
-            failed_tools.append({'tool': tool_name, 'reason': str(exc)})
+            reason = str(exc)
+            failed_tools.append({'tool': tool_name, 'reason': reason})
             logger.error("工具 %s 执行失败: %s", tool_name, exc, exc_info=True)
+            user_log(scan_id, "fingerprint_detect", f"{tool_name} failed: {reason}", "error")
     
     if failed_tools:
         logger.warning(
@@ -260,7 +261,8 @@ def fingerprint_detect_flow(
             'url_count': int,
             'processed_records': int,
             'updated_count': int,
-            'not_found_count': int,
+            'created_count': int,
+            'snapshot_count': int,
             'executed_tasks': list,
             'tool_stats': dict
         }
@@ -275,6 +277,8 @@ def fingerprint_detect_flow(
             "="*60
         )
         
+        user_log(scan_id, "fingerprint_detect", "Starting fingerprint detection")
+        
         # 参数验证
         if scan_id is None:
             raise ValueError("scan_id 不能为空")
@@ -296,7 +300,8 @@ def fingerprint_detect_flow(
         urls_file, url_count = _export_urls(target_id, fingerprint_dir, source)
         
         if url_count == 0:
-            logger.warning("目标下没有可用的 URL，跳过指纹识别")
+            logger.warning("跳过指纹识别：没有 URL 可扫描 - Scan ID: %s", scan_id)
+            user_log(scan_id, "fingerprint_detect", "Skipped: no URLs to scan", "warning")
             return {
                 'success': True,
                 'scan_id': scan_id,
@@ -307,6 +312,7 @@ def fingerprint_detect_flow(
                 'processed_records': 0,
                 'updated_count': 0,
                 'created_count': 0,
+                'snapshot_count': 0,
                 'executed_tasks': ['export_urls_for_fingerprint'],
                 'tool_stats': {
                     'total': 0,
@@ -334,8 +340,6 @@ def fingerprint_detect_flow(
             source=source
         )
         
-        logger.info("="*60 + "\n✓ 指纹识别完成\n" + "="*60)
-        
         # 动态生成已执行的任务列表
         executed_tasks = ['export_urls_for_fingerprint']
         executed_tasks.extend([f'run_xingfinger ({tool})' for tool in tool_stats.keys()])
@@ -344,6 +348,11 @@ def fingerprint_detect_flow(
         total_processed = sum(stats['result'].get('processed_records', 0) for stats in tool_stats.values())
         total_updated = sum(stats['result'].get('updated_count', 0) for stats in tool_stats.values())
         total_created = sum(stats['result'].get('created_count', 0) for stats in tool_stats.values())
+        total_snapshots = sum(stats['result'].get('snapshot_count', 0) for stats in tool_stats.values())
+        
+        # 记录 Flow 完成
+        logger.info("✓ 指纹识别完成 - 识别指纹: %d", total_updated)
+        user_log(scan_id, "fingerprint_detect", f"fingerprint_detect completed: identified {total_updated} fingerprints")
         
         successful_tools = [name for name in enabled_tools.keys() 
                            if name not in [f['tool'] for f in failed_tools]]
@@ -358,6 +367,7 @@ def fingerprint_detect_flow(
             'processed_records': total_processed,
             'updated_count': total_updated,
             'created_count': total_created,
+            'snapshot_count': total_snapshots,
             'executed_tasks': executed_tasks,
             'tool_stats': {
                 'total': len(enabled_tools),

backend/apps/scan/flows/initiate_scan_flow.py

@@ -114,8 +114,11 @@ def initiate_scan_flow(
         
         # ==================== Task 2: 获取引擎配置 ====================
         from apps.scan.models import Scan
-        scan = Scan.objects.select_related('engine').get(id=scan_id)
-        engine_config = scan.engine.configuration
+        scan = Scan.objects.get(id=scan_id)
+        engine_config = scan.yaml_configuration
+        
+        # 使用 engine_names 进行显示
+        display_engine_name = ', '.join(scan.engine_names) if scan.engine_names else engine_name
         
         # ==================== Task 3: 解析配置，生成执行计划 ====================
         orchestrator = FlowOrchestrator(engine_config)

backend/apps/scan/flows/port_scan_flow.py

@@ -28,7 +28,7 @@ from apps.scan.handlers.scan_flow_handlers import (
     on_scan_flow_completed,
     on_scan_flow_failed,
 )
-from apps.scan.utils import config_parser, build_scan_command
+from apps.scan.utils import config_parser, build_scan_command, user_log
 
 logger = logging.getLogger(__name__)
 
@@ -265,6 +265,7 @@ def _run_scans_sequentially(
         
         # 3. 执行扫描任务
         logger.info("开始执行 %s 扫描（超时: %d秒）...", tool_name, config_timeout)
+        user_log(scan_id, "port_scan", f"Running {tool_name}: {command}")
         
         try:
             # 直接调用 task（串行执行）
@@ -286,26 +287,31 @@ def _run_scans_sequentially(
                 'result': result,
                 'timeout': config_timeout
             }
-            processed_records += result.get('processed_records', 0)
+            tool_records = result.get('processed_records', 0)
+            processed_records += tool_records
             logger.info(
                 "✓ 工具 %s 流式处理完成 - 记录数: %d",
-                tool_name, result.get('processed_records', 0)
+                tool_name, tool_records
             )
+            user_log(scan_id, "port_scan", f"{tool_name} completed: found {tool_records} ports")
             
         except subprocess.TimeoutExpired as exc:
             # 超时异常单独处理
             # 注意：流式处理任务超时时，已解析的数据已保存到数据库
-            reason = f"执行超时（配置: {config_timeout}秒）"
+            reason = f"timeout after {config_timeout}s"
             failed_tools.append({'tool': tool_name, 'reason': reason})
             logger.warning(
                 "⚠️ 工具 %s 执行超时 - 超时配置: %d秒\n"
                 "注意：超时前已解析的端口数据已保存到数据库，但扫描未完全完成。",
                 tool_name, config_timeout
             )
+            user_log(scan_id, "port_scan", f"{tool_name} failed: {reason}", "error")
         except Exception as exc:
             # 其他异常
-            failed_tools.append({'tool': tool_name, 'reason': str(exc)})
+            reason = str(exc)
+            failed_tools.append({'tool': tool_name, 'reason': reason})
             logger.error("工具 %s 执行失败: %s", tool_name, exc, exc_info=True)
+            user_log(scan_id, "port_scan", f"{tool_name} failed: {reason}", "error")
     
     if failed_tools:
         logger.warning(
@@ -420,6 +426,8 @@ def port_scan_flow(
             "="*60
         )
         
+        user_log(scan_id, "port_scan", "Starting port scan")
+        
         # Step 0: 创建工作目录
         from apps.scan.utils import setup_scan_directory
         port_scan_dir = setup_scan_directory(scan_workspace_dir, 'port_scan')
@@ -428,7 +436,8 @@ def port_scan_flow(
         targets_file, target_count, target_type = _export_scan_targets(target_id, port_scan_dir)
         
         if target_count == 0:
-            logger.warning("目标下没有可扫描的地址，跳过端口扫描")
+            logger.warning("跳过端口扫描：没有目标可扫描 - Scan ID: %s", scan_id)
+            user_log(scan_id, "port_scan", "Skipped: no targets to scan", "warning")
             return {
                 'success': True,
                 'scan_id': scan_id,
@@ -467,7 +476,9 @@ def port_scan_flow(
             target_name=target_name
         )
         
-        logger.info("="*60 + "\n✓ 端口扫描完成\n" + "="*60)
+        # 记录 Flow 完成
+        logger.info("✓ 端口扫描完成 - 发现端口: %d", processed_records)
+        user_log(scan_id, "port_scan", f"port_scan completed: found {processed_records} ports")
         
         # 动态生成已执行的任务列表
         executed_tasks = ['export_scan_targets', 'parse_config']

backend/apps/scan/flows/site_scan_flow.py

@@ -17,6 +17,7 @@ from apps.common.prefect_django_setup import setup_django_for_prefect
 import logging
 import os
 import subprocess
+import time
 from pathlib import Path
 from typing import Callable
 from prefect import flow
@@ -26,7 +27,7 @@ from apps.scan.handlers.scan_flow_handlers import (
     on_scan_flow_completed,
     on_scan_flow_failed,
 )
-from apps.scan.utils import config_parser, build_scan_command
+from apps.scan.utils import config_parser, build_scan_command, user_log
 
 logger = logging.getLogger(__name__)
 
@@ -198,20 +199,20 @@ def _run_scans_sequentially(
             "开始执行 %s 站点扫描 - URL数: %d, 最终超时: %ds",
             tool_name, total_urls, timeout
         )
+        user_log(scan_id, "site_scan", f"Running {tool_name}: {command}")
         
         # 3. 执行扫描任务
         try:
             # 流式执行扫描并实时保存结果
             result = run_and_stream_save_websites_task(
                 cmd=command,
-                tool_name=tool_name,  # 新增：工具名称
+                tool_name=tool_name,
                 scan_id=scan_id,
                 target_id=target_id,
                 cwd=str(site_scan_dir),
                 shell=True,
-                batch_size=1000,
                 timeout=timeout,
-                log_file=str(log_file)  # 新增：日志文件路径
+                log_file=str(log_file)
             )
             
             tool_stats[tool_name] = {
@@ -219,29 +220,35 @@ def _run_scans_sequentially(
                 'result': result,
                 'timeout': timeout
             }
-            processed_records += result.get('processed_records', 0)
+            tool_records = result.get('processed_records', 0)
+            tool_created = result.get('created_websites', 0)
+            processed_records += tool_records
             
             logger.info(
                 "✓ 工具 %s 流式处理完成 - 处理记录: %d, 创建站点: %d, 跳过: %d",
                 tool_name,
-                result.get('processed_records', 0),
-                result.get('created_websites', 0),
+                tool_records,
+                tool_created,
                 result.get('skipped_no_subdomain', 0) + result.get('skipped_failed', 0)
             )
+            user_log(scan_id, "site_scan", f"{tool_name} completed: found {tool_created} websites")
             
         except subprocess.TimeoutExpired as exc:
             # 超时异常单独处理
-            reason = f"执行超时（配置: {timeout}秒）"
+            reason = f"timeout after {timeout}s"
             failed_tools.append({'tool': tool_name, 'reason': reason})
             logger.warning(
                 "⚠️ 工具 %s 执行超时 - 超时配置: %d秒\n"
                 "注意：超时前已解析的站点数据已保存到数据库，但扫描未完全完成。",
                 tool_name, timeout
             )
+            user_log(scan_id, "site_scan", f"{tool_name} failed: {reason}", "error")
         except Exception as exc:
             # 其他异常
-            failed_tools.append({'tool': tool_name, 'reason': str(exc)})
+            reason = str(exc)
+            failed_tools.append({'tool': tool_name, 'reason': reason})
             logger.error("工具 %s 执行失败: %s", tool_name, exc, exc_info=True)
+            user_log(scan_id, "site_scan", f"{tool_name} failed: {reason}", "error")
     
     if failed_tools:
         logger.warning(
@@ -380,6 +387,8 @@ def site_scan_flow(
         if not scan_workspace_dir:
             raise ValueError("scan_workspace_dir 不能为空")
         
+        user_log(scan_id, "site_scan", "Starting site scan")
+        
         # Step 0: 创建工作目录
         from apps.scan.utils import setup_scan_directory
         site_scan_dir = setup_scan_directory(scan_workspace_dir, 'site_scan')
@@ -390,7 +399,8 @@ def site_scan_flow(
         )
         
         if total_urls == 0:
-            logger.warning("目标下没有可用的站点URL，跳过站点扫描")
+            logger.warning("跳过站点扫描：没有站点 URL 可扫描 - Scan ID: %s", scan_id)
+            user_log(scan_id, "site_scan", "Skipped: no site URLs to scan", "warning")
             return {
                 'success': True,
                 'scan_id': scan_id,
@@ -433,8 +443,6 @@ def site_scan_flow(
             target_name=target_name
         )
         
-        logger.info("="*60 + "\n✓ 站点扫描完成\n" + "="*60)
-        
         # 动态生成已执行的任务列表
         executed_tasks = ['export_site_urls', 'parse_config']
         executed_tasks.extend([f'run_and_stream_save_websites ({tool})' for tool in tool_stats.keys()])
@@ -444,6 +452,10 @@ def site_scan_flow(
         total_skipped_no_subdomain = sum(stats['result'].get('skipped_no_subdomain', 0) for stats in tool_stats.values())
         total_skipped_failed = sum(stats['result'].get('skipped_failed', 0) for stats in tool_stats.values())
         
+        # 记录 Flow 完成
+        logger.info("✓ 站点扫描完成 - 创建站点: %d", total_created)
+        user_log(scan_id, "site_scan", f"site_scan completed: found {total_created} websites")
+        
         return {
             'success': True,
             'scan_id': scan_id,

backend/apps/scan/flows/subdomain_discovery_flow.py

@@ -30,7 +30,7 @@ from apps.scan.handlers.scan_flow_handlers import (
     on_scan_flow_completed,
     on_scan_flow_failed,
 )
-from apps.scan.utils import build_scan_command, ensure_wordlist_local
+from apps.scan.utils import build_scan_command, ensure_wordlist_local, user_log
 from apps.engine.services.wordlist_service import WordlistService
 from apps.common.normalizer import normalize_domain
 from apps.common.validators import validate_domain
@@ -77,7 +77,8 @@ def _validate_and_normalize_target(target_name: str) -> str:
 def _run_scans_parallel(
     enabled_tools: dict,
     domain_name: str,
-    result_dir: Path
+    result_dir: Path,
+    scan_id: int
 ) -> tuple[list, list, list]:
     """
     并行运行所有启用的子域名扫描工具
@@ -86,6 +87,7 @@ def _run_scans_parallel(
         enabled_tools: 启用的工具配置字典 {'tool_name': {'timeout': 600, ...}}
         domain_name: 目标域名
         result_dir: 结果输出目录
+        scan_id: 扫描任务 ID（用于记录日志）
         
     Returns:
         tuple: (result_files, failed_tools, successful_tool_names)
@@ -137,6 +139,9 @@ def _run_scans_parallel(
             f"提交任务 - 工具: {tool_name}, 超时: {timeout}s, 输出: {output_file}"
         )
         
+        # 记录工具开始执行日志
+        user_log(scan_id, "subdomain_discovery", f"Running {tool_name}: {command}")
+        
         future = run_subdomain_discovery_task.submit(
             tool=tool_name,
             command=command,
@@ -164,16 +169,19 @@ def _run_scans_parallel(
             if result:
                 result_files.append(result)
                 logger.info("✓ 扫描工具 %s 执行成功: %s", tool_name, result)
+                user_log(scan_id, "subdomain_discovery", f"{tool_name} completed")
             else:
                 failure_msg = f"{tool_name}: 未生成结果文件"
                 failures.append(failure_msg)
                 failed_tools.append({'tool': tool_name, 'reason': '未生成结果文件'})
                 logger.warning("⚠️ 扫描工具 %s 未生成结果文件", tool_name)
+                user_log(scan_id, "subdomain_discovery", f"{tool_name} failed: no output file", "error")
         except Exception as e:
             failure_msg = f"{tool_name}: {str(e)}"
             failures.append(failure_msg)
             failed_tools.append({'tool': tool_name, 'reason': str(e)})
             logger.warning("⚠️ 扫描工具 %s 执行失败: %s", tool_name, str(e))
+            user_log(scan_id, "subdomain_discovery", f"{tool_name} failed: {str(e)}", "error")
     
     # 4. 检查是否有成功的工具
     if not result_files:
@@ -203,7 +211,8 @@ def _run_single_tool(
     tool_config: dict,
     command_params: dict,
     result_dir: Path,
-    scan_type: str = 'subdomain_discovery'
+    scan_type: str = 'subdomain_discovery',
+    scan_id: int = None
 ) -> str:
     """
     运行单个扫描工具
@@ -214,6 +223,7 @@ def _run_single_tool(
         command_params: 命令参数
         result_dir: 结果目录
         scan_type: 扫描类型
+        scan_id: 扫描 ID（用于记录用户日志）
         
     Returns:
         str: 输出文件路径，失败返回空字符串
@@ -242,7 +252,9 @@ def _run_single_tool(
     if timeout == 'auto':
         timeout = 3600
     
-    logger.info(f"执行 {tool_name}: timeout={timeout}s")
+    logger.info(f"执行 {tool_name}: {command}")
+    if scan_id:
+        user_log(scan_id, scan_type, f"Running {tool_name}: {command}")
     
     try:
         result = run_subdomain_discovery_task(
@@ -401,7 +413,6 @@ def subdomain_discovery_flow(
             logger.warning("目标域名无效，跳过子域名发现扫描: %s", e)
             return _empty_result(scan_id, target_name, scan_workspace_dir)
         
-        # 验证成功后打印日志
         logger.info(
             "="*60 + "\n" +
             "开始子域名发现扫描\n" +
@@ -410,6 +421,7 @@ def subdomain_discovery_flow(
             f"  Workspace: {scan_workspace_dir}\n" +
             "="*60
         )
+        user_log(scan_id, "subdomain_discovery", f"Starting subdomain discovery for {domain_name}")
         
         # 解析配置
         passive_tools = scan_config.get('passive_tools', {})
@@ -429,23 +441,22 @@ def subdomain_discovery_flow(
         successful_tool_names = []
         
         # ==================== Stage 1: 被动收集（并行）====================
-        logger.info("=" * 40)
-        logger.info("Stage 1: 被动收集（并行）")
-        logger.info("=" * 40)
-        
         if enabled_passive_tools:
+            logger.info("=" * 40)
+            logger.info("Stage 1: 被动收集（并行）")
+            logger.info("=" * 40)
             logger.info("启用工具: %s", ', '.join(enabled_passive_tools.keys()))
+            user_log(scan_id, "subdomain_discovery", f"Stage 1: passive collection ({', '.join(enabled_passive_tools.keys())})")
             result_files, stage1_failed, stage1_success = _run_scans_parallel(
                 enabled_tools=enabled_passive_tools,
                 domain_name=domain_name,
-                result_dir=result_dir
+                result_dir=result_dir,
+                scan_id=scan_id
             )
             all_result_files.extend(result_files)
             failed_tools.extend(stage1_failed)
             successful_tool_names.extend(stage1_success)
             executed_tasks.extend([f'passive ({tool})' for tool in stage1_success])
-        else:
-            logger.warning("未启用任何被动收集工具")
         
         # 合并 Stage 1 结果
         timestamp = datetime.now().strftime('%Y%m%d_%H%M%S')
@@ -456,7 +467,6 @@ def subdomain_discovery_flow(
         else:
             # 创建空文件
             Path(current_result).touch()
-            logger.warning("Stage 1 无结果，创建空文件")
         
         # ==================== Stage 2: 字典爆破（可选）====================
         bruteforce_enabled = bruteforce_config.get('enabled', False)
@@ -464,6 +474,7 @@ def subdomain_discovery_flow(
             logger.info("=" * 40)
             logger.info("Stage 2: 字典爆破")
             logger.info("=" * 40)
+            user_log(scan_id, "subdomain_discovery", "Stage 2: bruteforce")
             
             bruteforce_tool_config = bruteforce_config.get('subdomain_bruteforce', {})
             wordlist_name = bruteforce_tool_config.get('wordlist_name', 'dns_wordlist.txt')
@@ -496,22 +507,16 @@ def subdomain_discovery_flow(
                         **bruteforce_tool_config,
                         'timeout': timeout_value,
                     }
-                    logger.info(
-                        "subdomain_bruteforce 使用自动 timeout: %s 秒 (字典行数=%s, 3秒/行)",
-                        timeout_value,
-                        line_count_int,
-                    )
 
-                brute_output = str(result_dir / f"subs_brute_{timestamp}.txt")
                 brute_result = _run_single_tool(
                     tool_name='subdomain_bruteforce',
                     tool_config=bruteforce_tool_config,
                     command_params={
                         'domain': domain_name,
                         'wordlist': local_wordlist_path,
-                        'output_file': brute_output
                     },
-                    result_dir=result_dir
+                    result_dir=result_dir,
+                    scan_id=scan_id
                 )
                 
                 if brute_result:
@@ -522,11 +527,16 @@ def subdomain_discovery_flow(
                     )
                     successful_tool_names.append('subdomain_bruteforce')
                     executed_tasks.append('bruteforce')
+                    logger.info("✓ subdomain_bruteforce 执行完成")
+                    user_log(scan_id, "subdomain_discovery", "subdomain_bruteforce completed")
                 else:
                     failed_tools.append({'tool': 'subdomain_bruteforce', 'reason': '执行失败'})
+                    logger.warning("⚠️ subdomain_bruteforce 执行失败")
+                    user_log(scan_id, "subdomain_discovery", "subdomain_bruteforce failed: execution failed", "error")
             except Exception as exc:
-                logger.warning("字典准备失败，跳过字典爆破: %s", exc)
                 failed_tools.append({'tool': 'subdomain_bruteforce', 'reason': str(exc)})
+                logger.warning("字典准备失败，跳过字典爆破: %s", exc)
+                user_log(scan_id, "subdomain_discovery", f"subdomain_bruteforce failed: {str(exc)}", "error")
         
         # ==================== Stage 3: 变异生成 + 验证（可选）====================
         permutation_enabled = permutation_config.get('enabled', False)
@@ -534,6 +544,7 @@ def subdomain_discovery_flow(
             logger.info("=" * 40)
             logger.info("Stage 3: 变异生成 + 存活验证（流式管道）")
             logger.info("=" * 40)
+            user_log(scan_id, "subdomain_discovery", "Stage 3: permutation + resolve")
             
             permutation_tool_config = permutation_config.get('subdomain_permutation_resolve', {})
             
@@ -587,20 +598,19 @@ def subdomain_discovery_flow(
                         'tool': 'subdomain_permutation_resolve',
                         'reason': f"采样检测到泛解析 (膨胀率 {ratio:.1f}x)"
                     })
+                    user_log(scan_id, "subdomain_discovery", f"subdomain_permutation_resolve skipped: wildcard detected (ratio {ratio:.1f}x)", "warning")
                 else:
                     # === Step 3.2: 采样通过，执行完整变异 ===
                     logger.info("采样检测通过，执行完整变异...")
                     
-                    permuted_output = str(result_dir / f"subs_permuted_{timestamp}.txt")
-                    
                     permuted_result = _run_single_tool(
                         tool_name='subdomain_permutation_resolve',
                         tool_config=permutation_tool_config,
                         command_params={
                             'input_file': current_result,
-                            'output_file': permuted_output,
                         },
-                        result_dir=result_dir
+                        result_dir=result_dir,
+                        scan_id=scan_id
                     )
                     
                     if permuted_result:
@@ -611,15 +621,21 @@ def subdomain_discovery_flow(
                         )
                         successful_tool_names.append('subdomain_permutation_resolve')
                         executed_tasks.append('permutation')
+                        logger.info("✓ subdomain_permutation_resolve 执行完成")
+                        user_log(scan_id, "subdomain_discovery", "subdomain_permutation_resolve completed")
                     else:
                         failed_tools.append({'tool': 'subdomain_permutation_resolve', 'reason': '执行失败'})
+                        logger.warning("⚠️ subdomain_permutation_resolve 执行失败")
+                        user_log(scan_id, "subdomain_discovery", "subdomain_permutation_resolve failed: execution failed", "error")
                         
             except subprocess.TimeoutExpired:
-                logger.warning(f"采样检测超时 ({SAMPLE_TIMEOUT}秒)，跳过变异")
                 failed_tools.append({'tool': 'subdomain_permutation_resolve', 'reason': '采样检测超时'})
+                logger.warning(f"采样检测超时 ({SAMPLE_TIMEOUT}秒)，跳过变异")
+                user_log(scan_id, "subdomain_discovery", "subdomain_permutation_resolve failed: sample detection timeout", "error")
             except Exception as e:
-                logger.warning(f"采样检测失败: {e}，跳过变异")
                 failed_tools.append({'tool': 'subdomain_permutation_resolve', 'reason': f'采样检测失败: {e}'})
+                logger.warning(f"采样检测失败: {e}，跳过变异")
+                user_log(scan_id, "subdomain_discovery", f"subdomain_permutation_resolve failed: {str(e)}", "error")
         
         # ==================== Stage 4: DNS 存活验证（可选）====================
         # 无论是否启用 Stage 3，只要 resolve.enabled 为 true 就会执行，对当前所有候选子域做统一 DNS 验证
@@ -628,6 +644,7 @@ def subdomain_discovery_flow(
             logger.info("=" * 40)
             logger.info("Stage 4: DNS 存活验证")
             logger.info("=" * 40)
+            user_log(scan_id, "subdomain_discovery", "Stage 4: DNS resolve")
             
             resolve_tool_config = resolve_config.get('subdomain_resolve', {})
 
@@ -651,30 +668,27 @@ def subdomain_discovery_flow(
                     **resolve_tool_config,
                     'timeout': timeout_value,
                 }
-                logger.info(
-                    "subdomain_resolve 使用自动 timeout: %s 秒 (候选子域数=%s, 3秒/域名)",
-                    timeout_value,
-                    line_count_int,
-                )
 
-            alive_output = str(result_dir / f"subs_alive_{timestamp}.txt")
-            
             alive_result = _run_single_tool(
                 tool_name='subdomain_resolve',
                 tool_config=resolve_tool_config,
                 command_params={
                     'input_file': current_result,
-                    'output_file': alive_output,
                 },
-                result_dir=result_dir
+                result_dir=result_dir,
+                scan_id=scan_id
             )
             
             if alive_result:
                 current_result = alive_result
                 successful_tool_names.append('subdomain_resolve')
                 executed_tasks.append('resolve')
+                logger.info("✓ subdomain_resolve 执行完成")
+                user_log(scan_id, "subdomain_discovery", "subdomain_resolve completed")
             else:
                 failed_tools.append({'tool': 'subdomain_resolve', 'reason': '执行失败'})
+                logger.warning("⚠️ subdomain_resolve 执行失败")
+                user_log(scan_id, "subdomain_discovery", "subdomain_resolve failed: execution failed", "error")
         
         # ==================== Final: 保存到数据库 ====================
         logger.info("=" * 40)
@@ -695,7 +709,9 @@ def subdomain_discovery_flow(
         processed_domains = save_result.get('processed_records', 0)
         executed_tasks.append('save_domains')
         
+        # 记录 Flow 完成
         logger.info("="*60 + "\n✓ 子域名发现扫描完成\n" + "="*60)
+        user_log(scan_id, "subdomain_discovery", f"subdomain_discovery completed: found {processed_domains} subdomains")
         
         return {
             'success': True,

backend/apps/scan/flows/url_fetch/domain_name_url_fetch_flow.py

@@ -59,6 +59,8 @@ def domain_name_url_fetch_flow(
     - IP 和 CIDR 类型会自动跳过（waymore 等工具不支持）
     - 工具会自动收集 *.target_name 的所有历史 URL，无需遍历子域名
     """
+    from apps.scan.utils import user_log
+    
     try:
         output_path = Path(output_dir)
         output_path.mkdir(parents=True, exist_ok=True)
@@ -145,6 +147,9 @@ def domain_name_url_fetch_flow(
                 timeout,
             )
 
+            # 记录工具开始执行日志
+            user_log(scan_id, "url_fetch", f"Running {tool_name}: {command}")
+
             future = run_url_fetcher_task.submit(
                 tool_name=tool_name,
                 command=command,
@@ -163,22 +168,28 @@ def domain_name_url_fetch_flow(
                 if result and result.get("success"):
                     result_files.append(result["output_file"])
                     successful_tools.append(tool_name)
+                    url_count = result.get("url_count", 0)
                     logger.info(
                         "✓ 工具 %s 执行成功 - 发现 URL: %d",
                         tool_name,
-                        result.get("url_count", 0),
+                        url_count,
                     )
+                    user_log(scan_id, "url_fetch", f"{tool_name} completed: found {url_count} urls")
                 else:
+                    reason = "未生成结果或无有效 URL"
                     failed_tools.append(
                         {
                             "tool": tool_name,
-                            "reason": "未生成结果或无有效 URL",
+                            "reason": reason,
                         }
                     )
                     logger.warning("⚠️ 工具 %s 未生成有效结果", tool_name)
+                    user_log(scan_id, "url_fetch", f"{tool_name} failed: {reason}", "error")
             except Exception as e:
-                failed_tools.append({"tool": tool_name, "reason": str(e)})
+                reason = str(e)
+                failed_tools.append({"tool": tool_name, "reason": reason})
                 logger.warning("⚠️ 工具 %s 执行失败: %s", tool_name, e)
+                user_log(scan_id, "url_fetch", f"{tool_name} failed: {reason}", "error")
 
         logger.info(
             "基于 domain_name 的 URL 获取完成 - 成功工具: %s, 失败工具: %s",

backend/apps/scan/flows/url_fetch/main_flow.py

@@ -25,6 +25,7 @@ from apps.scan.handlers.scan_flow_handlers import (
     on_scan_flow_completed,
     on_scan_flow_failed,
 )
+from apps.scan.utils import user_log
 
 from .domain_name_url_fetch_flow import domain_name_url_fetch_flow
 from .sites_url_fetch_flow import sites_url_fetch_flow
@@ -212,7 +213,6 @@ def _validate_and_stream_save_urls(
             target_id=target_id,
             cwd=str(url_fetch_dir),
             shell=True,
-            batch_size=500,
             timeout=timeout,
             log_file=str(log_file)
         )
@@ -292,6 +292,8 @@ def url_fetch_flow(
             "="*60
         )
         
+        user_log(scan_id, "url_fetch", "Starting URL fetch")
+        
         # Step 1: 准备工作目录
         logger.info("Step 1: 准备工作目录")
         from apps.scan.utils import setup_scan_directory
@@ -404,7 +406,9 @@ def url_fetch_flow(
                 target_id=target_id
             )
         
-        logger.info("="*60 + "\n✓ URL 获取扫描完成\n" + "="*60)
+        # 记录 Flow 完成
+        logger.info("✓ URL 获取完成 - 保存 endpoints: %d", saved_count)
+        user_log(scan_id, "url_fetch", f"url_fetch completed: found {saved_count} endpoints")
         
         # 构建已执行的任务列表
         executed_tasks = ['setup_directory', 'classify_tools']

backend/apps/scan/flows/url_fetch/sites_url_fetch_flow.py

@@ -116,7 +116,8 @@ def sites_url_fetch_flow(
             tools=enabled_tools,
             input_file=sites_file,
             input_type="sites_file",
-            output_dir=output_path
+            output_dir=output_path,
+            scan_id=scan_id
         )
         
         logger.info(

backend/apps/scan/flows/url_fetch/utils.py

@@ -152,7 +152,8 @@ def run_tools_parallel(
     tools: dict,
     input_file: str,
     input_type: str,
-    output_dir: Path
+    output_dir: Path,
+    scan_id: int
 ) -> tuple[list, list, list]:
     """
     并行执行工具列表
@@ -162,11 +163,13 @@ def run_tools_parallel(
         input_file: 输入文件路径
         input_type: 输入类型
         output_dir: 输出目录
+        scan_id: 扫描任务 ID（用于记录日志）
         
     Returns:
         tuple: (result_files, failed_tools, successful_tool_names)
     """
     from apps.scan.tasks.url_fetch import run_url_fetcher_task
+    from apps.scan.utils import user_log
 
     futures: dict[str, object] = {}
     failed_tools: list[dict] = []
@@ -192,6 +195,9 @@ def run_tools_parallel(
             exec_params["timeout"],
         )
 
+        # 记录工具开始执行日志
+        user_log(scan_id, "url_fetch", f"Running {tool_name}: {exec_params['command']}")
+
         # 提交并行任务
         future = run_url_fetcher_task.submit(
             tool_name=tool_name,
@@ -208,22 +214,28 @@ def run_tools_parallel(
             result = future.result()
             if result and result['success']:
                 result_files.append(result['output_file'])
+                url_count = result['url_count']
                 logger.info(
                     "✓ 工具 %s 执行成功 - 发现 URL: %d",
-                    tool_name, result['url_count']
+                    tool_name, url_count
                 )
+                user_log(scan_id, "url_fetch", f"{tool_name} completed: found {url_count} urls")
             else:
+                reason = '未生成结果或无有效URL'
                 failed_tools.append({
                     'tool': tool_name,
-                    'reason': '未生成结果或无有效URL'
+                    'reason': reason
                 })
                 logger.warning("⚠️ 工具 %s 未生成有效结果", tool_name)
+                user_log(scan_id, "url_fetch", f"{tool_name} failed: {reason}", "error")
         except Exception as e:
+            reason = str(e)
             failed_tools.append({
                 'tool': tool_name,
-                'reason': str(e)
+                'reason': reason
             })
             logger.warning("⚠️ 工具 %s 执行失败: %s", tool_name, e)
+            user_log(scan_id, "url_fetch", f"{tool_name} failed: {reason}", "error")
 
     # 计算成功的工具列表
     failed_tool_names = [f['tool'] for f in failed_tools]

backend/apps/scan/flows/vuln_scan/endpoints_vuln_scan_flow.py

@@ -12,7 +12,7 @@ from apps.scan.handlers.scan_flow_handlers import (
     on_scan_flow_completed,
     on_scan_flow_failed,
 )
-from apps.scan.utils import build_scan_command, ensure_nuclei_templates_local
+from apps.scan.utils import build_scan_command, ensure_nuclei_templates_local, user_log
 from apps.scan.tasks.vuln_scan import (
     export_endpoints_task,
     run_vuln_tool_task,
@@ -141,6 +141,7 @@ def endpoints_vuln_scan_flow(
             # Dalfox XSS 使用流式任务，一边解析一边保存漏洞结果
             if tool_name == "dalfox_xss":
                 logger.info("开始执行漏洞扫描工具 %s（流式保存漏洞结果，已提交任务）", tool_name)
+                user_log(scan_id, "vuln_scan", f"Running {tool_name}: {command}")
                 future = run_and_stream_save_dalfox_vulns_task.submit(
                     cmd=command,
                     tool_name=tool_name,
@@ -163,6 +164,7 @@ def endpoints_vuln_scan_flow(
             elif tool_name == "nuclei":
                 # Nuclei 使用流式任务
                 logger.info("开始执行漏洞扫描工具 %s（流式保存漏洞结果，已提交任务）", tool_name)
+                user_log(scan_id, "vuln_scan", f"Running {tool_name}: {command}")
                 future = run_and_stream_save_nuclei_vulns_task.submit(
                     cmd=command,
                     tool_name=tool_name,
@@ -185,6 +187,7 @@ def endpoints_vuln_scan_flow(
             else:
                 # 其他工具仍使用非流式执行逻辑
                 logger.info("开始执行漏洞扫描工具 %s（已提交任务）", tool_name)
+                user_log(scan_id, "vuln_scan", f"Running {tool_name}: {command}")
                 future = run_vuln_tool_task.submit(
                     tool_name=tool_name,
                     command=command,
@@ -203,24 +206,34 @@ def endpoints_vuln_scan_flow(
         # 统一收集所有工具的执行结果
         for tool_name, meta in tool_futures.items():
             future = meta["future"]
-            result = future.result()
+            try:
+                result = future.result()
 
-            if meta["mode"] == "streaming":
-                tool_results[tool_name] = {
-                    "command": meta["command"],
-                    "timeout": meta["timeout"],
-                    "processed_records": result.get("processed_records"),
-                    "created_vulns": result.get("created_vulns"),
-                    "command_log_file": meta["log_file"],
-                }
-            else:
-                tool_results[tool_name] = {
-                    "command": meta["command"],
-                    "timeout": meta["timeout"],
-                    "duration": result.get("duration"),
-                    "returncode": result.get("returncode"),
-                    "command_log_file": result.get("command_log_file"),
-                }
+                if meta["mode"] == "streaming":
+                    created_vulns = result.get("created_vulns", 0)
+                    tool_results[tool_name] = {
+                        "command": meta["command"],
+                        "timeout": meta["timeout"],
+                        "processed_records": result.get("processed_records"),
+                        "created_vulns": created_vulns,
+                        "command_log_file": meta["log_file"],
+                    }
+                    logger.info("✓ 工具 %s 执行完成 - 漏洞: %d", tool_name, created_vulns)
+                    user_log(scan_id, "vuln_scan", f"{tool_name} completed: found {created_vulns} vulnerabilities")
+                else:
+                    tool_results[tool_name] = {
+                        "command": meta["command"],
+                        "timeout": meta["timeout"],
+                        "duration": result.get("duration"),
+                        "returncode": result.get("returncode"),
+                        "command_log_file": result.get("command_log_file"),
+                    }
+                    logger.info("✓ 工具 %s 执行完成 - returncode=%s", tool_name, result.get("returncode"))
+                    user_log(scan_id, "vuln_scan", f"{tool_name} completed")
+            except Exception as e:
+                reason = str(e)
+                logger.error("工具 %s 执行失败: %s", tool_name, e, exc_info=True)
+                user_log(scan_id, "vuln_scan", f"{tool_name} failed: {reason}", "error")
 
         return {
             "success": True,

backend/apps/scan/flows/vuln_scan/main_flow.py

@@ -11,6 +11,7 @@ from apps.scan.handlers.scan_flow_handlers import (
     on_scan_flow_failed,
 )
 from apps.scan.configs.command_templates import get_command_template
+from apps.scan.utils import user_log
 from .endpoints_vuln_scan_flow import endpoints_vuln_scan_flow
 
 
@@ -72,6 +73,9 @@ def vuln_scan_flow(
         if not enabled_tools:
             raise ValueError("enabled_tools 不能为空")
 
+        logger.info("开始漏洞扫描 - Scan ID: %s, Target: %s", scan_id, target_name)
+        user_log(scan_id, "vuln_scan", "Starting vulnerability scan")
+
         # Step 1: 分类工具
         endpoints_tools, other_tools = _classify_vuln_tools(enabled_tools)
 
@@ -99,6 +103,14 @@ def vuln_scan_flow(
             enabled_tools=endpoints_tools,
         )
 
+        # 记录 Flow 完成
+        total_vulns = sum(
+            r.get("created_vulns", 0) 
+            for r in endpoint_result.get("tool_results", {}).values()
+        )
+        logger.info("✓ 漏洞扫描完成 - 新增漏洞: %d", total_vulns)
+        user_log(scan_id, "vuln_scan", f"vuln_scan completed: found {total_vulns} vulnerabilities")
+
         # 目前只有一个子 Flow，直接返回其结果
         return endpoint_result
 

backend/apps/scan/handlers/initiate_scan_flow_handlers.py

@@ -162,6 +162,8 @@ def on_initiate_scan_flow_completed(flow: Flow, flow_run: FlowRun, state: State)
     # 执行状态更新并获取统计数据
     stats = _update_completed_status()
     
+    # 注意：物化视图刷新已迁移到 pg_ivm 增量维护，无需手动标记刷新
+    
     # 发送通知（包含统计摘要）
     logger.info("准备发送扫描完成通知 - Scan ID: %s, Target: %s", scan_id, target_name)
     try:

backend/apps/scan/handlers/scan_flow_handlers.py

@@ -14,6 +14,7 @@ from prefect import Flow
 from prefect.client.schemas import FlowRun, State
 
 from apps.scan.utils.performance import FlowPerformanceTracker
+from apps.scan.utils import user_log
 
 logger = logging.getLogger(__name__)
 
@@ -136,6 +137,7 @@ def on_scan_flow_failed(flow: Flow, flow_run: FlowRun, state: State) -> None:
     - 更新阶段进度为 failed
     - 发送扫描失败通知
     - 记录性能指标（含错误信息）
+    - 写入 ScanLog 供前端显示
     
     Args:
         flow: Prefect Flow 对象
@@ -152,6 +154,11 @@ def on_scan_flow_failed(flow: Flow, flow_run: FlowRun, state: State) -> None:
     # 提取错误信息
     error_message = str(state.message) if state.message else "未知错误"
     
+    # 写入 ScanLog 供前端显示
+    stage = _get_stage_from_flow_name(flow.name)
+    if scan_id and stage:
+        user_log(scan_id, stage, f"Failed: {error_message}", "error")
+    
     # 记录性能指标（失败情况）
     tracker = _flow_trackers.pop(str(flow_run.id), None)
     if tracker:

backend/apps/scan/migrations/0001_initial.py

@@ -0,0 +1,136 @@
+# Generated by Django 5.2.7 on 2026-01-02 04:45
+
+import django.contrib.postgres.fields
+import django.db.models.deletion
+from django.db import migrations, models
+
+
+class Migration(migrations.Migration):
+
+    initial = True
+
+    dependencies = [
+        ('engine', '0001_initial'),
+        ('targets', '0001_initial'),
+    ]
+
+    operations = [
+        migrations.CreateModel(
+            name='NotificationSettings',
+            fields=[
+                ('id', models.BigAutoField(auto_created=True, primary_key=True, serialize=False, verbose_name='ID')),
+                ('discord_enabled', models.BooleanField(default=False, help_text='是否启用 Discord 通知')),
+                ('discord_webhook_url', models.URLField(blank=True, default='', help_text='Discord Webhook URL')),
+                ('categories', models.JSONField(default=dict, help_text='各分类通知开关，如 {"scan": true, "vulnerability": true, "asset": true, "system": false}')),
+                ('created_at', models.DateTimeField(auto_now_add=True)),
+                ('updated_at', models.DateTimeField(auto_now=True)),
+            ],
+            options={
+                'verbose_name': '通知设置',
+                'verbose_name_plural': '通知设置',
+                'db_table': 'notification_settings',
+            },
+        ),
+        migrations.CreateModel(
+            name='Notification',
+            fields=[
+                ('id', models.AutoField(primary_key=True, serialize=False)),
+                ('category', models.CharField(choices=[('scan', '扫描任务'), ('vulnerability', '漏洞发现'), ('asset', '资产发现'), ('system', '系统消息')], db_index=True, default='system', help_text='通知分类', max_length=20)),
+                ('level', models.CharField(choices=[('low', '低'), ('medium', '中'), ('high', '高'), ('critical', '严重')], db_index=True, default='low', help_text='通知级别', max_length=20)),
+                ('title', models.CharField(help_text='通知标题', max_length=200)),
+                ('message', models.CharField(help_text='通知内容', max_length=2000)),
+                ('created_at', models.DateTimeField(auto_now_add=True, help_text='创建时间')),
+                ('is_read', models.BooleanField(default=False, help_text='是否已读')),
+                ('read_at', models.DateTimeField(blank=True, help_text='阅读时间', null=True)),
+            ],
+            options={
+                'verbose_name': '通知',
+                'verbose_name_plural': '通知',
+                'db_table': 'notification',
+                'ordering': ['-created_at'],
+                'indexes': [models.Index(fields=['-created_at'], name='notificatio_created_c430f0_idx'), models.Index(fields=['category', '-created_at'], name='notificatio_categor_df0584_idx'), models.Index(fields=['level', '-created_at'], name='notificatio_level_0e5d12_idx'), models.Index(fields=['is_read', '-created_at'], name='notificatio_is_read_518ce0_idx')],
+            },
+        ),
+        migrations.CreateModel(
+            name='Scan',
+            fields=[
+                ('id', models.AutoField(primary_key=True, serialize=False)),
+                ('engine_ids', django.contrib.postgres.fields.ArrayField(base_field=models.IntegerField(), default=list, help_text='引擎 ID 列表', size=None)),
+                ('engine_names', models.JSONField(default=list, help_text='引擎名称列表，如 ["引擎A", "引擎B"]')),
+                ('yaml_configuration', models.TextField(default='', help_text='YAML 格式的扫描配置')),
+                ('created_at', models.DateTimeField(auto_now_add=True, help_text='任务创建时间')),
+                ('stopped_at', models.DateTimeField(blank=True, help_text='扫描结束时间', null=True)),
+                ('status', models.CharField(choices=[('cancelled', '已取消'), ('completed', '已完成'), ('failed', '失败'), ('initiated', '初始化'), ('running', '运行中')], db_index=True, default='initiated', help_text='任务状态', max_length=20)),
+                ('results_dir', models.CharField(blank=True, default='', help_text='结果存储目录', max_length=100)),
+                ('container_ids', django.contrib.postgres.fields.ArrayField(base_field=models.CharField(max_length=100), blank=True, default=list, help_text='容器 ID 列表（Docker Container ID）', size=None)),
+                ('error_message', models.CharField(blank=True, default='', help_text='错误信息', max_length=2000)),
+                ('deleted_at', models.DateTimeField(blank=True, db_index=True, help_text='删除时间（NULL表示未删除）', null=True)),
+                ('progress', models.IntegerField(default=0, help_text='扫描进度 0-100')),
+                ('current_stage', models.CharField(blank=True, default='', help_text='当前扫描阶段', max_length=50)),
+                ('stage_progress', models.JSONField(default=dict, help_text='各阶段进度详情')),
+                ('cached_subdomains_count', models.IntegerField(default=0, help_text='缓存的子域名数量')),
+                ('cached_websites_count', models.IntegerField(default=0, help_text='缓存的网站数量')),
+                ('cached_endpoints_count', models.IntegerField(default=0, help_text='缓存的端点数量')),
+                ('cached_ips_count', models.IntegerField(default=0, help_text='缓存的IP地址数量')),
+                ('cached_directories_count', models.IntegerField(default=0, help_text='缓存的目录数量')),
+                ('cached_vulns_total', models.IntegerField(default=0, help_text='缓存的漏洞总数')),
+                ('cached_vulns_critical', models.IntegerField(default=0, help_text='缓存的严重漏洞数量')),
+                ('cached_vulns_high', models.IntegerField(default=0, help_text='缓存的高危漏洞数量')),
+                ('cached_vulns_medium', models.IntegerField(default=0, help_text='缓存的中危漏洞数量')),
+                ('cached_vulns_low', models.IntegerField(default=0, help_text='缓存的低危漏洞数量')),
+                ('stats_updated_at', models.DateTimeField(blank=True, help_text='统计数据最后更新时间', null=True)),
+                ('target', models.ForeignKey(help_text='扫描目标', on_delete=django.db.models.deletion.CASCADE, related_name='scans', to='targets.target')),
+                ('worker', models.ForeignKey(blank=True, help_text='执行扫描的 Worker 节点', null=True, on_delete=django.db.models.deletion.SET_NULL, related_name='scans', to='engine.workernode')),
+            ],
+            options={
+                'verbose_name': '扫描任务',
+                'verbose_name_plural': '扫描任务',
+                'db_table': 'scan',
+                'ordering': ['-created_at'],
+                'indexes': [models.Index(fields=['-created_at'], name='scan_created_0bb6c7_idx'), models.Index(fields=['target'], name='scan_target__718b9d_idx'), models.Index(fields=['deleted_at', '-created_at'], name='scan_deleted_eb17e8_idx')],
+            },
+        ),
+        migrations.CreateModel(
+            name='ScheduledScan',
+            fields=[
+                ('id', models.AutoField(primary_key=True, serialize=False)),
+                ('name', models.CharField(help_text='任务名称', max_length=200)),
+                ('engine_ids', django.contrib.postgres.fields.ArrayField(base_field=models.IntegerField(), default=list, help_text='引擎 ID 列表', size=None)),
+                ('engine_names', models.JSONField(default=list, help_text='引擎名称列表，如 ["引擎A", "引擎B"]')),
+                ('yaml_configuration', models.TextField(default='', help_text='YAML 格式的扫描配置')),
+                ('cron_expression', models.CharField(default='0 2 * * *', help_text='Cron 表达式，格式：分 时 日 月 周', max_length=100)),
+                ('is_enabled', models.BooleanField(db_index=True, default=True, help_text='是否启用')),
+                ('run_count', models.IntegerField(default=0, help_text='已执行次数')),
+                ('last_run_time', models.DateTimeField(blank=True, help_text='上次执行时间', null=True)),
+                ('next_run_time', models.DateTimeField(blank=True, help_text='下次执行时间', null=True)),
+                ('created_at', models.DateTimeField(auto_now_add=True, help_text='创建时间')),
+                ('updated_at', models.DateTimeField(auto_now=True, help_text='更新时间')),
+                ('organization', models.ForeignKey(blank=True, help_text='扫描组织（设置后执行时动态获取组织下所有目标）', null=True, on_delete=django.db.models.deletion.CASCADE, related_name='scheduled_scans', to='targets.organization')),
+                ('target', models.ForeignKey(blank=True, help_text='扫描单个目标（与 organization 二选一）', null=True, on_delete=django.db.models.deletion.CASCADE, related_name='scheduled_scans', to='targets.target')),
+            ],
+            options={
+                'verbose_name': '定时扫描任务',
+                'verbose_name_plural': '定时扫描任务',
+                'db_table': 'scheduled_scan',
+                'ordering': ['-created_at'],
+                'indexes': [models.Index(fields=['-created_at'], name='scheduled_s_created_9b9c2e_idx'), models.Index(fields=['is_enabled', '-created_at'], name='scheduled_s_is_enab_23d660_idx'), models.Index(fields=['name'], name='scheduled_s_name_bf332d_idx')],
+            },
+        ),
+        migrations.CreateModel(
+            name='ScanLog',
+            fields=[
+                ('id', models.BigAutoField(primary_key=True, serialize=False)),
+                ('level', models.CharField(choices=[('info', 'Info'), ('warning', 'Warning'), ('error', 'Error')], default='info', help_text='日志级别', max_length=10)),
+                ('content', models.TextField(help_text='日志内容')),
+                ('created_at', models.DateTimeField(auto_now_add=True, db_index=True, help_text='创建时间')),
+                ('scan', models.ForeignKey(db_index=True, help_text='关联的扫描任务', on_delete=django.db.models.deletion.CASCADE, related_name='logs', to='scan.scan')),
+            ],
+            options={
+                'verbose_name': '扫描日志',
+                'verbose_name_plural': '扫描日志',
+                'db_table': 'scan_log',
+                'ordering': ['created_at'],
+                'indexes': [models.Index(fields=['scan', 'created_at'], name='scan_log_scan_id_e8c8f5_idx')],
+            },
+        ),
+    ]

backend/apps/scan/models.py

@@ -20,11 +20,19 @@ class Scan(models.Model):
 
     target = models.ForeignKey('targets.Target', on_delete=models.CASCADE, related_name='scans', help_text='扫描目标')
 
-    engine = models.ForeignKey(
-        'engine.ScanEngine',
-        on_delete=models.CASCADE,
-        related_name='scans',
-        help_text='使用的扫描引擎'
+    # 多引擎支持字段
+    engine_ids = ArrayField(
+        models.IntegerField(),
+        default=list,
+        help_text='引擎 ID 列表'
+    )
+    engine_names = models.JSONField(
+        default=list,
+        help_text='引擎名称列表，如 ["引擎A", "引擎B"]'
+    )
+    yaml_configuration = models.TextField(
+        default='',
+        help_text='YAML 格式的扫描配置'
     )
 
     created_at = models.DateTimeField(auto_now_add=True, help_text='任务创建时间')
@@ -98,6 +106,55 @@ class Scan(models.Model):
         return f"Scan #{self.id} - {self.target.name}"
 
 
+class ScanLog(models.Model):
+    """扫描日志模型
+    
+    存储扫描过程中的关键处理日志，用于前端实时查看扫描进度。
+    
+    日志类型：
+    - 阶段开始/完成/失败
+    - 处理进度（如 "Progress: 50/120"）
+    - 发现结果统计（如 "Found 120 subdomains"）
+    - 错误信息
+    
+    日志格式：[stage_name] message
+    """
+    
+    class Level(models.TextChoices):
+        INFO = 'info', 'Info'
+        WARNING = 'warning', 'Warning'
+        ERROR = 'error', 'Error'
+    
+    id = models.BigAutoField(primary_key=True)
+    scan = models.ForeignKey(
+        'Scan',
+        on_delete=models.CASCADE,
+        related_name='logs',
+        db_index=True,
+        help_text='关联的扫描任务'
+    )
+    level = models.CharField(
+        max_length=10,
+        choices=Level.choices,
+        default=Level.INFO,
+        help_text='日志级别'
+    )
+    content = models.TextField(help_text='日志内容')
+    created_at = models.DateTimeField(auto_now_add=True, db_index=True, help_text='创建时间')
+    
+    class Meta:
+        db_table = 'scan_log'
+        verbose_name = '扫描日志'
+        verbose_name_plural = '扫描日志'
+        ordering = ['created_at']
+        indexes = [
+            models.Index(fields=['scan', 'created_at']),
+        ]
+    
+    def __str__(self):
+        return f"[{self.level}] {self.content[:50]}"
+
+
 class ScheduledScan(models.Model):
     """
     定时扫描任务模型
@@ -118,12 +175,19 @@ class ScheduledScan(models.Model):
     # 基本信息
     name = models.CharField(max_length=200, help_text='任务名称')
     
-    # 关联的扫描引擎
-    engine = models.ForeignKey(
-        'engine.ScanEngine',
-        on_delete=models.CASCADE,
-        related_name='scheduled_scans',
-        help_text='使用的扫描引擎'
+    # 多引擎支持字段
+    engine_ids = ArrayField(
+        models.IntegerField(),
+        default=list,
+        help_text='引擎 ID 列表'
+    )
+    engine_names = models.JSONField(
+        default=list,
+        help_text='引擎名称列表，如 ["引擎A", "引擎B"]'
+    )
+    yaml_configuration = models.TextField(
+        default='',
+        help_text='YAML 格式的扫描配置'
     )
     
     # 关联的组织（组织扫描模式：执行时动态获取组织下所有目标）

backend/apps/scan/notifications/consumers.py

@@ -5,12 +5,13 @@ WebSocket Consumer - 通知实时推送
 import json
 import logging
 import asyncio
-from channels.generic.websocket import AsyncWebsocketConsumer
+
+from apps.common.websocket_auth import AuthenticatedWebsocketConsumer
 
 logger = logging.getLogger(__name__)
 
 
-class NotificationConsumer(AsyncWebsocketConsumer):
+class NotificationConsumer(AuthenticatedWebsocketConsumer):
     """
     通知 WebSocket Consumer
     
@@ -23,9 +24,9 @@ class NotificationConsumer(AsyncWebsocketConsumer):
         super().__init__(*args, **kwargs)
         self.heartbeat_task = None  # 心跳任务
     
-    async def connect(self):
+    async def on_connect(self):
         """
-        客户端连接时调用
+        客户端连接时调用（已通过认证）
         加入通知广播组
         """
         # 通知组名（所有客户端共享）

backend/apps/scan/notifications/services.py

@@ -305,6 +305,7 @@ def _push_via_api_callback(notification: Notification, server_url: str) -> None:
     通过 HTTP 请求 Server 容器的 /api/callbacks/notification/ 接口。
     Worker 无法直接访问 Redis，需要由 Server 代为推送 WebSocket。
     """
+    import os
     import requests
     
     try:
@@ -318,8 +319,14 @@ def _push_via_api_callback(notification: Notification, server_url: str) -> None:
             'created_at': notification.created_at.isoformat()
         }
         
+        # 构建请求头（包含 Worker API Key）
+        headers = {'Content-Type': 'application/json'}
+        worker_api_key = os.environ.get("WORKER_API_KEY", "")
+        if worker_api_key:
+            headers["X-Worker-API-Key"] = worker_api_key
+        
         # verify=False: 远程 Worker 回调 Server 时可能使用自签名证书
-        resp = requests.post(callback_url, json=data, timeout=5, verify=False)
+        resp = requests.post(callback_url, json=data, headers=headers, timeout=5, verify=False)
         resp.raise_for_status()
         
         logger.debug(f"通知回调推送成功 - ID: {notification.id}")

backend/apps/scan/notifications/views.py

@@ -7,8 +7,7 @@ from typing import Any
 from django.http import JsonResponse
 from django.utils import timezone
 from rest_framework import status
-from rest_framework.decorators import api_view, permission_classes
-from rest_framework.permissions import AllowAny
+from rest_framework.decorators import api_view
 from rest_framework.request import Request
 from rest_framework.response import Response
 from rest_framework.views import APIView
@@ -198,12 +197,13 @@ class NotificationSettingsView(APIView):
 # ============================================
 
 @api_view(['POST'])
-@permission_classes([AllowAny])  # Worker 容器无认证，可考虑添加 Token 验证
+# 权限由全局 IsAuthenticatedOrPublic 处理，/api/callbacks/* 需要 Worker API Key 认证
 def notification_callback(request):
     """
     接收 Worker 的通知推送请求
     
     Worker 容器无法直接访问 Redis，通过此 API 回调让 Server 推送 WebSocket。
+    需要 Worker API Key 认证（X-Worker-API-Key Header）。
     
     POST /api/callbacks/notification/
     {

backend/apps/scan/repositories/django_scan_repository.py

@@ -16,7 +16,6 @@ from django.utils import timezone
 
 from apps.scan.models import Scan
 from apps.targets.models import Target
-from apps.engine.models import ScanEngine
 from apps.common.definitions import ScanStatus
 from apps.common.decorators import auto_ensure_db_connection
 
@@ -40,7 +39,7 @@ class DjangoScanRepository:
         
         Args:
             scan_id: 扫描任务 ID
-            prefetch_relations: 是否预加载关联对象（engine, target）
+            prefetch_relations: 是否预加载关联对象（target, worker）
                               默认 False，只在需要展示关联信息时设为 True
             for_update: 是否加锁（用于更新场景）
         
@@ -56,7 +55,7 @@ class DjangoScanRepository:
             
             # 预加载关联对象（性能优化：默认不加载）
             if prefetch_relations:
-                queryset = queryset.select_related('engine', 'target')
+                queryset = queryset.select_related('target', 'worker')
             
             return queryset.get(id=scan_id)
         except Scan.DoesNotExist:  # type: ignore  # pylint: disable=no-member
@@ -79,7 +78,7 @@ class DjangoScanRepository:
         
         Note:
             - 使用默认的阻塞模式（等待锁释放）
-            - 不包含关联对象（engine, target），如需关联对象请使用 get_by_id()
+            - 不包含关联对象（target, worker），如需关联对象请使用 get_by_id()
         """
         try:
             return Scan.objects.select_for_update().get(id=scan_id)  # type: ignore  # pylint: disable=no-member
@@ -103,7 +102,9 @@ class DjangoScanRepository:
     
     def create(self,
         target: Target,
-        engine: ScanEngine,
+        engine_ids: List[int],
+        engine_names: List[str],
+        yaml_configuration: str,
         results_dir: str,
         status: ScanStatus = ScanStatus.INITIATED
     ) -> Scan:
@@ -112,7 +113,9 @@ class DjangoScanRepository:
         
         Args:
             target: 扫描目标
-            engine: 扫描引擎
+            engine_ids: 引擎 ID 列表
+            engine_names: 引擎名称列表
+            yaml_configuration: YAML 格式的扫描配置
             results_dir: 结果目录
             status: 初始状态
         
@@ -121,7 +124,9 @@ class DjangoScanRepository:
         """
         scan = Scan(
             target=target,
-            engine=engine,
+            engine_ids=engine_ids,
+            engine_names=engine_names,
+            yaml_configuration=yaml_configuration,
             results_dir=results_dir,
             status=status,
             container_ids=[]
@@ -231,14 +236,14 @@ class DjangoScanRepository:
         获取所有扫描任务
         
         Args:
-            prefetch_relations: 是否预加载关联对象（engine, target）
+            prefetch_relations: 是否预加载关联对象（target, worker）
         
         Returns:
             Scan QuerySet
         """
         queryset = Scan.objects.all()  # type: ignore  # pylint: disable=no-member
         if prefetch_relations:
-            queryset = queryset.select_related('engine', 'target')
+            queryset = queryset.select_related('target', 'worker')
         return queryset.order_by('-created_at')
     
     

backend/apps/scan/repositories/scheduled_scan_repository.py

@@ -29,7 +29,9 @@ class ScheduledScanDTO:
     """
     id: Optional[int] = None
     name: str = ''
-    engine_id: int = 0
+    engine_ids: List[int] = None  # 多引擎支持
+    engine_names: List[str] = None  # 引擎名称列表
+    yaml_configuration: str = ''  # YAML 格式的扫描配置
     organization_id: Optional[int] = None  # 组织扫描模式
     target_id: Optional[int] = None  # 目标扫描模式
     cron_expression: Optional[str] = None
@@ -40,6 +42,11 @@ class ScheduledScanDTO:
     created_at: Optional[datetime] = None
     updated_at: Optional[datetime] = None
     
+    def __post_init__(self):
+        if self.engine_ids is None:
+            self.engine_ids = []
+        if self.engine_names is None:
+            self.engine_names = []
 
 
 @auto_ensure_db_connection
@@ -56,7 +63,7 @@ class DjangoScheduledScanRepository:
     def get_by_id(self, scheduled_scan_id: int) -> Optional[ScheduledScan]:
         """根据 ID 查询定时扫描任务"""
         try:
-            return ScheduledScan.objects.select_related('engine', 'organization', 'target').get(id=scheduled_scan_id)
+            return ScheduledScan.objects.select_related('organization', 'target').get(id=scheduled_scan_id)
         except ScheduledScan.DoesNotExist:
             return None
     
@@ -67,7 +74,7 @@ class DjangoScheduledScanRepository:
         Returns:
             QuerySet
         """
-        return ScheduledScan.objects.select_related('engine', 'organization', 'target').order_by('-created_at')
+        return ScheduledScan.objects.select_related('organization', 'target').order_by('-created_at')
 
     def get_all(self, page: int = 1, page_size: int = 10) -> Tuple[List[ScheduledScan], int]:
         """
@@ -87,7 +94,7 @@ class DjangoScheduledScanRepository:
     def get_enabled(self) -> List[ScheduledScan]:
         """获取所有启用的定时扫描任务"""
         return list(
-            ScheduledScan.objects.select_related('engine', 'target')
+            ScheduledScan.objects.select_related('target')
             .filter(is_enabled=True)
             .order_by('-created_at')
         )
@@ -105,7 +112,9 @@ class DjangoScheduledScanRepository:
         with transaction.atomic():
             scheduled_scan = ScheduledScan.objects.create(
                 name=dto.name,
-                engine_id=dto.engine_id,
+                engine_ids=dto.engine_ids,
+                engine_names=dto.engine_names,
+                yaml_configuration=dto.yaml_configuration,
                 organization_id=dto.organization_id,  # 组织扫描模式
                 target_id=dto.target_id if not dto.organization_id else None,  # 目标扫描模式
                 cron_expression=dto.cron_expression,
@@ -134,8 +143,12 @@ class DjangoScheduledScanRepository:
                 # 更新基本字段
                 if dto.name:
                     scheduled_scan.name = dto.name
-                if dto.engine_id:
-                    scheduled_scan.engine_id = dto.engine_id
+                if dto.engine_ids is not None:
+                    scheduled_scan.engine_ids = dto.engine_ids
+                if dto.engine_names is not None:
+                    scheduled_scan.engine_names = dto.engine_names
+                if dto.yaml_configuration is not None:
+                    scheduled_scan.yaml_configuration = dto.yaml_configuration
                 if dto.cron_expression is not None:
                     scheduled_scan.cron_expression = dto.cron_expression
                 if dto.is_enabled is not None:

backend/apps/scan/serializers.py

@@ -1,18 +1,89 @@
 from rest_framework import serializers
 from django.db.models import Count
+import yaml
 
-from .models import Scan, ScheduledScan
+from .models import Scan, ScheduledScan, ScanLog
+
+
+# ==================== 扫描日志序列化器 ====================
+
+class ScanLogSerializer(serializers.ModelSerializer):
+    """扫描日志序列化器"""
+    
+    class Meta:
+        model = ScanLog
+        fields = ['id', 'level', 'content', 'created_at']
+
+
+# ==================== 通用验证 Mixin ====================
+
+class DuplicateKeyLoader(yaml.SafeLoader):
+    """自定义 YAML Loader，检测重复 key"""
+    pass
+
+
+def _check_duplicate_keys(loader, node, deep=False):
+    """检测 YAML mapping 中的重复 key"""
+    mapping = {}
+    for key_node, value_node in node.value:
+        key = loader.construct_object(key_node, deep=deep)
+        if key in mapping:
+            raise yaml.constructor.ConstructorError(
+                "while constructing a mapping", node.start_mark,
+                f"发现重复的配置项 '{key}'，后面的配置会覆盖前面的配置，请删除重复项", key_node.start_mark
+            )
+        mapping[key] = loader.construct_object(value_node, deep=deep)
+    return mapping
+
+
+DuplicateKeyLoader.add_constructor(
+    yaml.resolver.BaseResolver.DEFAULT_MAPPING_TAG,
+    _check_duplicate_keys
+)
+
+
+class ScanConfigValidationMixin:
+    """扫描配置验证 Mixin，提供通用的验证方法"""
+    
+    def validate_configuration(self, value):
+        """验证 YAML 配置格式，包括检测重复 key"""
+        import yaml
+        
+        if not value or not value.strip():
+            raise serializers.ValidationError("configuration 不能为空")
+        
+        try:
+            # 使用自定义 Loader 检测重复 key
+            yaml.load(value, Loader=DuplicateKeyLoader)
+        except yaml.YAMLError as e:
+            raise serializers.ValidationError(f"无效的 YAML 格式: {str(e)}")
+        
+        return value
+    
+    def validate_engine_ids(self, value):
+        """验证引擎 ID 列表"""
+        if not value:
+            raise serializers.ValidationError("engine_ids 不能为空，请至少选择一个扫描引擎")
+        return value
+    
+    def validate_engine_names(self, value):
+        """验证引擎名称列表"""
+        if not value:
+            raise serializers.ValidationError("engine_names 不能为空")
+        return value
+
+
+# ==================== 扫描任务序列化器 ====================
 
 
 class ScanSerializer(serializers.ModelSerializer):
     """扫描任务序列化器"""
     target_name = serializers.SerializerMethodField()
-    engine_name = serializers.SerializerMethodField()
     
     class Meta:
         model = Scan
         fields = [
-            'id', 'target', 'target_name', 'engine', 'engine_name',
+            'id', 'target', 'target_name', 'engine_ids', 'engine_names',
             'created_at', 'stopped_at', 'status', 'results_dir',
             'container_ids', 'error_message'
         ]
@@ -24,10 +95,6 @@ class ScanSerializer(serializers.ModelSerializer):
     def get_target_name(self, obj):
         """获取目标名称"""
         return obj.target.name if obj.target else None
-    
-    def get_engine_name(self, obj):
-        """获取引擎名称"""
-        return obj.engine.name if obj.engine else None
 
 
 class ScanHistorySerializer(serializers.ModelSerializer):
@@ -36,11 +103,12 @@ class ScanHistorySerializer(serializers.ModelSerializer):
     为前端扫描历史页面提供优化的数据格式，包括：
     - 扫描汇总统计（子域名、端点、漏洞数量）
     - 进度百分比和当前阶段
+    - 执行节点信息
     """
     
     # 字段映射
     target_name = serializers.CharField(source='target.name', read_only=True)
-    engine_name = serializers.CharField(source='engine.name', read_only=True)
+    worker_name = serializers.CharField(source='worker.name', read_only=True, allow_null=True)
     
     # 计算字段
     summary = serializers.SerializerMethodField()
@@ -53,9 +121,9 @@ class ScanHistorySerializer(serializers.ModelSerializer):
     class Meta:
         model = Scan
         fields = [
-            'id', 'target', 'target_name', 'engine', 'engine_name', 
-            'created_at', 'status', 'error_message', 'summary', 'progress',
-            'current_stage', 'stage_progress'
+            'id', 'target', 'target_name', 'engine_ids', 'engine_names', 
+            'worker_name', 'created_at', 'status', 'error_message', 'summary', 
+            'progress', 'current_stage', 'stage_progress'
         ]
     
     def get_summary(self, obj):
@@ -86,12 +154,12 @@ class ScanHistorySerializer(serializers.ModelSerializer):
         return summary
 
 
-class QuickScanSerializer(serializers.Serializer):
+class QuickScanSerializer(ScanConfigValidationMixin, serializers.Serializer):
     """
     快速扫描序列化器
     
     功能：
-    - 接收目标列表和引擎配置
+    - 接收目标列表和 YAML 配置
     - 自动创建/获取目标
     - 立即发起扫描
     """
@@ -105,10 +173,24 @@ class QuickScanSerializer(serializers.Serializer):
         help_text='目标列表，每个目标包含 name 字段'
     )
     
-    # 扫描引擎 ID
-    engine_id = serializers.IntegerField(
+    # YAML 配置（必填）
+    configuration = serializers.CharField(
         required=True,
-        help_text='使用的扫描引擎 ID (必填)'
+        help_text='YAML 格式的扫描配置（必填）'
+    )
+    
+    # 扫描引擎 ID 列表（必填，用于记录和显示）
+    engine_ids = serializers.ListField(
+        child=serializers.IntegerField(),
+        required=True,
+        help_text='使用的扫描引擎 ID 列表（必填）'
+    )
+    
+    # 引擎名称列表（必填，用于记录和显示）
+    engine_names = serializers.ListField(
+        child=serializers.CharField(),
+        required=True,
+        help_text='引擎名称列表（必填）'
     )
     
     def validate_targets(self, value):
@@ -138,7 +220,6 @@ class ScheduledScanSerializer(serializers.ModelSerializer):
     """定时扫描任务序列化器（用于列表和详情）"""
     
     # 关联字段
-    engine_name = serializers.CharField(source='engine.name', read_only=True)
     organization_id = serializers.IntegerField(source='organization.id', read_only=True, allow_null=True)
     organization_name = serializers.CharField(source='organization.name', read_only=True, allow_null=True)
     target_id = serializers.IntegerField(source='target.id', read_only=True, allow_null=True)
@@ -149,7 +230,7 @@ class ScheduledScanSerializer(serializers.ModelSerializer):
         model = ScheduledScan
         fields = [
             'id', 'name',
-            'engine', 'engine_name',
+            'engine_ids', 'engine_names',
             'organization_id', 'organization_name',
             'target_id', 'target_name',
             'scan_mode',
@@ -169,7 +250,7 @@ class ScheduledScanSerializer(serializers.ModelSerializer):
         return 'organization' if obj.organization_id else 'target'
 
 
-class CreateScheduledScanSerializer(serializers.Serializer):
+class CreateScheduledScanSerializer(ScanConfigValidationMixin, serializers.Serializer):
     """创建定时扫描任务序列化器
     
     扫描模式（二选一）：
@@ -178,7 +259,26 @@ class CreateScheduledScanSerializer(serializers.Serializer):
     """
     
     name = serializers.CharField(max_length=200, help_text='任务名称')
-    engine_id = serializers.IntegerField(help_text='扫描引擎 ID')
+    
+    # YAML 配置（必填）
+    configuration = serializers.CharField(
+        required=True,
+        help_text='YAML 格式的扫描配置（必填）'
+    )
+    
+    # 扫描引擎 ID 列表（必填，用于记录和显示）
+    engine_ids = serializers.ListField(
+        child=serializers.IntegerField(),
+        required=True,
+        help_text='扫描引擎 ID 列表（必填）'
+    )
+    
+    # 引擎名称列表（必填，用于记录和显示）
+    engine_names = serializers.ListField(
+        child=serializers.CharField(),
+        required=True,
+        help_text='引擎名称列表（必填）'
+    )
     
     # 组织扫描模式
     organization_id = serializers.IntegerField(
@@ -215,11 +315,71 @@ class CreateScheduledScanSerializer(serializers.Serializer):
         return data
 
 
+class InitiateScanSerializer(ScanConfigValidationMixin, serializers.Serializer):
+    """发起扫描任务序列化器
+    
+    扫描模式（二选一）：
+    - 组织扫描：提供 organization_id，扫描组织下所有目标
+    - 目标扫描：提供 target_id，扫描单个目标
+    """
+    
+    # YAML 配置（必填）
+    configuration = serializers.CharField(
+        required=True,
+        help_text='YAML 格式的扫描配置（必填）'
+    )
+    
+    # 扫描引擎 ID 列表（必填）
+    engine_ids = serializers.ListField(
+        child=serializers.IntegerField(),
+        required=True,
+        help_text='扫描引擎 ID 列表（必填）'
+    )
+    
+    # 引擎名称列表（必填）
+    engine_names = serializers.ListField(
+        child=serializers.CharField(),
+        required=True,
+        help_text='引擎名称列表（必填）'
+    )
+    
+    # 组织扫描模式
+    organization_id = serializers.IntegerField(
+        required=False,
+        allow_null=True,
+        help_text='组织 ID（组织扫描模式）'
+    )
+    
+    # 目标扫描模式
+    target_id = serializers.IntegerField(
+        required=False,
+        allow_null=True,
+        help_text='目标 ID（目标扫描模式）'
+    )
+    
+    def validate(self, data):
+        """验证 organization_id 和 target_id 互斥"""
+        organization_id = data.get('organization_id')
+        target_id = data.get('target_id')
+        
+        if not organization_id and not target_id:
+            raise serializers.ValidationError('必须提供 organization_id 或 target_id 其中之一')
+        
+        if organization_id and target_id:
+            raise serializers.ValidationError('organization_id 和 target_id 只能提供其中之一')
+        
+        return data
+
+
 class UpdateScheduledScanSerializer(serializers.Serializer):
     """更新定时扫描任务序列化器"""
     
     name = serializers.CharField(max_length=200, required=False, help_text='任务名称')
-    engine_id = serializers.IntegerField(required=False, help_text='扫描引擎 ID')
+    engine_ids = serializers.ListField(
+        child=serializers.IntegerField(),
+        required=False,
+        help_text='扫描引擎 ID 列表'
+    )
     
     # 组织扫描模式
     organization_id = serializers.IntegerField(
@@ -237,6 +397,12 @@ class UpdateScheduledScanSerializer(serializers.Serializer):
     
     cron_expression = serializers.CharField(max_length=100, required=False, help_text='Cron 表达式')
     is_enabled = serializers.BooleanField(required=False, help_text='是否启用')
+    
+    def validate_engine_ids(self, value):
+        """验证引擎 ID 列表"""
+        if value is not None and not value:
+            raise serializers.ValidationError("engine_ids 不能为空")
+        return value
 
 
 class ToggleScheduledScanSerializer(serializers.Serializer):

backend/apps/scan/services/scan_creation_service.py

@@ -10,7 +10,7 @@
 import uuid
 import logging
 import threading
-from typing import List
+from typing import List, Tuple
 from datetime import datetime
 from pathlib import Path
 from django.conf import settings
@@ -20,6 +20,7 @@ from django.core.exceptions import ValidationError, ObjectDoesNotExist
 
 from apps.scan.models import Scan
 from apps.scan.repositories import DjangoScanRepository
+from apps.scan.utils.config_merger import merge_engine_configs, ConfigConflictError
 from apps.targets.repositories import DjangoTargetRepository, DjangoOrganizationRepository
 from apps.engine.repositories import DjangoEngineRepository
 from apps.targets.models import Target
@@ -142,6 +143,106 @@ class ScanCreationService:
         
         return targets, engine
     
+    def prepare_initiate_scan_multi_engine(
+        self,
+        organization_id: int | None = None,
+        target_id: int | None = None,
+        engine_ids: List[int] | None = None
+    ) -> Tuple[List[Target], str, List[str], List[int]]:
+        """
+        准备多引擎扫描任务所需的数据
+        
+        职责：
+            1. 参数验证（必填项、互斥参数）
+            2. 资源查询（Engines、Organization、Target）
+            3. 合并引擎配置（检测冲突）
+            4. 返回准备好的目标列表、合并配置和引擎信息
+        
+        Args:
+            organization_id: 组织ID（可选）
+            target_id: 目标ID（可选）
+            engine_ids: 扫描引擎ID列表（必填）
+        
+        Returns:
+            (目标列表, 合并配置, 引擎名称列表, 引擎ID列表) - 供 create_scans 方法使用
+        
+        Raises:
+            ValidationError: 参数验证失败或业务规则不满足
+            ObjectDoesNotExist: 资源不存在（Organization/Target/ScanEngine）
+            ConfigConflictError: 引擎配置存在冲突
+        
+        Note:
+            - organization_id 和 target_id 必须二选一
+            - 如果提供 organization_id，返回该组织下所有目标
+            - 如果提供 target_id，返回单个目标列表
+        """
+        # 1. 参数验证
+        if not engine_ids:
+            raise ValidationError('缺少必填参数: engine_ids')
+        
+        if not organization_id and not target_id:
+            raise ValidationError('必须提供 organization_id 或 target_id 其中之一')
+        
+        if organization_id and target_id:
+            raise ValidationError('organization_id 和 target_id 只能提供其中之一')
+        
+        # 2. 查询所有扫描引擎
+        engines = []
+        for engine_id in engine_ids:
+            engine = self.engine_repo.get_by_id(engine_id)
+            if not engine:
+                logger.error("扫描引擎不存在 - Engine ID: %s", engine_id)
+                raise ObjectDoesNotExist(f'ScanEngine ID {engine_id} 不存在')
+            engines.append(engine)
+        
+        # 3. 合并引擎配置（可能抛出 ConfigConflictError）
+        engine_configs = [(e.name, e.configuration or '') for e in engines]
+        merged_configuration = merge_engine_configs(engine_configs)
+        engine_names = [e.name for e in engines]
+        
+        logger.debug(
+            "引擎配置合并成功 - 引擎: %s",
+            ', '.join(engine_names)
+        )
+        
+        # 4. 根据参数获取目标列表
+        targets = []
+        
+        if organization_id:
+            # 根据组织ID获取所有目标
+            organization = self.organization_repo.get_by_id(organization_id)
+            if not organization:
+                logger.error("组织不存在 - Organization ID: %s", organization_id)
+                raise ObjectDoesNotExist(f'Organization ID {organization_id} 不存在')
+            
+            targets = self.organization_repo.get_targets(organization_id)
+            
+            if not targets:
+                raise ValidationError(f'组织 ID {organization_id} 下没有目标')
+            
+            logger.debug(
+                "准备发起扫描 - 组织: %s, 目标数量: %d, 引擎: %s",
+                organization.name,
+                len(targets),
+                ', '.join(engine_names)
+            )
+        else:
+            # 根据目标ID获取单个目标
+            target = self.target_repo.get_by_id(target_id)
+            if not target:
+                logger.error("目标不存在 - Target ID: %s", target_id)
+                raise ObjectDoesNotExist(f'Target ID {target_id} 不存在')
+            
+            targets = [target]
+            
+            logger.debug(
+                "准备发起扫描 - 目标: %s, 引擎: %s",
+                target.name,
+                ', '.join(engine_names)
+            )
+        
+        return targets, merged_configuration, engine_names, engine_ids
+    
     def _generate_scan_workspace_dir(self) -> str:
         """
         生成 Scan 工作空间目录路径
@@ -179,7 +280,9 @@ class ScanCreationService:
     def create_scans(
         self,
         targets: List[Target],
-        engine: ScanEngine,
+        engine_ids: List[int],
+        engine_names: List[str],
+        yaml_configuration: str,
         scheduled_scan_name: str | None = None
     ) -> List[Scan]:
         """
@@ -187,7 +290,9 @@ class ScanCreationService:
         
         Args:
             targets: 目标列表
-            engine: 扫描引擎对象
+            engine_ids: 引擎 ID 列表
+            engine_names: 引擎名称列表
+            yaml_configuration: YAML 格式的扫描配置
             scheduled_scan_name: 定时扫描任务名称（可选，用于通知显示）
         
         Returns:
@@ -205,7 +310,9 @@ class ScanCreationService:
                 scan_workspace_dir = self._generate_scan_workspace_dir()
                 scan = Scan(
                     target=target,
-                    engine=engine,
+                    engine_ids=engine_ids,
+                    engine_names=engine_names,
+                    yaml_configuration=yaml_configuration,
                     results_dir=scan_workspace_dir,
                     status=ScanStatus.INITIATED,
                     container_ids=[],
@@ -236,13 +343,15 @@ class ScanCreationService:
             return []
         
         # 第三步：分发任务到 Workers
+        # 使用第一个引擎名称作为显示名称，或者合并显示
+        display_engine_name = ', '.join(engine_names) if engine_names else ''
         scan_data = [
             {
                 'scan_id': scan.id,
                 'target_name': scan.target.name,
                 'target_id': scan.target.id,
                 'results_dir': scan.results_dir,
-                'engine_name': scan.engine.name,
+                'engine_name': display_engine_name,
                 'scheduled_scan_name': scheduled_scan_name,
             }
             for scan in created_scans

backend/apps/scan/services/scan_service.py

@@ -96,14 +96,34 @@ class ScanService:
             organization_id, target_id, engine_id
         )
     
+    def prepare_initiate_scan_multi_engine(
+        self,
+        organization_id: int | None = None,
+        target_id: int | None = None,
+        engine_ids: List[int] | None = None
+    ) -> tuple[List[Target], str, List[str], List[int]]:
+        """
+        为创建多引擎扫描任务做准备
+        
+        Returns:
+            (目标列表, 合并配置, 引擎名称列表, 引擎ID列表)
+        """
+        return self.creation_service.prepare_initiate_scan_multi_engine(
+            organization_id, target_id, engine_ids
+        )
+    
     def create_scans(
         self,
         targets: List[Target],
-        engine: ScanEngine,
+        engine_ids: List[int],
+        engine_names: List[str],
+        yaml_configuration: str,
         scheduled_scan_name: str | None = None
     ) -> List[Scan]:
         """批量创建扫描任务（委托给 ScanCreationService）"""
-        return self.creation_service.create_scans(targets, engine, scheduled_scan_name)
+        return self.creation_service.create_scans(
+            targets, engine_ids, engine_names, yaml_configuration, scheduled_scan_name
+        )
     
     # ==================== 状态管理方法（委托给 ScanStateService） ====================
     

backend/apps/scan/services/scheduled_scan_service.py

@@ -14,6 +14,7 @@ from django.core.exceptions import ValidationError
 
 from apps.scan.models import ScheduledScan
 from apps.scan.repositories import DjangoScheduledScanRepository, ScheduledScanDTO
+from apps.scan.utils.config_merger import merge_engine_configs, ConfigConflictError
 from apps.engine.repositories import DjangoEngineRepository
 from apps.targets.services import TargetService
 
@@ -53,12 +54,13 @@ class ScheduledScanService:
     
     def create(self, dto: ScheduledScanDTO) -> ScheduledScan:
         """
-        创建定时扫描任务
+        创建定时扫描任务（使用引擎 ID 合并配置）
         
         流程：
         1. 验证参数
-        2. 创建数据库记录
-        3. 计算并设置 next_run_time
+        2. 合并引擎配置
+        3. 创建数据库记录
+        4. 计算并设置 next_run_time
         
         Args:
             dto: 定时扫描 DTO
@@ -68,11 +70,66 @@ class ScheduledScanService:
         
         Raises:
             ValidationError: 参数验证失败
+            ConfigConflictError: 引擎配置冲突
         """
         # 1. 验证参数
         self._validate_create_dto(dto)
         
-        # 2. 创建数据库记录
+        # 2. 合并引擎配置
+        engines = []
+        engine_names = []
+        for engine_id in dto.engine_ids:
+            engine = self.engine_repo.get_by_id(engine_id)
+            if engine:
+                engines.append((engine.name, engine.configuration or ''))
+                engine_names.append(engine.name)
+        
+        merged_configuration = merge_engine_configs(engines)
+        
+        # 设置 DTO 的合并配置和引擎名称
+        dto.engine_names = engine_names
+        dto.yaml_configuration = merged_configuration
+        
+        # 3. 创建数据库记录
+        scheduled_scan = self.repo.create(dto)
+        
+        # 4. 如果有 cron 表达式且已启用，计算下次执行时间
+        if scheduled_scan.cron_expression and scheduled_scan.is_enabled:
+            next_run_time = self._calculate_next_run_time(scheduled_scan)
+            if next_run_time:
+                self.repo.update_next_run_time(scheduled_scan.id, next_run_time)
+                scheduled_scan.next_run_time = next_run_time
+        
+        logger.info(
+            "创建定时扫描任务 - ID: %s, 名称: %s, 下次执行: %s",
+            scheduled_scan.id, scheduled_scan.name, scheduled_scan.next_run_time
+        )
+        
+        return scheduled_scan
+    
+    def create_with_configuration(self, dto: ScheduledScanDTO) -> ScheduledScan:
+        """
+        创建定时扫描任务（直接使用前端传递的配置）
+        
+        流程：
+        1. 验证参数
+        2. 直接使用 dto.yaml_configuration
+        3. 创建数据库记录
+        4. 计算并设置 next_run_time
+        
+        Args:
+            dto: 定时扫描 DTO（必须包含 yaml_configuration）
+        
+        Returns:
+            创建的 ScheduledScan 对象
+        
+        Raises:
+            ValidationError: 参数验证失败
+        """
+        # 1. 验证参数
+        self._validate_create_dto_with_configuration(dto)
+        
+        # 2. 创建数据库记录（直接使用 dto 中的配置）
         scheduled_scan = self.repo.create(dto)
         
         # 3. 如果有 cron 表达式且已启用，计算下次执行时间
@@ -90,18 +147,33 @@ class ScheduledScanService:
         return scheduled_scan
     
     def _validate_create_dto(self, dto: ScheduledScanDTO) -> None:
-        """验证创建 DTO"""
+        """验证创建 DTO（使用引擎 ID）"""
+        # 基础验证
+        self._validate_base_dto(dto)
+        
+        if not dto.engine_ids:
+            raise ValidationError('必须选择扫描引擎')
+        
+        # 验证所有引擎是否存在
+        for engine_id in dto.engine_ids:
+            if not self.engine_repo.get_by_id(engine_id):
+                raise ValidationError(f'扫描引擎 ID {engine_id} 不存在')
+    
+    def _validate_create_dto_with_configuration(self, dto: ScheduledScanDTO) -> None:
+        """验证创建 DTO（使用前端传递的配置）"""
+        # 基础验证
+        self._validate_base_dto(dto)
+        
+        if not dto.yaml_configuration:
+            raise ValidationError('配置不能为空')
+    
+    def _validate_base_dto(self, dto: ScheduledScanDTO) -> None:
+        """验证 DTO 的基础字段（公共逻辑）"""
         from apps.targets.repositories import DjangoOrganizationRepository
         
         if not dto.name:
             raise ValidationError('任务名称不能为空')
         
-        if not dto.engine_id:
-            raise ValidationError('必须选择扫描引擎')
-        
-        if not self.engine_repo.get_by_id(dto.engine_id):
-            raise ValidationError(f'扫描引擎 ID {dto.engine_id} 不存在')
-        
         # 验证扫描模式（organization_id 和 target_id 互斥）
         if not dto.organization_id and not dto.target_id:
             raise ValidationError('必须选择组织或扫描目标')
@@ -138,11 +210,28 @@ class ScheduledScanService:
         
         Returns:
             更新后的 ScheduledScan 对象
+        
+        Raises:
+            ConfigConflictError: 引擎配置冲突
         """
         existing = self.repo.get_by_id(scheduled_scan_id)
         if not existing:
             return None
         
+        # 如果引擎变更，重新合并配置
+        if dto.engine_ids is not None:
+            engines = []
+            engine_names = []
+            for engine_id in dto.engine_ids:
+                engine = self.engine_repo.get_by_id(engine_id)
+                if engine:
+                    engines.append((engine.name, engine.configuration or ''))
+                    engine_names.append(engine.name)
+            
+            merged_configuration = merge_engine_configs(engines)
+            dto.engine_names = engine_names
+            dto.yaml_configuration = merged_configuration
+        
         # 更新数据库记录
         scheduled_scan = self.repo.update(scheduled_scan_id, dto)
         if not scheduled_scan:
@@ -292,21 +381,25 @@ class ScheduledScanService:
         立即触发扫描（支持组织扫描和目标扫描两种模式）
         
         复用 ScanService 的逻辑，与 API 调用保持一致。
+        使用存储的 yaml_configuration 而不是重新合并。
         """
         from apps.scan.services.scan_service import ScanService
         
         scan_service = ScanService()
         
-        # 1. 准备扫描所需数据（复用 API 的逻辑）
-        targets, engine = scan_service.prepare_initiate_scan(
+        # 1. 准备扫描所需数据（使用存储的多引擎配置）
+        targets, _, _, _ = scan_service.prepare_initiate_scan_multi_engine(
             organization_id=scheduled_scan.organization_id,
             target_id=scheduled_scan.target_id,
-            engine_id=scheduled_scan.engine_id
+            engine_ids=scheduled_scan.engine_ids
         )
         
-        # 2. 创建扫描任务，传递定时扫描名称用于通知显示
+        # 2. 创建扫描任务，使用存储的合并配置
         created_scans = scan_service.create_scans(
-            targets, engine,
+            targets=targets,
+            engine_ids=scheduled_scan.engine_ids,
+            engine_names=scheduled_scan.engine_names,
+            yaml_configuration=scheduled_scan.yaml_configuration,
             scheduled_scan_name=scheduled_scan.name
         )
         

backend/apps/scan/tasks/fingerprint_detect/run_xingfinger_task.py

@@ -4,7 +4,6 @@ xingfinger 执行任务
 流式执行 xingfinger 命令并实时更新 tech 字段
 """
 
-import importlib
 import json
 import logging
 import subprocess
@@ -15,93 +14,97 @@ from django.db import connection
 from prefect import task
 
 from apps.scan.utils import execute_stream
+from apps.asset.dtos.snapshot import WebsiteSnapshotDTO
+from apps.asset.repositories.snapshot import DjangoWebsiteSnapshotRepository
 
 logger = logging.getLogger(__name__)
 
 
-# 数据源映射：source → (module_path, model_name, url_field)
-SOURCE_MODEL_MAP = {
-    'website': ('apps.asset.models', 'WebSite', 'url'),
-    # 以后扩展：
-    # 'endpoint': ('apps.asset.models', 'Endpoint', 'url'),
-    # 'directory': ('apps.asset.models', 'Directory', 'url'),
-}
-
-
-def _get_model_class(source: str):
-    """根据数据源类型获取 Model 类"""
-    if source not in SOURCE_MODEL_MAP:
-        raise ValueError(f"不支持的数据源: {source}")
-    
-    module_path, model_name, _ = SOURCE_MODEL_MAP[source]
-    module = importlib.import_module(module_path)
-    return getattr(module, model_name)
-
-
-def parse_xingfinger_line(line: str) -> tuple[str, list[str]] | None:
+def parse_xingfinger_line(line: str) -> dict | None:
     """
     解析 xingfinger 单行 JSON 输出
     
-    xingfinger 静默模式输出格式：
-    {"url": "https://example.com", "cms": "WordPress,PHP,nginx", ...}
+    xingfinger 输出格式：
+    {"url": "...", "cms": "...", "server": "BWS/1.1", "status_code": 200, "length": 642831, "title": "..."}
     
     Returns:
-        tuple: (url, tech_list) 或 None（解析失败时）
+        dict: 包含 url, techs, server, title, status_code, content_length 的字典
+        None: 解析失败或 URL 为空时
     """
     try:
         item = json.loads(line)
         url = item.get('url', '').strip()
-        cms = item.get('cms', '')
         
-        if not url or not cms:
+        if not url:
             return None
         
         # cms 字段按逗号分割，去除空白
-        techs = [t.strip() for t in cms.split(',') if t.strip()]
+        cms = item.get('cms', '')
+        techs = [t.strip() for t in cms.split(',') if t.strip()] if cms else []
         
-        return (url, techs) if techs else None
+        return {
+            'url': url,
+            'techs': techs,
+            'server': item.get('server', ''),
+            'title': item.get('title', ''),
+            'status_code': item.get('status_code'),
+            'content_length': item.get('length'),
+        }
         
     except json.JSONDecodeError:
         return None
 
 
-def bulk_merge_tech_field(
-    source: str,
-    url_techs_map: dict[str, list[str]],
+def bulk_merge_website_fields(
+    records: list[dict],
     target_id: int
 ) -> dict:
     """
-    批量合并 tech 数组字段（PostgreSQL 原生 SQL）
+    批量合并更新 WebSite 字段（PostgreSQL 原生 SQL）
+    
+    合并策略：
+    - tech：数组合并去重
+    - title, webserver, status_code, content_length：只在原值为空/NULL 时更新
     
-    使用 PostgreSQL 原生 SQL 实现高效的数组合并去重操作。
     如果 URL 对应的记录不存在，会自动创建新记录。
     
+    Args:
+        records: 解析后的记录列表，每个包含 {url, techs, server, title, status_code, content_length}
+        target_id: 目标 ID
+    
     Returns:
         dict: {'updated_count': int, 'created_count': int}
     """
-    Model = _get_model_class(source)
-    table_name = Model._meta.db_table
+    from apps.asset.models import WebSite
+    table_name = WebSite._meta.db_table
     
     updated_count = 0
     created_count = 0
     
     with connection.cursor() as cursor:
-        for url, techs in url_techs_map.items():
-            if not techs:
-                continue
+        for record in records:
+            url = record['url']
+            techs = record.get('techs', [])
+            server = record.get('server', '') or ''
+            title = record.get('title', '') or ''
+            status_code = record.get('status_code')
+            content_length = record.get('content_length')
             
-            # 先尝试更新（PostgreSQL 数组合并去重）
-            sql = f"""
+            # 先尝试更新（合并策略）
+            update_sql = f"""
                 UPDATE {table_name}
-                SET tech = (
-                    SELECT ARRAY(SELECT DISTINCT unnest(
+                SET 
+                    tech = (SELECT ARRAY(SELECT DISTINCT unnest(
                         COALESCE(tech, ARRAY[]::varchar[]) || %s::varchar[]
-                    ))
-                )
+                    ))),
+                    title = CASE WHEN title = '' OR title IS NULL THEN %s ELSE title END,
+                    webserver = CASE WHEN webserver = '' OR webserver IS NULL THEN %s ELSE webserver END,
+                    status_code = CASE WHEN status_code IS NULL THEN %s ELSE status_code END,
+                    content_length = CASE WHEN content_length IS NULL THEN %s ELSE content_length END
                 WHERE url = %s AND target_id = %s
             """
             
-            cursor.execute(sql, [techs, url, target_id])
+            cursor.execute(update_sql, [techs, title, server, status_code, content_length, url, target_id])
             
             if cursor.rowcount > 0:
                 updated_count += cursor.rowcount
@@ -114,20 +117,26 @@ def bulk_merge_tech_field(
                     
                     # 插入新记录（带冲突处理）
                     insert_sql = f"""
-                        INSERT INTO {table_name} (target_id, url, host, tech, created_at)
-                        VALUES (%s, %s, %s, %s::varchar[], NOW())
+                        INSERT INTO {table_name} (
+                            target_id, url, host, location, title, webserver, 
+                            response_body, content_type, tech, status_code, content_length,
+                            response_headers, created_at
+                        )
+                        VALUES (%s, %s, %s, '', %s, %s, '', '', %s::varchar[], %s, %s, '', NOW())
                         ON CONFLICT (target_id, url) DO UPDATE SET
-                            tech = (
-                                SELECT ARRAY(SELECT DISTINCT unnest(
-                                    COALESCE({table_name}.tech, ARRAY[]::varchar[]) || EXCLUDED.tech
-                                ))
-                            )
+                            tech = (SELECT ARRAY(SELECT DISTINCT unnest(
+                                COALESCE({table_name}.tech, ARRAY[]::varchar[]) || EXCLUDED.tech
+                            ))),
+                            title = CASE WHEN {table_name}.title = '' OR {table_name}.title IS NULL THEN EXCLUDED.title ELSE {table_name}.title END,
+                            webserver = CASE WHEN {table_name}.webserver = '' OR {table_name}.webserver IS NULL THEN EXCLUDED.webserver ELSE {table_name}.webserver END,
+                            status_code = CASE WHEN {table_name}.status_code IS NULL THEN EXCLUDED.status_code ELSE {table_name}.status_code END,
+                            content_length = CASE WHEN {table_name}.content_length IS NULL THEN EXCLUDED.content_length ELSE {table_name}.content_length END
                     """
-                    cursor.execute(insert_sql, [target_id, url, host, techs])
+                    cursor.execute(insert_sql, [target_id, url, host, title, server, techs, status_code, content_length])
                     created_count += 1
                     
                 except Exception as e:
-                    logger.warning("创建 %s 记录失败 (url=%s): %s", source, url, e)
+                    logger.warning("创建 WebSite 记录失败 (url=%s): %s", url, e)
     
     return {
         'updated_count': updated_count,
@@ -141,12 +150,12 @@ def _parse_xingfinger_stream_output(
     cwd: Optional[str] = None,
     timeout: Optional[int] = None,
     log_file: Optional[str] = None
-) -> Generator[tuple[str, list[str]], None, None]:
+) -> Generator[dict, None, None]:
     """
     流式解析 xingfinger 命令输出
     
     基于 execute_stream 实时处理 xingfinger 命令的 stdout，将每行 JSON 输出
-    转换为 (url, tech_list) 格式
+    转换为完整字段字典
     """
     logger.info("开始流式解析 xingfinger 命令输出 - 命令: %s", cmd)
     
@@ -193,43 +202,46 @@ def run_xingfinger_and_stream_update_tech_task(
     batch_size: int = 100
 ) -> dict:
     """
-    流式执行 xingfinger 命令并实时更新 tech 字段
-    
-    根据 source 参数更新对应表的 tech 字段：
-    - website → WebSite.tech
-    - endpoint → Endpoint.tech（以后扩展）
+    流式执行 xingfinger 命令，保存快照并合并更新资产表
     
     处理流程：
     1. 流式执行 xingfinger 命令
-    2. 实时解析 JSON 输出
-    3. 累积到 batch_size 条后批量更新数据库
-    4. 使用 PostgreSQL 原生 SQL 进行数组合并去重
-    5. 如果记录不存在，自动创建
+    2. 实时解析 JSON 输出（完整字段）
+    3. 累积到 batch_size 条后批量处理：
+       - 保存快照（WebsiteSnapshot）
+       - 合并更新资产表（WebSite）
+    
+    合并策略：
+    - tech：数组合并去重
+    - title, webserver, status_code, content_length：只在原值为空时更新
     
     Returns:
         dict: {
             'processed_records': int,
             'updated_count': int,
             'created_count': int,
+            'snapshot_count': int,
             'batch_count': int
         }
     """
     logger.info(
-        "开始执行 xingfinger 并更新 tech - target_id=%s, source=%s, timeout=%s秒",
-        target_id, source, timeout
+        "开始执行 xingfinger - scan_id=%s, target_id=%s, timeout=%s秒",
+        scan_id, target_id, timeout
     )
     
     data_generator = None
+    snapshot_repo = DjangoWebsiteSnapshotRepository()
     
     try:
         # 初始化统计
         processed_records = 0
         updated_count = 0
         created_count = 0
+        snapshot_count = 0
         batch_count = 0
         
-        # 当前批次的 URL -> techs 映射
-        url_techs_map = {}
+        # 当前批次的记录列表
+        batch_records = []
         
         # 流式处理
         data_generator = _parse_xingfinger_stream_output(
@@ -240,47 +252,43 @@ def run_xingfinger_and_stream_update_tech_task(
             log_file=log_file
         )
         
-        for url, techs in data_generator:
+        for record in data_generator:
             processed_records += 1
+            batch_records.append(record)
             
-            # 累积到 url_techs_map
-            if url in url_techs_map:
-                # 合并同一 URL 的多次识别结果
-                url_techs_map[url].extend(techs)
-            else:
-                url_techs_map[url] = techs
-            
-            # 达到批次大小，执行批量更新
-            if len(url_techs_map) >= batch_size:
+            # 达到批次大小，执行批量处理
+            if len(batch_records) >= batch_size:
                 batch_count += 1
-                result = bulk_merge_tech_field(source, url_techs_map, target_id)
-                updated_count += result['updated_count']
-                created_count += result.get('created_count', 0)
-                
-                logger.debug(
-                    "批次 %d 完成 - 更新: %d, 创建: %d",
-                    batch_count, result['updated_count'], result.get('created_count', 0)
+                result = _process_batch(
+                    batch_records, scan_id, target_id, batch_count, snapshot_repo
                 )
+                updated_count += result['updated_count']
+                created_count += result['created_count']
+                snapshot_count += result['snapshot_count']
                 
                 # 清空批次
-                url_techs_map = {}
+                batch_records = []
         
         # 处理最后一批
-        if url_techs_map:
+        if batch_records:
             batch_count += 1
-            result = bulk_merge_tech_field(source, url_techs_map, target_id)
+            result = _process_batch(
+                batch_records, scan_id, target_id, batch_count, snapshot_repo
+            )
             updated_count += result['updated_count']
-            created_count += result.get('created_count', 0)
+            created_count += result['created_count']
+            snapshot_count += result['snapshot_count']
         
         logger.info(
-            "✓ xingfinger 执行完成 - 处理记录: %d, 更新: %d, 创建: %d, 批次: %d",
-            processed_records, updated_count, created_count, batch_count
+            "✓ xingfinger 执行完成 - 处理: %d, 更新: %d, 创建: %d, 快照: %d, 批次: %d",
+            processed_records, updated_count, created_count, snapshot_count, batch_count
         )
         
         return {
             'processed_records': processed_records,
             'updated_count': updated_count,
             'created_count': created_count,
+            'snapshot_count': snapshot_count,
             'batch_count': batch_count
         }
         
@@ -298,3 +306,67 @@ def run_xingfinger_and_stream_update_tech_task(
                 data_generator.close()
             except Exception as e:
                 logger.debug("关闭生成器时出错: %s", e)
+
+
+def _process_batch(
+    records: list[dict],
+    scan_id: int,
+    target_id: int,
+    batch_num: int,
+    snapshot_repo: DjangoWebsiteSnapshotRepository
+) -> dict:
+    """
+    处理一个批次的数据：保存快照 + 合并更新资产表
+    
+    Args:
+        records: 解析后的记录列表
+        scan_id: 扫描任务 ID
+        target_id: 目标 ID
+        batch_num: 批次编号
+        snapshot_repo: 快照仓库
+    
+    Returns:
+        dict: {'updated_count': int, 'created_count': int, 'snapshot_count': int}
+    """
+    # 1. 构建快照 DTO 列表
+    snapshot_dtos = []
+    for record in records:
+        # 从 URL 提取 host
+        parsed = urlparse(record['url'])
+        host = parsed.hostname or ''
+        
+        dto = WebsiteSnapshotDTO(
+            scan_id=scan_id,
+            target_id=target_id,
+            url=record['url'],
+            host=host,
+            title=record.get('title', '') or '',
+            status_code=record.get('status_code'),
+            content_length=record.get('content_length'),
+            webserver=record.get('server', '') or '',
+            tech=record.get('techs', []),
+        )
+        snapshot_dtos.append(dto)
+    
+    # 2. 保存快照
+    snapshot_count = 0
+    if snapshot_dtos:
+        try:
+            snapshot_repo.save_snapshots(snapshot_dtos)
+            snapshot_count = len(snapshot_dtos)
+        except Exception as e:
+            logger.warning("批次 %d 保存快照失败: %s", batch_num, e)
+    
+    # 3. 合并更新资产表
+    merge_result = bulk_merge_website_fields(records, target_id)
+    
+    logger.debug(
+        "批次 %d 完成 - 更新: %d, 创建: %d, 快照: %d",
+        batch_num, merge_result['updated_count'], merge_result['created_count'], snapshot_count
+    )
+    
+    return {
+        'updated_count': merge_result['updated_count'],
+        'created_count': merge_result['created_count'],
+        'snapshot_count': snapshot_count
+    }

backend/apps/scan/tasks/site_scan/run_and_stream_save_websites_task.py

@@ -30,7 +30,6 @@ from typing import Generator, Optional, Dict, Any, TYPE_CHECKING
 from django.db import IntegrityError, OperationalError, DatabaseError
 from dataclasses import dataclass
 from urllib.parse import urlparse, urlunparse
-from dateutil.parser import parse as parse_datetime
 from psycopg2 import InterfaceError
 
 from apps.asset.dtos.snapshot import WebsiteSnapshotDTO
@@ -62,6 +61,18 @@ class ServiceSet:
         )
 
 
+def _sanitize_string(value: str) -> str:
+    """
+    清理字符串中的 NUL 字符和其他不可打印字符
+    
+    PostgreSQL 不允许字符串字段包含 NUL (0x00) 字符
+    """
+    if not value:
+        return value
+    # 移除 NUL 字符
+    return value.replace('\x00', '')
+
+
 def normalize_url(url: str) -> str:
     """
     标准化 URL，移除默认端口号
@@ -117,69 +128,50 @@ def normalize_url(url: str) -> str:
         return url
 
 
+def _extract_hostname(url: str) -> str:
+    """
+    从 URL 提取主机名
+    
+    Args:
+        url: URL 字符串
+    
+    Returns:
+        str: 提取的主机名（小写）
+    """
+    try:
+        if url:
+            parsed = urlparse(url)
+            if parsed.hostname:
+                return parsed.hostname
+            # 降级方案：手动提取
+            return url.replace('http://', '').replace('https://', '').split('/')[0].split(':')[0]
+        return ''
+    except Exception as e:
+        logger.debug("提取主机名失败: %s", e)
+        return ''
+
+
 class HttpxRecord:
     """httpx 扫描记录数据类"""
     
     def __init__(self, data: Dict[str, Any]):
-        self.url = data.get('url', '')
-        self.input = data.get('input', '')
-        self.title = data.get('title', '')
-        self.status_code = data.get('status_code')
-        self.content_length = data.get('content_length')
-        self.content_type = data.get('content_type', '')
-        self.location = data.get('location', '')
-        self.webserver = data.get('webserver', '')
-        self.body_preview = data.get('body_preview', '')
-        self.tech = data.get('tech', [])
-        self.vhost = data.get('vhost')
-        self.failed = data.get('failed', False)
-        self.timestamp = data.get('timestamp')
+        self.url = _sanitize_string(data.get('url', ''))
+        self.input = _sanitize_string(data.get('input', ''))
+        self.title = _sanitize_string(data.get('title', ''))
+        self.status_code = data.get('status_code')  # int，不需要清理
+        self.content_length = data.get('content_length')  # int，不需要清理
+        self.content_type = _sanitize_string(data.get('content_type', ''))
+        self.location = _sanitize_string(data.get('location', ''))
+        self.webserver = _sanitize_string(data.get('webserver', ''))
+        self.response_body = _sanitize_string(data.get('body', ''))
+        self.tech = [_sanitize_string(t) for t in data.get('tech', []) if isinstance(t, str)]  # 列表中的字符串也需要清理
+        self.vhost = data.get('vhost')  # bool，不需要清理
+        self.failed = data.get('failed', False)  # bool，不需要清理
+        self.response_headers = _sanitize_string(data.get('raw_header', ''))
         
-        # 从 URL 中提取主机名
-        self.host = self._extract_hostname()
-    
-    def _extract_hostname(self) -> str:
-        """
-        从 URL 或 input 字段提取主机名
-        
-        优先级：
-        1. 使用 urlparse 解析 URL 获取 hostname
-        2. 从 input 字段提取（处理可能包含协议的情况）
-        3. 从 URL 字段手动提取（降级方案）
-        
-        Returns:
-            str: 提取的主机名（小写）
-        """
-        try:
-            # 方法 1: 使用 urlparse 解析 URL
-            if self.url:
-                parsed = urlparse(self.url)
-                if parsed.hostname:
-                    return parsed.hostname
-            
-            # 方法 2: 从 input 字段提取
-            if self.input:
-                host = self.input.strip().lower()
-                # 移除协议前缀
-                if host.startswith(('http://', 'https://')):
-                    host = host.split('//', 1)[1].split('/')[0]
-                return host
-            
-            # 方法 3: 从 URL 手动提取（降级方案）
-            if self.url:
-                return self.url.replace('http://', '').replace('https://', '').split('/')[0]
-            
-            # 兜底：返回空字符串
-            return ''
-            
-        except Exception as e:
-            # 异常处理：尽力从 input 或 URL 提取
-            logger.debug("提取主机名失败: %s，使用降级方案", e)
-            if self.input:
-                return self.input.strip().lower()
-            if self.url:
-                return self.url.replace('http://', '').replace('https://', '').split('/')[0]
-            return ''
+        # 从 URL 中提取主机名（优先使用 httpx 返回的 host，否则自动提取）
+        httpx_host = _sanitize_string(data.get('host', ''))
+        self.host = httpx_host if httpx_host else _extract_hostname(self.url)
 
 
 def _save_batch_with_retry(
@@ -227,39 +219,31 @@ def _save_batch_with_retry(
             }
         
         except (OperationalError, DatabaseError, InterfaceError) as e:
-            # 数据库连接/操作错误，可重试
+            # 数据库级错误（连接中断、表结构不匹配等）：按指数退避重试，最终失败时抛出异常让 Flow 失败
             if attempt < max_retries - 1:
-                wait_time = 2 ** attempt  # 指数退避: 1s, 2s, 4s
+                wait_time = 2 ** attempt
                 logger.warning(
                     "批次 %d 保存失败（第 %d 次尝试），%d秒后重试: %s",
                     batch_num, attempt + 1, wait_time, str(e)[:100]
                 )
                 time.sleep(wait_time)
             else:
-                logger.error("批次 %d 保存失败（已重试 %d 次）: %s", batch_num, max_retries, e)
-                return {
-                    'success': False,
-                    'created_websites': 0,
-                    'skipped_failed': 0
-                }
-        
-        except Exception as e:
-            # 其他未知错误 - 检查是否为连接问题
-            error_str = str(e).lower()
-            if 'connection' in error_str and attempt < max_retries - 1:
-                logger.warning(
-                    "批次 %d 连接相关错误（尝试 %d/%d）: %s，Repository 装饰器会自动重连",
-                    batch_num, attempt + 1, max_retries, str(e)
+                logger.error(
+                    "批次 %d 保存失败（已重试 %d 次），将终止任务: %s",
+                    batch_num,
+                    max_retries,
+                    e,
+                    exc_info=True,
                 )
-                time.sleep(2)
-            else:
-                logger.error("批次 %d 未知错误: %s", batch_num, e, exc_info=True)
-                return {
-                    'success': False,
-                    'created_websites': 0,
-                    'skipped_failed': 0
-                }
-    
+                # 让上层 Task 感知失败，从而标记整个扫描为失败
+                raise
+
+        except Exception as e:
+            # 其他未知异常也不再吞掉，直接抛出以便 Flow 标记为失败
+            logger.error("批次 %d 未知错误: %s", batch_num, e, exc_info=True)
+            raise
+
+    # 理论上不会走到这里，保留兜底返回值以满足类型约束
     return {
         'success': False,
         'created_websites': 0,
@@ -327,42 +311,39 @@ def _save_batch(
             skipped_failed += 1
             continue
         
-        # 解析时间戳
-        created_at = None
-        if hasattr(record, 'timestamp') and record.timestamp:
-            try:
-                created_at = parse_datetime(record.timestamp)
-            except (ValueError, TypeError) as e:
-                logger.warning(f"无法解析时间戳 {record.timestamp}: {e}")
-        
-        # 使用 input 字段（原始扫描的 URL）而不是 url 字段（重定向后的 URL）
-        # 原因：避免多个不同的输入 URL 重定向到同一个 URL 时产生唯一约束冲突
-        # 例如：http://example.com 和 https://example.com 都重定向到 https://example.com
-        # 如果使用 record.url，两条记录会有相同的 url，导致数据库冲突
-        # 如果使用 record.input，两条记录保留原始输入，不会冲突
-        normalized_url = normalize_url(record.input)
-        
-        # 提取 host 字段（域名或IP地址）
-        host = record.host if record.host else ''
-        
-        # 创建 WebsiteSnapshot DTO
-        snapshot_dto = WebsiteSnapshotDTO(
-            scan_id=scan_id,
-            target_id=target_id,  # 主关联字段
-            url=normalized_url,  # 保存原始输入 URL（归一化后）
-            host=host,  # 主机名（域名或IP地址）
-            location=record.location,  # location 字段保存重定向信息
-            title=record.title[:1000] if record.title else '',
-            web_server=record.webserver[:200] if record.webserver else '',
-            body_preview=record.body_preview[:1000] if record.body_preview else '',
-            content_type=record.content_type[:200] if record.content_type else '',
-            tech=record.tech if isinstance(record.tech, list) else [],
-            status=record.status_code,
-            content_length=record.content_length,
-            vhost=record.vhost
-        )
-        
-        snapshot_items.append(snapshot_dto)
+        try:
+            # 使用 input 字段（原始扫描的 URL）而不是 url 字段（重定向后的 URL）
+            # 原因：避免多个不同的输入 URL 重定向到同一个 URL 时产生唯一约束冲突
+            # 例如：http://example.com 和 https://example.com 都重定向到 https://example.com
+            # 如果使用 record.url，两条记录会有相同的 url，导致数据库冲突
+            # 如果使用 record.input，两条记录保留原始输入，不会冲突
+            normalized_url = normalize_url(record.input) if record.input else normalize_url(record.url)
+            
+            # 提取 host 字段（域名或IP地址）
+            host = record.host if record.host else ''
+            
+            # 创建 WebsiteSnapshot DTO
+            snapshot_dto = WebsiteSnapshotDTO(
+                scan_id=scan_id,
+                target_id=target_id,  # 主关联字段
+                url=normalized_url,  # 保存原始输入 URL（归一化后）
+                host=host,  # 主机名（域名或IP地址）
+                location=record.location if record.location else '',
+                title=record.title if record.title else '',
+                webserver=record.webserver if record.webserver else '',
+                response_body=record.response_body if record.response_body else '',
+                content_type=record.content_type if record.content_type else '',
+                tech=record.tech if isinstance(record.tech, list) else [],
+                status_code=record.status_code,
+                content_length=record.content_length,
+                vhost=record.vhost,
+                response_headers=record.response_headers if record.response_headers else '',
+            )
+            
+            snapshot_items.append(snapshot_dto)
+        except Exception as e:
+            logger.error("处理记录失败: %s，错误: %s", record.url, e)
+            continue
     
     # ========== Step 3: 保存快照并同步到资产表（通过快照 Service）==========
     if snapshot_items:
@@ -384,28 +365,31 @@ def _parse_and_validate_line(line: str) -> Optional[HttpxRecord]:
         Optional[HttpxRecord]: 有效的 httpx 扫描记录，或 None 如果验证失败
     
     验证步骤：
-        1. 解析 JSON 格式
-        2. 验证数据类型为字典
-        3. 创建 HttpxRecord 对象
-        4. 验证必要字段（url）
+        1. 清理 NUL 字符
+        2. 解析 JSON 格式
+        3. 验证数据类型为字典
+        4. 创建 HttpxRecord 对象
+        5. 验证必要字段（url）
     """
     try:
-        # 步骤 1: 解析 JSON
+        # 步骤 1: 清理 NUL 字符后再解析 JSON
+        line = _sanitize_string(line)
+        
+        # 步骤 2: 解析 JSON
         try:
             line_data = json.loads(line, strict=False)
         except json.JSONDecodeError:
-            # logger.info("跳过非 JSON 行: %s", line)
             return None
         
-        # 步骤 2: 验证数据类型
+        # 步骤 3: 验证数据类型
         if not isinstance(line_data, dict):
             logger.info("跳过非字典数据")
             return None
         
-        # 步骤 3: 创建记录
+        # 步骤 4: 创建记录
         record = HttpxRecord(line_data)
         
-        # 步骤 4: 验证必要字段
+        # 步骤 5: 验证必要字段
         if not record.url:
             logger.info("URL 为空，跳过 - 数据: %s", str(line_data)[:200])
             return None
@@ -414,7 +398,7 @@ def _parse_and_validate_line(line: str) -> Optional[HttpxRecord]:
         return record
     
     except Exception:
-        logger.info("跳过无法解析的行: %s", line[:100])
+        logger.info("跳过无法解析的行: %s", line[:100] if line else 'empty')
         return None
 
 
@@ -462,8 +446,8 @@ def _parse_httpx_stream_output(
             # yield 一条有效记录
             yield record
             
-            # 每处理 1000 条记录输出一次进度
-            if valid_records % 1000 == 0:
+            # 每处理 5 条记录输出一次进度
+            if valid_records % 5 == 0:
                 logger.info("已解析 %d 条有效记录...", valid_records)
                 
     except subprocess.TimeoutExpired as e:
@@ -602,8 +586,8 @@ def _process_records_in_batches(
             _process_batch(batch, scan_id, target_id, batch_num, total_stats, failed_batches, services)
             batch = []  # 清空批次
             
-            # 每20个批次输出进度
-            if batch_num % 20 == 0:
+            # 每 2 个批次输出进度
+            if batch_num % 2 == 0:
                 logger.info("进度: 已处理 %d 批次，%d 条记录", batch_num, total_records)
     
     # 保存最后一批
@@ -674,11 +658,7 @@ def _cleanup_resources(data_generator) -> None:
             logger.error("关闭生成器时出错: %s", gen_close_error)
 
 
-@task(
-    name='run_and_stream_save_websites',
-    retries=0,
-    log_prints=True
-)
+@task(name='run_and_stream_save_websites', retries=0)
 def run_and_stream_save_websites_task(
     cmd: str,
     tool_name: str,
@@ -686,7 +666,7 @@ def run_and_stream_save_websites_task(
     target_id: int,
     cwd: Optional[str] = None,
     shell: bool = False,
-    batch_size: int = 1000,
+    batch_size: int = 10,
     timeout: Optional[int] = None,
     log_file: Optional[str] = None
 ) -> dict:

backend/apps/scan/tasks/url_fetch/run_and_stream_save_urls_task.py

@@ -2,8 +2,8 @@
 基于 execute_stream 的流式 URL 验证任务
 
 主要功能：
-    1. 实时执行 httpx 命令验证 URL 存活
-    2. 流式处理命令输出，解析存活的 URL
+    1. 实时执行 httpx 命令验证 URL
+    2. 流式处理命令输出，解析 URL 信息
     3. 批量保存到数据库（Endpoint 表）
     4. 避免一次性加载所有 URL 到内存
 
@@ -14,7 +14,7 @@
     - 使用 execute_stream 实时处理输出
     - 流式处理避免内存溢出
     - 批量操作减少数据库交互
-    - 只保存存活的 URL（status 2xx/3xx）
+    - 保存所有有效 URL（包括 4xx/5xx，便于安全分析）
 """
 
 import logging
@@ -23,10 +23,11 @@ import subprocess
 import time
 from pathlib import Path
 from prefect import task
-from typing import Generator, Optional
+from typing import Generator, Optional, Dict, Any
 from django.db import IntegrityError, OperationalError, DatabaseError
 from psycopg2 import InterfaceError
 from dataclasses import dataclass
+from urllib.parse import urlparse
 
 from apps.asset.services.snapshot import EndpointSnapshotsService
 from apps.scan.utils import execute_stream
@@ -63,7 +64,53 @@ def _sanitize_string(value: str) -> str:
     return value.replace('\x00', '')
 
 
-def _parse_and_validate_line(line: str) -> Optional[dict]:
+def _extract_hostname(url: str) -> str:
+    """
+    从 URL 提取主机名
+    
+    Args:
+        url: URL 字符串
+    
+    Returns:
+        str: 提取的主机名（小写）
+    """
+    try:
+        if url:
+            parsed = urlparse(url)
+            if parsed.hostname:
+                return parsed.hostname
+            # 降级方案：手动提取
+            return url.replace('http://', '').replace('https://', '').split('/')[0].split(':')[0]
+        return ''
+    except Exception as e:
+        logger.debug("提取主机名失败: %s", e)
+        return ''
+
+
+class HttpxRecord:
+    """httpx 扫描记录数据类"""
+    
+    def __init__(self, data: Dict[str, Any]):
+        self.url = _sanitize_string(data.get('url', ''))
+        self.input = _sanitize_string(data.get('input', ''))
+        self.title = _sanitize_string(data.get('title', ''))
+        self.status_code = data.get('status_code')  # int，不需要清理
+        self.content_length = data.get('content_length')  # int，不需要清理
+        self.content_type = _sanitize_string(data.get('content_type', ''))
+        self.location = _sanitize_string(data.get('location', ''))
+        self.webserver = _sanitize_string(data.get('webserver', ''))
+        self.response_body = _sanitize_string(data.get('body', ''))
+        self.tech = [_sanitize_string(t) for t in data.get('tech', []) if isinstance(t, str)]  # 列表中的字符串也需要清理
+        self.vhost = data.get('vhost')  # bool，不需要清理
+        self.failed = data.get('failed', False)  # bool，不需要清理
+        self.response_headers = _sanitize_string(data.get('raw_header', ''))
+        
+        # 从 URL 中提取主机名（优先使用 httpx 返回的 host，否则自动提取）
+        httpx_host = _sanitize_string(data.get('host', ''))
+        self.host = httpx_host if httpx_host else _extract_hostname(self.url)
+
+
+def _parse_and_validate_line(line: str) -> Optional[HttpxRecord]:
     """
     解析并验证单行 httpx JSON 输出
     
@@ -71,9 +118,7 @@ def _parse_and_validate_line(line: str) -> Optional[dict]:
         line: 单行输出数据
     
     Returns:
-        Optional[dict]: 有效的 httpx 记录，或 None 如果验证失败
-        
-    只返回存活的 URL（2xx/3xx 状态码）
+        Optional[HttpxRecord]: 有效的 httpx 记录，或 None 如果验证失败
     """
     try:
         # 清理 NUL 字符后再解析 JSON
@@ -83,7 +128,6 @@ def _parse_and_validate_line(line: str) -> Optional[dict]:
         try:
             line_data = json.loads(line, strict=False)
         except json.JSONDecodeError:
-            # logger.info("跳过非 JSON 行: %s", line)
             return None
         
         # 验证数据类型
@@ -91,32 +135,15 @@ def _parse_and_validate_line(line: str) -> Optional[dict]:
             logger.info("跳过非字典数据")
             return None
         
-        # 获取必要字段
-        url = line_data.get('url', '').strip()
-        status_code = line_data.get('status_code')
+        # 创建记录
+        record = HttpxRecord(line_data)
         
-        if not url:
+        # 验证必要字段
+        if not record.url:
             logger.info("URL 为空，跳过 - 数据: %s", str(line_data)[:200])
             return None
         
-        # 只保存存活的 URL（2xx 或 3xx）
-        if status_code and (200 <= status_code < 400):
-            return {
-                'url': _sanitize_string(url),
-                'host': _sanitize_string(line_data.get('host', '')),
-                'status_code': status_code,
-                'title': _sanitize_string(line_data.get('title', '')),
-                'content_length': line_data.get('content_length', 0),
-                'content_type': _sanitize_string(line_data.get('content_type', '')),
-                'webserver': _sanitize_string(line_data.get('webserver', '')),
-                'location': _sanitize_string(line_data.get('location', '')),
-                'tech': line_data.get('tech', []),
-                'body_preview': _sanitize_string(line_data.get('body_preview', '')),
-                'vhost': line_data.get('vhost', False),
-            }
-        else:
-            logger.debug("URL 不存活（状态码: %s），跳过: %s", status_code, url)
-            return None
+        return record
     
     except Exception:
         logger.info("跳过无法解析的行: %s", line[:100] if line else 'empty')
@@ -130,7 +157,7 @@ def _parse_httpx_stream_output(
     shell: bool = False,
     timeout: Optional[int] = None,
     log_file: Optional[str] = None
-) -> Generator[dict, None, None]:
+) -> Generator[HttpxRecord, None, None]:
     """
     流式解析 httpx 命令输出
     
@@ -143,7 +170,7 @@ def _parse_httpx_stream_output(
         log_file: 日志文件路径
     
     Yields:
-        dict: 每次 yield 一条存活的 URL 记录
+        HttpxRecord: 每次 yield 一条存活的 URL 记录
     """
     logger.info("开始流式解析 httpx 输出 - 命令: %s", cmd)
     
@@ -173,8 +200,8 @@ def _parse_httpx_stream_output(
             # yield 一条有效记录（存活的 URL）
             yield record
             
-            # 每处理 500 条记录输出一次进度
-            if valid_records % 500 == 0:
+            # 每处理 100 条记录输出一次进度
+            if valid_records % 100 == 0:
                 logger.info("已解析 %d 条存活的 URL...", valid_records)
                 
     except subprocess.TimeoutExpired as e:
@@ -191,6 +218,78 @@ def _parse_httpx_stream_output(
     )
 
 
+def _validate_task_parameters(cmd: str, target_id: int, scan_id: int, cwd: Optional[str]) -> None:
+    """
+    验证任务参数的有效性
+    
+    Args:
+        cmd: 扫描命令
+        target_id: 目标ID
+        scan_id: 扫描ID
+        cwd: 工作目录
+        
+    Raises:
+        ValueError: 参数验证失败
+    """
+    if not cmd or not cmd.strip():
+        raise ValueError("扫描命令不能为空")
+    
+    if target_id is None:
+        raise ValueError("target_id 不能为 None，必须指定目标ID")
+        
+    if scan_id is None:
+        raise ValueError("scan_id 不能为 None，必须指定扫描ID")
+    
+    # 验证工作目录（如果指定）
+    if cwd and not Path(cwd).exists():
+        raise ValueError(f"工作目录不存在: {cwd}")
+
+
+def _build_final_result(stats: dict) -> dict:
+    """
+    构建最终结果并输出日志
+    
+    Args:
+        stats: 处理统计信息
+        
+    Returns:
+        dict: 最终结果
+    """
+    logger.info(
+        "✓ URL 验证任务完成 - 处理记录: %d（%d 批次），创建端点: %d，跳过（失败）: %d",
+        stats['processed_records'], stats['batch_count'], stats['created_endpoints'],
+        stats['skipped_failed']
+    )
+    
+    # 如果没有创建任何记录，给出明确提示
+    if stats['created_endpoints'] == 0:
+        logger.warning(
+            "⚠️  没有创建任何端点记录！可能原因：1) 命令输出格式问题 2) 重复数据被忽略 3) 所有请求都失败"
+        )
+    
+    return {
+        'processed_records': stats['processed_records'],
+        'created_endpoints': stats['created_endpoints'],
+        'skipped_failed': stats['skipped_failed']
+    }
+
+
+def _cleanup_resources(data_generator) -> None:
+    """
+    清理任务资源
+    
+    Args:
+        data_generator: 数据生成器
+    """
+    # 确保生成器被正确关闭
+    if data_generator is not None:
+        try:
+            data_generator.close()
+            logger.debug("已关闭数据生成器")
+        except Exception as gen_close_error:
+            logger.error("关闭生成器时出错: %s", gen_close_error)
+
+
 def _save_batch_with_retry(
     batch: list,
     scan_id: int,
@@ -211,14 +310,19 @@ def _save_batch_with_retry(
         max_retries: 最大重试次数
     
     Returns:
-        dict: {'success': bool, 'saved_count': int}
+        dict: {
+            'success': bool,
+            'created_endpoints': int,
+            'skipped_failed': int
+        }
     """
     for attempt in range(max_retries):
         try:
-            count = _save_batch(batch, scan_id, target_id, batch_num, services)
+            stats = _save_batch(batch, scan_id, target_id, batch_num, services)
             return {
                 'success': True,
-                'saved_count': count
+                'created_endpoints': stats.get('created_endpoints', 0),
+                'skipped_failed': stats.get('skipped_failed', 0)
             }
 
         except IntegrityError as e:
@@ -226,7 +330,8 @@ def _save_batch_with_retry(
             logger.error("批次 %d 数据完整性错误，跳过: %s", batch_num, str(e)[:100])
             return {
                 'success': False,
-                'saved_count': 0
+                'created_endpoints': 0,
+                'skipped_failed': 0
             }
 
         except (OperationalError, DatabaseError, InterfaceError) as e:
@@ -257,7 +362,8 @@ def _save_batch_with_retry(
     # 理论上不会走到这里，保留兜底返回值以满足类型约束
     return {
         'success': False,
-        'saved_count': 0
+        'created_endpoints': 0,
+        'skipped_failed': 0
     }
 
 
@@ -267,49 +373,72 @@ def _save_batch(
     target_id: int,
     batch_num: int,
     services: ServiceSet
-) -> int:
+) -> dict:
     """
     保存一个批次的数据到数据库
     
     Args:
-        batch: 数据批次，list of dict
+        batch: 数据批次，list of HttpxRecord
         scan_id: 扫描任务 ID
         target_id: 目标 ID
         batch_num: 批次编号
         services: Service 集合
     
     Returns:
-        int: 创建的记录数
+        dict: 包含创建和跳过记录的统计信息
     """
+    # 参数验证
+    if not isinstance(batch, list):
+        raise TypeError(f"batch 必须是 list 类型，实际: {type(batch).__name__}")
+    
     if not batch:
         logger.debug("批次 %d 为空，跳过处理", batch_num)
-        return 0
+        return {
+            'created_endpoints': 0,
+            'skipped_failed': 0
+        }
+    
+    # 统计变量
+    skipped_failed = 0
     
     # 批量构造 Endpoint 快照 DTO
     from apps.asset.dtos.snapshot import EndpointSnapshotDTO
     
     snapshots = []
     for record in batch:
+        # 跳过失败的请求
+        if record.failed:
+            skipped_failed += 1
+            continue
+        
         try:
+            # Endpoint URL 直接使用原始值，不做标准化
+            # 原因：Endpoint URL 来自 waymore/katana，包含路径和参数，标准化可能改变含义
+            url = record.input if record.input else record.url
+            
+            # 提取 host 字段（域名或IP地址）
+            host = record.host if record.host else ''
+            
             dto = EndpointSnapshotDTO(
                 scan_id=scan_id,
-                url=record['url'],
-                host=record.get('host', ''),
-                title=record.get('title', ''),
-                status_code=record.get('status_code'),
-                content_length=record.get('content_length', 0),
-                location=record.get('location', ''),
-                webserver=record.get('webserver', ''),
-                content_type=record.get('content_type', ''),
-                tech=record.get('tech', []),
-                body_preview=record.get('body_preview', ''),
-                vhost=record.get('vhost', False),
-                matched_gf_patterns=[],
                 target_id=target_id,
+                url=url,
+                host=host,
+                title=record.title if record.title else '',
+                status_code=record.status_code,
+                content_length=record.content_length,
+                location=record.location if record.location else '',
+                webserver=record.webserver if record.webserver else '',
+                content_type=record.content_type if record.content_type else '',
+                tech=record.tech if isinstance(record.tech, list) else [],
+                response_body=record.response_body if record.response_body else '',
+                vhost=record.vhost if record.vhost else False,
+                matched_gf_patterns=[],
+                response_headers=record.response_headers if record.response_headers else '',
             )
             snapshots.append(dto)
         except Exception as e:
-            logger.error("处理记录失败: %s，错误: %s", record.get('url', 'Unknown'), e)
+            logger.error("处理记录失败: %s，错误: %s", record.url, e)
             continue
     
     if snapshots:
@@ -318,15 +447,69 @@ def _save_batch(
             services.snapshot.save_and_sync(snapshots)
             count = len(snapshots)
             logger.info(
-                "批次 %d: 保存了 %d 个存活的 URL（共 %d 个）",
-                batch_num, count, len(batch)
+                "批次 %d: 保存了 %d 个存活的 URL（共 %d 个，跳过失败: %d）",
+                batch_num, count, len(batch), skipped_failed
             )
-            return count
+            return {
+                'created_endpoints': count,
+                'skipped_failed': skipped_failed
+            }
         except Exception as e:
             logger.error("批次 %d 批量保存失败: %s", batch_num, e)
             raise
     
-    return 0
+    return {
+        'created_endpoints': 0,
+        'skipped_failed': skipped_failed
+    }
+
+
+def _accumulate_batch_stats(total_stats: dict, batch_result: dict) -> None:
+    """
+    累加批次统计信息
+    
+    Args:
+        total_stats: 总统计信息字典
+        batch_result: 批次结果字典
+    """
+    total_stats['created_endpoints'] += batch_result.get('created_endpoints', 0)
+    total_stats['skipped_failed'] += batch_result.get('skipped_failed', 0)
+
+
+def _process_batch(
+    batch: list,
+    scan_id: int,
+    target_id: int,
+    batch_num: int,
+    total_stats: dict,
+    failed_batches: list,
+    services: ServiceSet
+) -> None:
+    """
+    处理单个批次
+    
+    Args:
+        batch: 数据批次
+        scan_id: 扫描ID
+        target_id: 目标ID
+        batch_num: 批次编号
+        total_stats: 总统计信息
+        failed_batches: 失败批次列表
+        services: Service 集合（必须，依赖注入）
+    """
+    result = _save_batch_with_retry(
+        batch, scan_id, target_id, batch_num, services
+    )
+    
+    # 累计统计信息（失败时可能有部分数据已保存）
+    _accumulate_batch_stats(total_stats, result)
+    
+    if not result['success']:
+        failed_batches.append(batch_num)
+        logger.warning(
+            "批次 %d 保存失败，但已累计统计信息：创建端点=%d",
+            batch_num, result.get('created_endpoints', 0)
+        )
 
 
 def _process_records_in_batches(
@@ -337,7 +520,7 @@ def _process_records_in_batches(
     services: ServiceSet
 ) -> dict:
     """
-    分批处理记录并保存到数据库
+    流式处理记录并分批保存
     
     Args:
         data_generator: 数据生成器
@@ -347,14 +530,23 @@ def _process_records_in_batches(
         services: Service 集合
         
     Returns:
-        dict: 处理统计结果
+        dict: 处理统计信息
+        
+    Raises:
+        RuntimeError: 存在失败批次时抛出
     """
-    batch = []
-    batch_num = 0
     total_records = 0
-    total_saved = 0
+    batch_num = 0
     failed_batches = []
+    batch = []
     
+    # 统计信息
+    total_stats = {
+        'created_endpoints': 0,
+        'skipped_failed': 0
+    }
+    
+    # 流式读取生成器并分批保存
     for record in data_generator:
         batch.append(record)
         total_records += 1
@@ -362,46 +554,35 @@ def _process_records_in_batches(
         # 达到批次大小，执行保存
         if len(batch) >= batch_size:
             batch_num += 1
-            result = _save_batch_with_retry(
-                batch, scan_id, target_id, batch_num, services
-            )
-            
-            if result['success']:
-                total_saved += result['saved_count']
-            else:
-                failed_batches.append(batch_num)
-            
+            _process_batch(batch, scan_id, target_id, batch_num, total_stats, failed_batches, services)
             batch = []  # 清空批次
             
             # 每 10 个批次输出进度
             if batch_num % 10 == 0:
-                logger.info(
-                    "进度: 已处理 %d 批次，%d 条记录，保存 %d 条",
-                    batch_num, total_records, total_saved
-                )
+                logger.info("进度: 已处理 %d 批次，%d 条记录", batch_num, total_records)
     
     # 保存最后一批
     if batch:
         batch_num += 1
-        result = _save_batch_with_retry(
-            batch, scan_id, target_id, batch_num, services
+        _process_batch(batch, scan_id, target_id, batch_num, total_stats, failed_batches, services)
+    
+    # 检查失败批次
+    if failed_batches:
+        error_msg = (
+            f"流式保存 URL 验证结果时出现失败批次，处理记录: {total_records}，"
+            f"失败批次: {failed_batches}"
         )
-        
-        if result['success']:
-            total_saved += result['saved_count']
-        else:
-            failed_batches.append(batch_num)
+        logger.warning(error_msg)
+        raise RuntimeError(error_msg)
     
     return {
         'processed_records': total_records,
-        'saved_urls': total_saved,
-        'failed_urls': total_records - total_saved,
         'batch_count': batch_num,
-        'failed_batches': failed_batches
+        **total_stats
     }
 
 
-@task(name="run_and_stream_save_urls", retries=3, retry_delay_seconds=10)
+@task(name="run_and_stream_save_urls", retries=0)
 def run_and_stream_save_urls_task(
     cmd: str,
     tool_name: str,
@@ -409,7 +590,7 @@ def run_and_stream_save_urls_task(
     target_id: int,
     cwd: Optional[str] = None,
     shell: bool = False,
-    batch_size: int = 500,
+    batch_size: int = 100,
     timeout: Optional[int] = None,
     log_file: Optional[str] = None
 ) -> dict:
@@ -417,17 +598,18 @@ def run_and_stream_save_urls_task(
     执行 httpx 验证并流式保存存活的 URL
     
     该任务将：
-    1. 执行 httpx 命令验证 URL 存活
-    2. 流式处理输出，实时解析
-    3. 批量保存存活的 URL 到 Endpoint 表
+    1. 验证输入参数
+    2. 初始化资源（缓存、生成器）
+    3. 流式处理记录并分批保存
+    4. 构建并返回结果统计
     
     Args:
         cmd: httpx 命令
         tool_name: 工具名称（'httpx'）
         scan_id: 扫描任务 ID
         target_id: 目标 ID
-        cwd: 工作目录
-        shell: 是否使用 shell 执行
+        cwd: 工作目录（可选）
+        shell: 是否使用 shell 执行（默认 False）
         batch_size: 批次大小（默认 500）
         timeout: 超时时间（秒）
         log_file: 日志文件路径
@@ -435,11 +617,14 @@ def run_and_stream_save_urls_task(
     Returns:
         dict: {
             'processed_records': int,  # 处理的记录总数
-            'saved_urls': int,         # 保存的存活 URL 数
-            'failed_urls': int,        # 失败/死链数
-            'batch_count': int,        # 批次数
-            'failed_batches': list     # 失败的批次号
+            'created_endpoints': int,  # 创建的端点记录数
+            'skipped_failed': int,     # 因请求失败跳过的记录数
         }
+    
+    Raises:
+        ValueError: 参数验证失败
+        RuntimeError: 命令执行或数据库操作失败
+        subprocess.TimeoutExpired: 命令执行超时
     """
     logger.info(
         "开始执行流式 URL 验证任务 - target_id=%s, 超时=%s秒, 命令: %s",
@@ -449,33 +634,30 @@ def run_and_stream_save_urls_task(
     data_generator = None
     
     try:
-        # 1. 初始化资源
+        # 1. 验证参数
+        _validate_task_parameters(cmd, target_id, scan_id, cwd)
+        
+        # 2. 初始化资源
         data_generator = _parse_httpx_stream_output(
             cmd, tool_name, cwd, shell, timeout, log_file
         )
         services = ServiceSet.create_default()
         
-        # 2. 流式处理记录并分批保存
+        # 3. 流式处理记录并分批保存
         stats = _process_records_in_batches(
             data_generator, scan_id, target_id, batch_size, services
         )
         
-        # 3. 输出最终统计
-        logger.info(
-            "✓ URL 验证任务完成 - 处理: %d, 存活: %d, 失败: %d",
-            stats['processed_records'],
-            stats['saved_urls'],
-            stats['failed_urls']
-        )
-        
-        return stats
+        # 4. 构建最终结果
+        return _build_final_result(stats)
         
     except subprocess.TimeoutExpired:
+        # 超时异常直接向上传播，保留异常类型
         logger.warning(
             "⚠️ URL 验证任务超时 - target_id=%s, 超时=%s秒",
             target_id, timeout
         )
-        raise
+        raise  # 直接重新抛出，不包装
     
     except Exception as e:
         error_msg = f"流式执行 URL 验证任务失败: {e}"
@@ -483,12 +665,5 @@ def run_and_stream_save_urls_task(
         raise RuntimeError(error_msg) from e
     
     finally:
-        # 清理资源
-        if data_generator is not None:
-            try:
-                # 确保生成器被正确关闭
-                data_generator.close()
-            except (GeneratorExit, StopIteration):
-                pass
-            except Exception as e:
-                logger.warning("关闭数据生成器时出错: %s", e)
+        # 5. 清理资源
+        _cleanup_resources(data_generator)

backend/apps/scan/urls.py

@@ -1,6 +1,6 @@
 from django.urls import path, include
 from rest_framework.routers import DefaultRouter
-from .views import ScanViewSet, ScheduledScanViewSet
+from .views import ScanViewSet, ScheduledScanViewSet, ScanLogListView
 from .notifications.views import notification_callback
 from apps.asset.views import (
     SubdomainSnapshotViewSet, WebsiteSnapshotViewSet, DirectorySnapshotViewSet,
@@ -31,6 +31,8 @@ urlpatterns = [
     path('', include(router.urls)),
     # Worker 回调 API
     path('callbacks/notification/', notification_callback, name='notification-callback'),
+    # 扫描日志 API
+    path('scans/<int:scan_id>/logs/', ScanLogListView.as_view(), name='scan-logs-list'),
     # 嵌套路由：/api/scans/{scan_pk}/xxx/
     path('scans/<int:scan_pk>/subdomains/', scan_subdomains_list, name='scan-subdomains-list'),
     path('scans/<int:scan_pk>/subdomains/export/', scan_subdomains_export, name='scan-subdomains-export'),

backend/apps/scan/utils/__init__.py

@@ -11,6 +11,7 @@ from .wordlist_helpers import ensure_wordlist_local
 from .nuclei_helpers import ensure_nuclei_templates_local
 from .performance import FlowPerformanceTracker, CommandPerformanceTracker
 from .workspace_utils import setup_scan_workspace, setup_scan_directory
+from .user_logger import user_log
 from . import config_parser
 
 __all__ = [
@@ -31,6 +32,8 @@ __all__ = [
     # 性能监控
     'FlowPerformanceTracker',      # Flow 性能追踪器（含系统资源采样）
     'CommandPerformanceTracker',   # 命令性能追踪器
+    # 扫描日志
+    'user_log',              # 用户可见扫描日志记录
     # 配置解析
     'config_parser',
 ]

backend/apps/scan/utils/command_executor.py

@@ -48,7 +48,7 @@ ENABLE_COMMAND_LOGGING = getattr(settings, 'ENABLE_COMMAND_LOGGING', True)
 # 动态并发控制阈值（可在 Django settings 中覆盖）
 SCAN_CPU_HIGH = getattr(settings, 'SCAN_CPU_HIGH', 90.0)   # CPU 高水位（百分比）
 SCAN_MEM_HIGH = getattr(settings, 'SCAN_MEM_HIGH', 80.0)   # 内存高水位（百分比）
-SCAN_LOAD_CHECK_INTERVAL = getattr(settings, 'SCAN_LOAD_CHECK_INTERVAL', 30)  # 负载检查间隔（秒）
+SCAN_LOAD_CHECK_INTERVAL = getattr(settings, 'SCAN_LOAD_CHECK_INTERVAL', 180)  # 负载检查间隔（秒）
 SCAN_COMMAND_STARTUP_DELAY = getattr(settings, 'SCAN_COMMAND_STARTUP_DELAY', 5)  # 命令启动前等待（秒）
 
 _ACTIVE_COMMANDS = 0
@@ -74,7 +74,7 @@ def _wait_for_system_load() -> None:
             return
 
         logger.info(
-            "系统负载较高，暂缓启动: cpu=%.1f%% (阈值 %.1f%%), mem=%.1f%% (阈值 %.1f%%)",
+            "系统负载较高，任务将排队执行，防止oom: cpu=%.1f%% (阈值 %.1f%%), mem=%.1f%% (阈值 %.1f%%)",
             cpu,
             SCAN_CPU_HIGH,
             mem,

backend/apps/scan/utils/config_merger.py

@@ -0,0 +1,80 @@
+"""
+配置合并工具模块
+
+提供多引擎 YAML 配置的冲突检测和合并功能。
+"""
+
+from typing import List, Tuple
+
+import yaml
+
+
+class ConfigConflictError(Exception):
+    """配置冲突异常
+    
+    当两个或多个引擎定义相同的顶层扫描类型键时抛出。
+    """
+    
+    def __init__(self, conflicts: List[Tuple[str, str, str]]):
+        """
+        参数:
+            conflicts: (键, 引擎1名称, 引擎2名称) 元组列表
+        """
+        self.conflicts = conflicts
+        msg = "; ".join([f"{k} 同时存在于「{e1}」和「{e2}」" for k, e1, e2 in conflicts])
+        super().__init__(f"扫描类型冲突: {msg}")
+
+
+def merge_engine_configs(engines: List[Tuple[str, str]]) -> str:
+    """
+    合并多个引擎的 YAML 配置。
+    
+    参数:
+        engines: (引擎名称, 配置YAML) 元组列表
+    
+    返回:
+        合并后的 YAML 字符串
+    
+    异常:
+        ConfigConflictError: 当顶层键冲突时
+    """
+    if not engines:
+        return ""
+    
+    if len(engines) == 1:
+        return engines[0][1]
+    
+    # 追踪每个顶层键属于哪个引擎
+    key_to_engine: dict[str, str] = {}
+    conflicts: List[Tuple[str, str, str]] = []
+    
+    for engine_name, config_yaml in engines:
+        if not config_yaml or not config_yaml.strip():
+            continue
+            
+        try:
+            parsed = yaml.safe_load(config_yaml)
+        except yaml.YAMLError:
+            # 无效 YAML 跳过
+            continue
+            
+        if not isinstance(parsed, dict):
+            continue
+            
+        # 检查顶层键冲突
+        for key in parsed.keys():
+            if key in key_to_engine:
+                conflicts.append((key, key_to_engine[key], engine_name))
+            else:
+                key_to_engine[key] = engine_name
+    
+    if conflicts:
+        raise ConfigConflictError(conflicts)
+    
+    # 无冲突，用双换行符连接配置
+    configs = []
+    for _, config_yaml in engines:
+        if config_yaml and config_yaml.strip():
+            configs.append(config_yaml.strip())
+    
+    return "\n\n".join(configs)

backend/apps/scan/utils/user_logger.py

@@ -0,0 +1,56 @@
+"""
+扫描日志记录器
+
+提供统一的日志记录接口，用于在 Flow 中记录用户可见的扫描进度日志。
+
+特性：
+- 简单的函数式 API
+- 只写入数据库（ScanLog 表），不写 Python logging
+- 错误容忍（数据库失败不影响扫描执行）
+
+职责分离：
+- user_log: 用户可见日志（写数据库，前端展示）
+- logger: 开发者日志（写日志文件/控制台，调试用）
+
+使用示例：
+    from apps.scan.utils import user_log
+    
+    # 用户日志（写数据库）
+    user_log(scan_id, "port_scan", "Starting port scan")
+    user_log(scan_id, "port_scan", "naabu completed: found 120 ports")
+    
+    # 开发者日志（写日志文件）
+    logger.info("✓ 工具 %s 执行完成 - 记录数: %d", tool_name, count)
+"""
+
+import logging
+from django.db import DatabaseError
+
+logger = logging.getLogger(__name__)
+
+
+def user_log(scan_id: int, stage: str, message: str, level: str = "info"):
+    """
+    记录用户可见的扫描日志（只写数据库）
+    
+    Args:
+        scan_id: 扫描任务 ID
+        stage: 阶段名称，如 "port_scan", "site_scan"
+        message: 日志消息
+        level: 日志级别，默认 "info"，可选 "warning", "error"
+        
+    数据库 content 格式: "[{stage}] {message}"
+    """
+    formatted = f"[{stage}] {message}"
+    
+    try:
+        from apps.scan.models import ScanLog
+        ScanLog.objects.create(
+            scan_id=scan_id,
+            level=level,
+            content=formatted
+        )
+    except DatabaseError as e:
+        logger.error("ScanLog write failed - scan_id=%s, error=%s", scan_id, e)
+    except Exception as e:
+        logger.error("ScanLog write unexpected error - scan_id=%s, error=%s", scan_id, e)

backend/apps/scan/utils/wordlist_helpers.py

@@ -96,7 +96,13 @@ def ensure_wordlist_local(wordlist_name: str) -> str:
         ssl_context.check_hostname = False
         ssl_context.verify_mode = ssl.CERT_NONE
         
-        with urllib_request.urlopen(download_url, context=ssl_context) as resp:
+        # 创建带 API Key 的请求
+        req = urllib_request.Request(download_url)
+        worker_api_key = os.getenv('WORKER_API_KEY', '')
+        if worker_api_key:
+            req.add_header('X-Worker-API-Key', worker_api_key)
+        
+        with urllib_request.urlopen(req, context=ssl_context) as resp:
             if resp.status != 200:
                 raise RuntimeError(f"下载字典失败，HTTP {resp.status}")
             data = resp.read()

backend/apps/scan/views/__init__.py

@@ -2,8 +2,10 @@
 
 from .scan_views import ScanViewSet
 from .scheduled_scan_views import ScheduledScanViewSet
+from .scan_log_views import ScanLogListView
 
 __all__ = [
     'ScanViewSet',
     'ScheduledScanViewSet',
+    'ScanLogListView',
 ]

backend/apps/scan/views/scan_log_views.py

@@ -0,0 +1,56 @@
+"""
+扫描日志 API
+
+提供扫描日志查询接口，支持游标分页用于增量轮询。
+"""
+
+from rest_framework.views import APIView
+from rest_framework.response import Response
+
+from apps.scan.models import ScanLog
+from apps.scan.serializers import ScanLogSerializer
+
+
+class ScanLogListView(APIView):
+    """
+    GET /scans/{scan_id}/logs/
+    
+    游标分页 API，用于增量查询日志
+    
+    查询参数：
+    - afterId: 只返回此 ID 之后的日志（用于增量轮询，避免时间戳重复导致的重复日志）
+    - limit: 返回数量限制（默认 200，最大 1000）
+    
+    返回：
+    - results: 日志列表
+    - hasMore: 是否还有更多日志
+    """
+    
+    def get(self, request, scan_id: int):
+        # 参数解析
+        after_id = request.query_params.get('afterId')
+        try:
+            limit = min(int(request.query_params.get('limit', 200)), 1000)
+        except (ValueError, TypeError):
+            limit = 200
+        
+        # 查询日志（按 ID 排序，ID 是自增的，保证顺序一致）
+        queryset = ScanLog.objects.filter(scan_id=scan_id).order_by('id')
+        
+        # 游标过滤（使用 ID 而非时间戳，避免同一时间戳多条日志导致重复）
+        if after_id:
+            try:
+                queryset = queryset.filter(id__gt=int(after_id))
+            except (ValueError, TypeError):
+                pass
+        
+        # 限制返回数量（多取一条用于判断 hasMore）
+        logs = list(queryset[:limit + 1])
+        has_more = len(logs) > limit
+        if has_more:
+            logs = logs[:limit]
+        
+        return Response({
+            'results': ScanLogSerializer(logs, many=True).data,
+            'hasMore': has_more,
+        })

backend/apps/scan/views/scan_views.py

@@ -9,13 +9,14 @@ import logging
 
 from apps.common.response_helpers import success_response, error_response
 from apps.common.error_codes import ErrorCodes
+from apps.scan.utils.config_merger import ConfigConflictError
 
 logger = logging.getLogger(__name__)
 
 from ..models import Scan, ScheduledScan
 from ..serializers import (
     ScanSerializer, ScanHistorySerializer, QuickScanSerializer,
-    ScheduledScanSerializer, CreateScheduledScanSerializer,
+    InitiateScanSerializer, ScheduledScanSerializer, CreateScheduledScanSerializer,
     UpdateScheduledScanSerializer, ToggleScheduledScanSerializer
 )
 from ..services.scan_service import ScanService
@@ -110,7 +111,7 @@ class ScanViewSet(viewsets.ModelViewSet):
         快速扫描接口
         
         功能：
-        1. 接收目标列表和引擎配置
+        1. 接收目标列表和 YAML 配置
         2. 自动解析输入（支持 URL、域名、IP、CIDR）
         3. 批量创建 Target、Website、Endpoint 资产
         4. 立即发起批量扫描
@@ -118,7 +119,9 @@ class ScanViewSet(viewsets.ModelViewSet):
         请求参数：
         {
             "targets": [{"name": "example.com"}, {"name": "https://example.com/api"}],
-            "engine_id": 1
+            "configuration": "subdomain_discovery:\n  enabled: true\n  ...",
+            "engine_ids": [1, 2],  // 可选，用于记录
+            "engine_names": ["引擎A", "引擎B"]  // 可选，用于记录
         }
         
         支持的输入格式：
@@ -133,7 +136,9 @@ class ScanViewSet(viewsets.ModelViewSet):
         serializer.is_valid(raise_exception=True)
         
         targets_data = serializer.validated_data['targets']
-        engine_id = serializer.validated_data.get('engine_id')
+        configuration = serializer.validated_data['configuration']
+        engine_ids = serializer.validated_data.get('engine_ids', [])
+        engine_names = serializer.validated_data.get('engine_names', [])
         
         try:
             # 提取输入字符串列表
@@ -141,7 +146,7 @@ class ScanViewSet(viewsets.ModelViewSet):
             
             # 1. 使用 QuickScanService 解析输入并创建资产
             quick_scan_service = QuickScanService()
-            result = quick_scan_service.process_quick_scan(inputs, engine_id)
+            result = quick_scan_service.process_quick_scan(inputs, engine_ids[0] if engine_ids else None)
             
             targets = result['targets']
             
@@ -153,17 +158,13 @@ class ScanViewSet(viewsets.ModelViewSet):
                     status_code=status.HTTP_400_BAD_REQUEST
                 )
             
-            # 2. 获取扫描引擎
-            engine_service = EngineService()
-            engine = engine_service.get_engine(engine_id)
-            if not engine:
-                raise ValidationError(f'扫描引擎 ID {engine_id} 不存在')
-            
-            # 3. 批量发起扫描
+            # 2. 直接使用前端传递的配置创建扫描
             scan_service = ScanService()
             created_scans = scan_service.create_scans(
                 targets=targets,
-                engine=engine
+                engine_ids=engine_ids,
+                engine_names=engine_names,
+                yaml_configuration=configuration
             )
             
             # 检查是否成功创建扫描任务
@@ -214,31 +215,53 @@ class ScanViewSet(viewsets.ModelViewSet):
         请求参数:
         - organization_id: 组织ID (int, 可选)
         - target_id: 目标ID (int, 可选)
-        - engine_id: 扫描引擎ID (int, 必填)
+        - configuration: YAML 配置字符串 (str, 必填)
+        - engine_ids: 扫描引擎ID列表 (list[int], 必填)
+        - engine_names: 引擎名称列表 (list[str], 必填)
         
         注意: organization_id 和 target_id 二选一
         
         返回:
         - 扫描任务详情（单个或多个）
         """
-        # 获取请求数据
-        organization_id = request.data.get('organization_id')
-        target_id = request.data.get('target_id')
-        engine_id = request.data.get('engine_id')
+        # 使用 serializer 验证请求数据
+        serializer = InitiateScanSerializer(data=request.data)
+        serializer.is_valid(raise_exception=True)
+        
+        # 获取验证后的数据
+        organization_id = serializer.validated_data.get('organization_id')
+        target_id = serializer.validated_data.get('target_id')
+        configuration = serializer.validated_data['configuration']
+        engine_ids = serializer.validated_data['engine_ids']
+        engine_names = serializer.validated_data['engine_names']
         
         try:
-            # 步骤1：准备扫描所需的数据（验证参数、查询资源、返回目标列表和引擎）
+            # 获取目标列表
             scan_service = ScanService()
-            targets, engine = scan_service.prepare_initiate_scan(
-                organization_id=organization_id,
-                target_id=target_id,
-                engine_id=engine_id
-            )
             
-            # 步骤2：批量创建扫描记录并分发扫描任务
+            if organization_id:
+                from apps.targets.repositories import DjangoOrganizationRepository
+                org_repo = DjangoOrganizationRepository()
+                organization = org_repo.get_by_id(organization_id)
+                if not organization:
+                    raise ObjectDoesNotExist(f'Organization ID {organization_id} 不存在')
+                targets = org_repo.get_targets(organization_id)
+                if not targets:
+                    raise ValidationError(f'组织 ID {organization_id} 下没有目标')
+            else:
+                from apps.targets.repositories import DjangoTargetRepository
+                target_repo = DjangoTargetRepository()
+                target = target_repo.get_by_id(target_id)
+                if not target:
+                    raise ObjectDoesNotExist(f'Target ID {target_id} 不存在')
+                targets = [target]
+            
+            # 直接使用前端传递的配置创建扫描
             created_scans = scan_service.create_scans(
                 targets=targets,
-                engine=engine
+                engine_ids=engine_ids,
+                engine_names=engine_names,
+                yaml_configuration=configuration
             )
             
             # 检查是否成功创建扫描任务

backend/apps/scan/views/scheduled_scan_views.py

@@ -17,6 +17,7 @@ from ..serializers import (
 )
 from ..services.scheduled_scan_service import ScheduledScanService
 from ..repositories import ScheduledScanDTO
+from ..utils.config_merger import ConfigConflictError
 from apps.common.pagination import BasePagination
 from apps.common.response_helpers import success_response, error_response
 from apps.common.error_codes import ErrorCodes
@@ -67,14 +68,16 @@ class ScheduledScanViewSet(viewsets.ModelViewSet):
             data = serializer.validated_data
             dto = ScheduledScanDTO(
                 name=data['name'],
-                engine_id=data['engine_id'],
+                engine_ids=data.get('engine_ids', []),
+                engine_names=data.get('engine_names', []),
+                yaml_configuration=data['configuration'],
                 organization_id=data.get('organization_id'),
                 target_id=data.get('target_id'),
                 cron_expression=data.get('cron_expression', '0 2 * * *'),
                 is_enabled=data.get('is_enabled', True),
             )
             
-            scheduled_scan = self.service.create(dto)
+            scheduled_scan = self.service.create_with_configuration(dto)
             response_serializer = ScheduledScanSerializer(scheduled_scan)
             
             return success_response(
@@ -98,7 +101,7 @@ class ScheduledScanViewSet(viewsets.ModelViewSet):
             data = serializer.validated_data
             dto = ScheduledScanDTO(
                 name=data.get('name'),
-                engine_id=data.get('engine_id'),
+                engine_ids=data.get('engine_ids'),
                 organization_id=data.get('organization_id'),
                 target_id=data.get('target_id'),
                 cron_expression=data.get('cron_expression'),
@@ -109,6 +112,16 @@ class ScheduledScanViewSet(viewsets.ModelViewSet):
             response_serializer = ScheduledScanSerializer(scheduled_scan)
             
             return success_response(data=response_serializer.data)
+        except ConfigConflictError as e:
+            return error_response(
+                code='CONFIG_CONFLICT',
+                message=str(e),
+                details=[
+                    {'key': k, 'engines': [e1, e2]} 
+                    for k, e1, e2 in e.conflicts
+                ],
+                status_code=status.HTTP_400_BAD_REQUEST
+            )
         except ValidationError as e:
             return error_response(
                 code=ErrorCodes.VALIDATION_ERROR,

backend/apps/targets/migrations/0001_initial.py

@@ -0,0 +1,52 @@
+# Generated by Django 5.2.7 on 2026-01-02 04:45
+
+from django.db import migrations, models
+
+
+class Migration(migrations.Migration):
+
+    initial = True
+
+    dependencies = [
+    ]
+
+    operations = [
+        migrations.CreateModel(
+            name='Target',
+            fields=[
+                ('id', models.AutoField(primary_key=True, serialize=False)),
+                ('name', models.CharField(blank=True, default='', help_text='目标标识（域名/IP/CIDR）', max_length=300)),
+                ('type', models.CharField(choices=[('domain', '域名'), ('ip', 'IP地址'), ('cidr', 'CIDR范围')], db_index=True, default='domain', help_text='目标类型', max_length=20)),
+                ('created_at', models.DateTimeField(auto_now_add=True, help_text='创建时间')),
+                ('last_scanned_at', models.DateTimeField(blank=True, help_text='最后扫描时间', null=True)),
+                ('deleted_at', models.DateTimeField(blank=True, db_index=True, help_text='删除时间（NULL表示未删除）', null=True)),
+            ],
+            options={
+                'verbose_name': '扫描目标',
+                'verbose_name_plural': '扫描目标',
+                'db_table': 'target',
+                'ordering': ['-created_at'],
+                'indexes': [models.Index(fields=['type'], name='target_type_36a73c_idx'), models.Index(fields=['-created_at'], name='target_created_67f489_idx'), models.Index(fields=['deleted_at', '-created_at'], name='target_deleted_9fc9da_idx'), models.Index(fields=['deleted_at', 'type'], name='target_deleted_306a89_idx'), models.Index(fields=['name'], name='target_name_f1c641_idx')],
+                'constraints': [models.UniqueConstraint(condition=models.Q(('deleted_at__isnull', True)), fields=('name',), name='unique_target_name_active')],
+            },
+        ),
+        migrations.CreateModel(
+            name='Organization',
+            fields=[
+                ('id', models.AutoField(primary_key=True, serialize=False)),
+                ('name', models.CharField(blank=True, default='', help_text='组织名称', max_length=300)),
+                ('description', models.CharField(blank=True, default='', help_text='组织描述', max_length=1000)),
+                ('created_at', models.DateTimeField(auto_now_add=True, help_text='创建时间')),
+                ('deleted_at', models.DateTimeField(blank=True, db_index=True, help_text='删除时间（NULL表示未删除）', null=True)),
+                ('targets', models.ManyToManyField(blank=True, help_text='所属目标列表', related_name='organizations', to='targets.target')),
+            ],
+            options={
+                'verbose_name': '组织',
+                'verbose_name_plural': '组织',
+                'db_table': 'organization',
+                'ordering': ['-created_at'],
+                'indexes': [models.Index(fields=['-created_at'], name='organizatio_created_012eac_idx'), models.Index(fields=['deleted_at', '-created_at'], name='organizatio_deleted_2c604f_idx'), models.Index(fields=['name'], name='organizatio_name_bcc2ee_idx')],
+                'constraints': [models.UniqueConstraint(condition=models.Q(('deleted_at__isnull', True)), fields=('name',), name='unique_organization_name_active')],
+            },
+        ),
+    ]

backend/config/settings.py

@@ -177,6 +177,10 @@ STATIC_URL = 'static/'
 
 DEFAULT_AUTO_FIELD = 'django.db.models.BigAutoField'
 
+# ==================== Worker API Key 配置 ====================
+# Worker 节点认证密钥（从环境变量读取）
+WORKER_API_KEY = os.environ.get('WORKER_API_KEY', '')
+
 # ==================== REST Framework 配置 ====================
 REST_FRAMEWORK = {
     'DEFAULT_PAGINATION_CLASS': 'apps.common.pagination.BasePagination',  # 使用基础分页器
@@ -186,6 +190,14 @@ REST_FRAMEWORK = {
         'apps.common.authentication.CsrfExemptSessionAuthentication',
     ],
     
+    # 全局权限配置：默认需要认证，公开端点和 Worker 端点在权限类中单独处理
+    'DEFAULT_PERMISSION_CLASSES': [
+        'apps.common.permissions.IsAuthenticatedOrPublic',
+    ],
+    
+    # 自定义异常处理器：统一 401/403 错误响应格式
+    'EXCEPTION_HANDLER': 'apps.common.exception_handlers.custom_exception_handler',
+    
     # JSON 命名格式转换：后端 snake_case ↔ 前端 camelCase
     'DEFAULT_RENDERER_CLASSES': (
         'djangorestframework_camel_case.render.CamelCaseJSONRenderer',  # 响应数据转换为 camelCase
@@ -207,6 +219,8 @@ REST_FRAMEWORK = {
 # 允许所有来源（前后端分离项目，安全性由认证系统保障）
 CORS_ALLOW_ALL_ORIGINS = os.getenv('CORS_ALLOW_ALL_ORIGINS', 'True').lower() == 'true'
 CORS_ALLOW_CREDENTIALS = True
+# 暴露额外的响应头给前端（Content-Disposition 用于文件下载获取文件名）
+CORS_EXPOSE_HEADERS = ['Content-Disposition']
 
 # ==================== CSRF 配置 ====================
 CSRF_TRUSTED_ORIGINS = os.getenv('CSRF_TRUSTED_ORIGINS', 'http://localhost:3000,http://127.0.0.1:3000').split(',')
@@ -355,25 +369,16 @@ HOST_WORDLISTS_DIR = '/opt/xingrin/wordlists'
 # ============================================
 # Worker 配置中心（任务容器从 /api/workers/config/ 获取）
 # ============================================
-# Worker 数据库/Redis 地址由 worker_views.py 的 config API 动态返回
+# Worker 数据库地址由 worker_views.py 的 config API 动态返回
 # 根据请求来源（本地/远程）返回不同的配置：
-# - 本地 Worker（Docker 网络内）：使用内部服务名（postgres, redis）
+# - 本地 Worker（Docker 网络内）：使用内部服务名 postgres
 # - 远程 Worker（公网访问）：使用 PUBLIC_HOST
 # 
-# 以下变量仅作为备用/兼容配置，实际配置由 API 动态生成
+# 注意：Redis 仅在 Server 容器内使用，Worker 不需要直接连接 Redis
 _db_host = DATABASES['default']['HOST']
 _is_internal_db = _db_host in ('postgres', 'localhost', '127.0.0.1')
 WORKER_DB_HOST = os.getenv('WORKER_DB_HOST', _db_host)
 
-# 远程 Worker 访问 Redis 的地址（自动推导）
-# - 如果 PUBLIC_HOST 是外部 IP → 使用 PUBLIC_HOST
-# - 如果 PUBLIC_HOST 是 Docker 内部名 → 使用 redis（本地部署）
-_is_internal_public = PUBLIC_HOST in ('server', 'localhost', '127.0.0.1')
-WORKER_REDIS_URL = os.getenv(
-    'WORKER_REDIS_URL',
-    'redis://redis:6379/0' if _is_internal_public else f'redis://{PUBLIC_HOST}:6379/0'
-)
-
 # 容器内挂载目标路径（统一使用 /opt/xingrin）
 CONTAINER_RESULTS_MOUNT = '/opt/xingrin/results'
 CONTAINER_LOGS_MOUNT = '/opt/xingrin/logs'

backend/config/urls.py

@@ -16,7 +16,6 @@ Including another URLconf
 """
 from django.contrib import admin
 from django.urls import path, include
-from rest_framework import permissions
 from drf_yasg.views import get_schema_view
 from drf_yasg import openapi
 
@@ -30,7 +29,6 @@ schema_view = get_schema_view(
       description="Web 应用侦察工具 API 文档",
    ),
    public=True,
-   permission_classes=(permissions.AllowAny,),
 )
 
 urlpatterns = [

backend/requirements.txt

@@ -41,6 +41,7 @@ python-dateutil==2.9.0
 pytz==2024.1
 validators==0.22.0
 PyYAML==6.0.1
+ruamel.yaml>=0.18.0  # 保留注释的 YAML 解析
 colorlog==6.8.2  # 彩色日志输出
 python-json-logger==2.0.7  # JSON 结构化日志
 Jinja2>=3.1.6  # 命令模板引擎

backend/scripts/generate_test_data_sql.py

@@ -180,6 +180,28 @@ def get_db_config() -> dict:
     }
 
 
+def generate_raw_response_headers(headers_dict: dict) -> str:
+    """
+    将响应头字典转换为原始 HTTP 响应头字符串格式
+    
+    Args:
+        headers_dict: 响应头字典
+    
+    Returns:
+        原始 HTTP 响应头字符串，格式如：
+        HTTP/1.1 200 OK
+        Server: nginx
+        Content-Type: text/html
+        ...
+    """
+    lines = ['HTTP/1.1 200 OK']
+    for key, value in headers_dict.items():
+        # 将下划线转换为连字符，并首字母大写
+        header_name = key.replace('_', '-').title()
+        lines.append(f'{header_name}: {value}')
+    return '\r\n'.join(lines)
+
+
 DB_CONFIG = get_db_config()
 
 
@@ -238,6 +260,12 @@ class TestDataGenerator:
     def clear_data(self):
         """清除所有测试数据"""
         cur = self.conn.cursor()
+        
+        # 先删除 IMMV（避免 pg_ivm 的 anyarray bug）
+        print("  删除 IMMV...")
+        cur.execute("DROP TABLE IF EXISTS asset_search_view CASCADE")
+        self.conn.commit()
+        
         tables = [
             # 指纹表
             'ehole_fingerprint', 'goby_fingerprint', 'wappalyzer_fingerprint',
@@ -254,6 +282,26 @@ class TestDataGenerator:
         for table in tables:
             cur.execute(f"DELETE FROM {table}")
         self.conn.commit()
+        
+        # 重建 IMMV
+        print("  重建 IMMV...")
+        cur.execute("""
+            SELECT pgivm.create_immv('asset_search_view', $$
+                SELECT 
+                    w.id,
+                    w.url,
+                    w.host,
+                    w.title,
+                    w.tech,
+                    w.status_code,
+                    w.response_headers,
+                    w.response_body,
+                    w.created_at,
+                    w.target_id
+                FROM website w
+            $$)
+        """)
+        self.conn.commit()
         print("  ✓ 数据清除完成\n")
 
     def create_workers(self) -> list:
@@ -548,6 +596,10 @@ class TestDataGenerator:
             'Authentication failed for protected resources.',
         ]
         
+        # 获取引擎名称映射
+        cur.execute("SELECT id, name FROM scan_engine WHERE id = ANY(%s)", (engine_ids,))
+        engine_name_map = {row[0]: row[1] for row in cur.fetchall()}
+        
         ids = []
         # 随机选择目标数量 - 增加到 80-120 个
         num_targets = min(random.randint(80, 120), len(target_ids))
@@ -558,7 +610,10 @@ class TestDataGenerator:
             num_scans = random.randint(3, 15)
             for _ in range(num_scans):
                 status = random.choices(statuses, weights=status_weights)[0]
-                engine_id = random.choice(engine_ids)
+                # 随机选择 1-3 个引擎
+                num_engines = random.randint(1, min(3, len(engine_ids)))
+                selected_engine_ids = random.sample(engine_ids, num_engines)
+                selected_engine_names = [engine_name_map.get(eid, f'Engine-{eid}') for eid in selected_engine_ids]
                 worker_id = random.choice(worker_ids) if worker_ids else None
                 
                 progress = random.randint(10, 95) if status == 'running' else (100 if status == 'completed' else random.randint(0, 50))
@@ -581,20 +636,20 @@ class TestDataGenerator:
                 
                 cur.execute("""
                     INSERT INTO scan (
-                        target_id, engine_id, status, worker_id, progress, current_stage,
+                        target_id, engine_ids, engine_names, yaml_configuration, status, worker_id, progress, current_stage,
                         results_dir, error_message, container_ids, stage_progress,
                         cached_subdomains_count, cached_websites_count, cached_endpoints_count,
                         cached_ips_count, cached_directories_count, cached_vulns_total,
                         cached_vulns_critical, cached_vulns_high, cached_vulns_medium, cached_vulns_low,
                         created_at, stopped_at, deleted_at
                     ) VALUES (
-                        %s, %s, %s, %s, %s, %s, %s, %s, %s, %s,
+                        %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s,
                         %s, %s, %s, %s, %s, %s, %s, %s, %s, %s,
                         NOW() - INTERVAL '%s days', %s, NULL
                     )
                     RETURNING id
                 """, (
-                    target_id, engine_id, status, worker_id, progress, stage,
+                    target_id, selected_engine_ids, json.dumps(selected_engine_names), '', status, worker_id, progress, stage,
                     f'/app/results/scan_{target_id}_{random.randint(1000, 9999)}', error_msg, '{}', '{}',
                     subdomains, websites, endpoints, ips, directories, vulns_total,
                     vulns_critical, vulns_high, vulns_medium, vulns_low,
@@ -651,6 +706,10 @@ class TestDataGenerator:
         num_schedules = random.randint(40, 50)
         selected = random.sample(schedule_templates, min(num_schedules, len(schedule_templates)))
         
+        # 获取引擎名称映射
+        cur.execute("SELECT id, name FROM scan_engine WHERE id = ANY(%s)", (engine_ids,))
+        engine_name_map = {row[0]: row[1] for row in cur.fetchall()}
+        
         count = 0
         for name_base, cron_template in selected:
             name = f'{name_base}-{suffix}-{count:02d}'
@@ -662,7 +721,11 @@ class TestDataGenerator:
             )
             enabled = random.random() > 0.3  # 70% 启用
             
-            engine_id = random.choice(engine_ids)
+            # 随机选择 1-3 个引擎
+            num_engines = random.randint(1, min(3, len(engine_ids)))
+            selected_engine_ids = random.sample(engine_ids, num_engines)
+            selected_engine_names = [engine_name_map.get(eid, f'Engine-{eid}') for eid in selected_engine_ids]
+            
             # 随机决定关联组织还是目标
             if org_ids and target_ids:
                 if random.random() > 0.5:
@@ -686,12 +749,12 @@ class TestDataGenerator:
             
             cur.execute("""
                 INSERT INTO scheduled_scan (
-                    name, engine_id, organization_id, target_id, cron_expression, is_enabled,
+                    name, engine_ids, engine_names, yaml_configuration, organization_id, target_id, cron_expression, is_enabled,
                     run_count, last_run_time, next_run_time, created_at, updated_at
-                ) VALUES (%s, %s, %s, %s, %s, %s, %s, %s, %s, NOW() - INTERVAL '%s days', NOW())
+                ) VALUES (%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, NOW() - INTERVAL '%s days', NOW())
                 ON CONFLICT DO NOTHING
             """, (
-                name, engine_id, org_id, target_id, cron, enabled,
+                name, selected_engine_ids, json.dumps(selected_engine_names), '', org_id, target_id, cron, enabled,
                 run_count if has_run else 0,
                 datetime.now() - timedelta(days=random.randint(0, 14), hours=random.randint(0, 23)) if has_run else None,
                 datetime.now() + timedelta(hours=random.randint(1, 336))  # 最多 2 周后
@@ -812,7 +875,7 @@ class TestDataGenerator:
         ]
         
         # 真实的 body preview 内容
-        body_previews = [
+        response_bodies = [
             '<!DOCTYPE html><html lang="en"><head><meta charset="UTF-8"><meta name="viewport" content="width=device-width, initial-scale=1.0"><title>Login - Enterprise Portal</title><link rel="stylesheet" href="/assets/css/main.css"></head><body><div id="app"></div><script src="/assets/js/bundle.js"></script></body></html>',
             '<!DOCTYPE html><html><head><title>Dashboard</title><meta name="description" content="Enterprise management dashboard for monitoring and analytics"><link rel="icon" href="/favicon.ico"></head><body><noscript>You need to enable JavaScript to run this app.</noscript><div id="root"></div></body></html>',
             '{"status":"ok","version":"2.4.1","environment":"production","timestamp":"2024-12-22T10:30:00Z","services":{"database":"healthy","cache":"healthy","queue":"healthy"},"uptime":864000}',
@@ -843,14 +906,27 @@ class TestDataGenerator:
                 # 生成固定 245 长度的 URL
                 url = generate_fixed_length_url(target_name, length=245, path_hint=f'website/{i:04d}')
                 
+                # 生成模拟的响应头数据
+                response_headers = {
+                    'server': random.choice(['nginx', 'Apache', 'cloudflare', 'Microsoft-IIS/10.0']),
+                    'content_type': 'text/html; charset=utf-8',
+                    'x_powered_by': random.choice(['PHP/8.2', 'ASP.NET', 'Express', None]),
+                    'x_frame_options': random.choice(['DENY', 'SAMEORIGIN', None]),
+                    'strict_transport_security': 'max-age=31536000; includeSubDomains' if random.choice([True, False]) else None,
+                    'set_cookie': f'session={random.randint(100000, 999999)}; HttpOnly; Secure' if random.choice([True, False]) else None,
+                }
+                # 移除 None 值
+                response_headers = {k: v for k, v in response_headers.items() if v is not None}
+                
                 batch_data.append((
                     url, target_id, target_name, random.choice(titles),
                     random.choice(webservers), random.choice(tech_stacks),
                     random.choice([200, 301, 302, 403, 404]),
                     random.randint(1000, 500000), 'text/html; charset=utf-8',
                     f'https://{target_name}/login' if random.choice([True, False]) else '',
-                    random.choice(body_previews),
-                    random.choice([True, False, None])
+                    random.choice(response_bodies),
+                    random.choice([True, False, None]),
+                    generate_raw_response_headers(response_headers)
                 ))
         
         # 批量插入
@@ -859,12 +935,12 @@ class TestDataGenerator:
             execute_values(cur, """
                 INSERT INTO website (
                     url, target_id, host, title, webserver, tech, status_code,
-                    content_length, content_type, location, body_preview, vhost,
-                    created_at
+                    content_length, content_type, location, response_body, vhost,
+                    response_headers, created_at
                 ) VALUES %s
                 ON CONFLICT DO NOTHING
                 RETURNING id
-            """, batch_data, template="(%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, NOW())")
+            """, batch_data, template="(%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, NOW())")
             ids = [row[0] for row in cur.fetchall()]
                     
         print(f"  ✓ 创建了 {len(batch_data)} 个网站\n")
@@ -965,7 +1041,7 @@ class TestDataGenerator:
         ]
         
         # 真实的 API 响应 body preview
-        body_previews = [
+        response_bodies = [
             '{"status":"success","data":{"user_id":12345,"username":"john_doe","email":"john@example.com","role":"user","created_at":"2024-01-15T10:30:00Z","last_login":"2024-12-22T08:45:00Z"}}',
             '{"success":true,"message":"Authentication successful","token":"eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJzdWIiOiIxMjM0NTY3ODkwIiwibmFtZSI6IkpvaG4gRG9lIiwiaWF0IjoxNTE2MjM5MDIyfQ.SflKxwRJSMeKKF2QT4fwpMeJf36POk6yJV_adQssw5c","expires_in":3600}',
             '{"error":"Unauthorized","code":"AUTH_FAILED","message":"Invalid credentials provided. Please check your username and password.","timestamp":"2024-12-22T15:30:45.123Z","request_id":"req_abc123xyz"}',
@@ -1017,14 +1093,27 @@ class TestDataGenerator:
                 # 生成 10-20 个 tags (gf_patterns)
                 tags = random.choice(gf_patterns)
                 
+                # 生成模拟的响应头数据
+                response_headers = {
+                    'server': random.choice(['nginx', 'gunicorn', 'uvicorn', 'Apache']),
+                    'content_type': 'application/json',
+                    'x_request_id': f'req_{random.randint(100000, 999999)}',
+                    'x_ratelimit_limit': str(random.choice([100, 1000, 5000])),
+                    'x_ratelimit_remaining': str(random.randint(0, 1000)),
+                    'cache_control': random.choice(['no-cache', 'max-age=3600', 'private', None]),
+                }
+                # 移除 None 值
+                response_headers = {k: v for k, v in response_headers.items() if v is not None}
+                
                 batch_data.append((
                     url, target_id, target_name, title,
                     random.choice(['nginx/1.24.0', 'gunicorn/21.2.0']),
                     random.choice([200, 201, 301, 400, 401, 403, 404, 500]),
                     random.randint(100, 50000), 'application/json',
                     tech_list,
-                    '', random.choice(body_previews),
-                    random.choice([True, False, None]), tags
+                    '', random.choice(response_bodies),
+                    random.choice([True, False, None]), tags,
+                    generate_raw_response_headers(response_headers)
                 ))
                 count += 1
         
@@ -1033,11 +1122,11 @@ class TestDataGenerator:
             execute_values(cur, """
                 INSERT INTO endpoint (
                     url, target_id, host, title, webserver, status_code, content_length,
-                    content_type, tech, location, body_preview, vhost, matched_gf_patterns,
-                    created_at
+                    content_type, tech, location, response_body, vhost, matched_gf_patterns,
+                    response_headers, created_at
                 ) VALUES %s
                 ON CONFLICT DO NOTHING
-            """, batch_data, template="(%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, NOW())")
+            """, batch_data, template="(%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, NOW())")
                 
         print(f"  ✓ 创建了 {count} 个端点\n")
 
@@ -1185,77 +1274,79 @@ class TestDataGenerator:
         print(f"  ✓ 创建了 {count} 个主机端口映射\n")
 
     def create_vulnerabilities(self, target_ids: list):
-        """创建漏洞"""
+        """创建漏洞（基于 website URL 前缀）"""
         print("🐛 创建漏洞...")
         cur = self.conn.cursor()
         
         vuln_types = [
-            'sql-injection-authentication-bypass-vulnerability-',  # 50 chars
-            'cross-site-scripting-xss-stored-persistent-attack-',  # 50 chars
-            'cross-site-request-forgery-csrf-token-validation--',  # 50 chars
-            'server-side-request-forgery-ssrf-internal-access--',  # 50 chars
-            'xml-external-entity-xxe-injection-vulnerability---',  # 50 chars
-            'remote-code-execution-rce-command-injection-flaw--',  # 50 chars
-            'local-file-inclusion-lfi-path-traversal-exploit---',  # 50 chars
-            'directory-traversal-arbitrary-file-read-access----',  # 50 chars
-            'authentication-bypass-session-management-flaw-----',  # 50 chars
-            'insecure-direct-object-reference-idor-access-ctrl-',  # 50 chars
-            'sensitive-data-exposure-information-disclosure----',  # 50 chars
-            'security-misconfiguration-default-credentials-----',  # 50 chars
-            'broken-access-control-privilege-escalation-vuln---',  # 50 chars
-            'cors-misconfiguration-cross-origin-data-leakage---',  # 50 chars
-            'subdomain-takeover-dns-misconfiguration-exploit---',  # 50 chars
-            'exposed-admin-panel-unauthorized-access-control---',  # 50 chars
-            'default-credentials-weak-authentication-bypass----',  # 50 chars
-            'information-disclosure-sensitive-data-exposure----',  # 50 chars
-            'command-injection-os-command-execution-exploit----',  # 50 chars
-            'ldap-injection-directory-service-manipulation-----',  # 50 chars
-            'xpath-injection-xml-query-manipulation-attack-----',  # 50 chars
-            'nosql-injection-mongodb-query-manipulation--------',  # 50 chars
-            'template-injection-ssti-server-side-execution-----',  # 50 chars
-            'deserialization-vulnerability-object-injection----',  # 50 chars
-            'jwt-vulnerability-token-forgery-authentication----',  # 50 chars
-            'open-redirect-url-redirection-phishing-attack-----',  # 50 chars
-            'http-request-smuggling-cache-poisoning-attack-----',  # 50 chars
-            'host-header-injection-password-reset-poisoning----',  # 50 chars
-            'clickjacking-ui-redressing-frame-injection--------',  # 50 chars
-            'session-fixation-authentication-session-attack----',  # 50 chars
+            'sql-injection-authentication-bypass-vulnerability-',
+            'cross-site-scripting-xss-stored-persistent-attack-',
+            'cross-site-request-forgery-csrf-token-validation--',
+            'server-side-request-forgery-ssrf-internal-access--',
+            'xml-external-entity-xxe-injection-vulnerability---',
+            'remote-code-execution-rce-command-injection-flaw--',
+            'local-file-inclusion-lfi-path-traversal-exploit---',
+            'directory-traversal-arbitrary-file-read-access----',
+            'authentication-bypass-session-management-flaw-----',
+            'insecure-direct-object-reference-idor-access-ctrl-',
+            'sensitive-data-exposure-information-disclosure----',
+            'security-misconfiguration-default-credentials-----',
+            'broken-access-control-privilege-escalation-vuln---',
+            'cors-misconfiguration-cross-origin-data-leakage---',
+            'subdomain-takeover-dns-misconfiguration-exploit---',
+            'exposed-admin-panel-unauthorized-access-control---',
+            'default-credentials-weak-authentication-bypass----',
+            'information-disclosure-sensitive-data-exposure----',
+            'command-injection-os-command-execution-exploit----',
+            'ldap-injection-directory-service-manipulation-----',
         ]
         
         sources = [
-            'nuclei-vulnerability-scanner--',  # 30 chars
-            'dalfox-xss-parameter-analysis-',  # 30 chars
-            'sqlmap-sql-injection-testing--',  # 30 chars
-            'crlfuzz-crlf-injection-finder-',  # 30 chars
-            'httpx-web-probe-fingerprint---',  # 30 chars
-            'manual-penetration-testing----',  # 30 chars
-            'burp-suite-professional-scan--',  # 30 chars
-            'owasp-zap-security-scanner----',  # 30 chars
-            'nmap-network-service-scanner--',  # 30 chars
-            'nikto-web-server-scanner------',  # 30 chars
-            'wpscan-wordpress-vuln-scan----',  # 30 chars
-            'dirsearch-directory-brute-----',  # 30 chars
-            'ffuf-web-fuzzer-content-disc--',  # 30 chars
-            'amass-subdomain-enumeration---',  # 30 chars
-            'subfinder-passive-subdomain---',  # 30 chars
-            'masscan-port-scanner-fast-----',  # 30 chars
-            'nessus-vulnerability-assess---',  # 30 chars
-            'qualys-cloud-security-scan----',  # 30 chars
-            'acunetix-web-vuln-scanner-----',  # 30 chars
-            'semgrep-static-code-analysis--',  # 30 chars
+            'nuclei-vulnerability-scanner--',
+            'dalfox-xss-parameter-analysis-',
+            'sqlmap-sql-injection-testing--',
+            'crlfuzz-crlf-injection-finder-',
+            'httpx-web-probe-fingerprint---',
+            'manual-penetration-testing----',
+            'burp-suite-professional-scan--',
+            'owasp-zap-security-scanner----',
         ]
         severities = ['unknown', 'info', 'low', 'medium', 'high', 'critical']
         
-        # 获取域名目标
-        cur.execute("SELECT id, name FROM target WHERE type = 'domain' AND deleted_at IS NULL LIMIT 80")
-        domain_targets = cur.fetchall()
+        # 漏洞路径后缀（会追加到 website URL 后面）
+        vuln_paths = [
+            '/api/users?id=1',
+            '/api/admin/config',
+            '/api/v1/auth/login',
+            '/api/v2/data/export',
+            '/admin/settings',
+            '/debug/console',
+            '/backup/db.sql',
+            '/.env',
+            '/.git/config',
+            '/wp-admin/',
+            '/phpmyadmin/',
+            '/api/graphql',
+            '/swagger.json',
+            '/actuator/health',
+            '/metrics',
+        ]
+        
+        # 获取所有 website 的 URL 和 target_id
+        cur.execute("SELECT id, url, target_id FROM website LIMIT 500")
+        websites = cur.fetchall()
+        
+        if not websites:
+            print("  ⚠ 没有 website 数据，跳过漏洞生成\n")
+            return
         
         count = 0
         batch_data = []
-        for target_id, target_name in domain_targets:
-            num = random.randint(30, 80)
+        for website_id, website_url, target_id in websites:
+            # 每个 website 生成 1-5 个漏洞
+            num_vulns = random.randint(1, 5)
             
-            for idx in range(num):
+            for idx in range(num_vulns):
                 severity = random.choice(severities)
                 cvss_ranges = {
                     'critical': (9.0, 10.0), 'high': (7.0, 8.9), 'medium': (4.0, 6.9),
@@ -1264,22 +1355,22 @@ class TestDataGenerator:
                 cvss_range = cvss_ranges.get(severity, (0.0, 10.0))
                 cvss_score = round(random.uniform(*cvss_range), 1)
                 
-                # 生成固定 245 长度的 URL
-                url = generate_fixed_length_url(target_name, length=245, path_hint=f'vuln/{idx:04d}')
+                # 漏洞 URL = website URL + 漏洞路径
+                # 先移除 website URL 中的查询参数
+                base_url = website_url.split('?')[0]
+                vuln_url = base_url + random.choice(vuln_paths)
                 
-                # 生成固定 300 长度的描述
                 description = generate_fixed_length_text(length=300, text_type='description')
                 
                 raw_output = json.dumps({
                     'template': f'CVE-2024-{random.randint(10000, 99999)}',
                     'matcher_name': 'default',
                     'severity': severity,
-                    'host': target_name,
-                    'matched_at': url,
+                    'matched_at': vuln_url,
                 })
                 
                 batch_data.append((
-                    target_id, url, random.choice(vuln_types), severity,
+                    target_id, vuln_url, random.choice(vuln_types), severity,
                     random.choice(sources), cvss_score, description, raw_output
                 ))
                 count += 1
@@ -1401,13 +1492,23 @@ class TestDataGenerator:
                 # 生成固定 245 长度的 URL
                 url = generate_fixed_length_url(target_name, length=245, path_hint=f'website-snap/{i:04d}')
                 
+                # 生成模拟的响应头数据
+                response_headers = {
+                    'server': random.choice(['nginx', 'Apache', 'cloudflare']),
+                    'content_type': 'text/html; charset=utf-8',
+                    'x_frame_options': random.choice(['DENY', 'SAMEORIGIN', None]),
+                }
+                # 移除 None 值
+                response_headers = {k: v for k, v in response_headers.items() if v is not None}
+                
                 batch_data.append((
                     scan_id, url, target_name, random.choice(titles),
                     random.choice(webservers), random.choice(tech_stacks),
                     random.choice([200, 301, 403]),
                     random.randint(1000, 50000), 'text/html; charset=utf-8',
                     '',  # location 字段
-                    '<!DOCTYPE html><html><head><title>Test</title></head><body>Content</body></html>'
+                    '<!DOCTYPE html><html><head><title>Test</title></head><body>Content</body></html>',
+                    generate_raw_response_headers(response_headers)
                 ))
                 count += 1
         
@@ -1415,11 +1516,12 @@ class TestDataGenerator:
         if batch_data:
             execute_values(cur, """
                 INSERT INTO website_snapshot (
-                    scan_id, url, host, title, web_server, tech, status,
-                    content_length, content_type, location, body_preview, created_at
+                    scan_id, url, host, title, webserver, tech, status_code,
+                    content_length, content_type, location, response_body,
+                    response_headers, created_at
                 ) VALUES %s
                 ON CONFLICT DO NOTHING
-            """, batch_data, template="(%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, NOW())")
+            """, batch_data, template="(%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, NOW())")
                 
         print(f"  ✓ 创建了 {count} 个网站快照\n")
 
@@ -1498,6 +1600,13 @@ class TestDataGenerator:
                 num_tags = random.randint(10, 20)
                 tags = random.sample(all_tags, min(num_tags, len(all_tags)))
                 
+                # 生成模拟的响应头数据
+                response_headers = {
+                    'server': 'nginx/1.24.0',
+                    'content_type': 'application/json',
+                    'x_request_id': f'req_{random.randint(100000, 999999)}',
+                }
+                
                 batch_data.append((
                     scan_id, url, target_name, title,
                     random.choice([200, 201, 401, 403, 404]),
@@ -1506,7 +1615,8 @@ class TestDataGenerator:
                     'nginx/1.24.0',
                     'application/json', tech_list,
                     '{"status":"ok","data":{}}',
-                    tags
+                    tags,
+                    generate_raw_response_headers(response_headers)
                 ))
                 count += 1
         
@@ -1515,11 +1625,11 @@ class TestDataGenerator:
             execute_values(cur, """
                 INSERT INTO endpoint_snapshot (
                     scan_id, url, host, title, status_code, content_length,
-                    location, webserver, content_type, tech, body_preview,
-                    matched_gf_patterns, created_at
+                    location, webserver, content_type, tech, response_body,
+                    matched_gf_patterns, response_headers, created_at
                 ) VALUES %s
                 ON CONFLICT DO NOTHING
-            """, batch_data, template="(%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, NOW())")
+            """, batch_data, template="(%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, NOW())")
                 
         print(f"  ✓ 创建了 {count} 个端点快照\n")
 
@@ -2543,9 +2653,10 @@ class MillionDataGenerator:
                 if len(batch_data) >= batch_size:
                     execute_values(cur, """
                         INSERT INTO website (url, target_id, host, title, webserver, tech, 
-                            status_code, content_length, content_type, location, body_preview, created_at)
+                            status_code, content_length, content_type, location, response_body, 
+                            vhost, response_headers, created_at)
                         VALUES %s ON CONFLICT DO NOTHING
-                    """, batch_data, template="(%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, NOW())")
+                    """, batch_data, template="(%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, NULL, '', NOW())")
                     self.conn.commit()
                     batch_data = []
                     print(f"    ✓ {count:,} / {target_count:,}")
@@ -2555,9 +2666,10 @@ class MillionDataGenerator:
         if batch_data:
             execute_values(cur, """
                 INSERT INTO website (url, target_id, host, title, webserver, tech, 
-                    status_code, content_length, content_type, location, body_preview, created_at)
+                    status_code, content_length, content_type, location, response_body, 
+                    vhost, response_headers, created_at)
                 VALUES %s ON CONFLICT DO NOTHING
-            """, batch_data, template="(%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, NOW())")
+            """, batch_data, template="(%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, NULL, '', NOW())")
             self.conn.commit()
                 
         print(f"  ✓ 创建了 {count:,} 个网站\n")
@@ -2631,10 +2743,10 @@ class MillionDataGenerator:
                 if len(batch_data) >= batch_size:
                     execute_values(cur, """
                         INSERT INTO endpoint (url, target_id, host, title, webserver, status_code,
-                            content_length, content_type, tech, location, body_preview, vhost, 
-                            matched_gf_patterns, created_at)
+                            content_length, content_type, tech, location, response_body, vhost, 
+                            matched_gf_patterns, response_headers, created_at)
                         VALUES %s ON CONFLICT DO NOTHING
-                    """, batch_data, template="(%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, NOW())")
+                    """, batch_data, template="(%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, '', NOW())")
                     self.conn.commit()
                     batch_data = []
                     print(f"    ✓ {count:,} / {target_count:,}")
@@ -2644,10 +2756,10 @@ class MillionDataGenerator:
         if batch_data:
             execute_values(cur, """
                 INSERT INTO endpoint (url, target_id, host, title, webserver, status_code,
-                    content_length, content_type, tech, location, body_preview, vhost, 
-                    matched_gf_patterns, created_at)
+                    content_length, content_type, tech, location, response_body, vhost, 
+                    matched_gf_patterns, response_headers, created_at)
                 VALUES %s ON CONFLICT DO NOTHING
-            """, batch_data, template="(%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, NOW())")
+            """, batch_data, template="(%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, '', NOW())")
             self.conn.commit()
                 
         print(f"  ✓ 创建了 {count:,} 个端点\n")

backend/scripts/worker-deploy/agent.sh

@@ -95,6 +95,7 @@ EOF
     
     RESPONSE=$(curl -k -s -X POST \
         -H "Content-Type: application/json" \
+        -H "X-Worker-API-Key: ${WORKER_API_KEY}" \
         -d "$REGISTER_DATA" \
         "${API_URL}/api/workers/register/" 2>/dev/null)
     
@@ -116,7 +117,7 @@ if [ -z "$WORKER_ID" ]; then
     # 等待 Server 就绪
     log "等待 Server 就绪..."
     for i in $(seq 1 30); do
-        if curl -k -s "${API_URL}/api/" > /dev/null 2>&1; then
+        if curl -k -s -H "X-Worker-API-Key: ${WORKER_API_KEY}" "${API_URL}/api/workers/config/?is_local=${IS_LOCAL}" > /dev/null 2>&1; then
             log "${GREEN}Server 已就绪${NC}"
             break
         fi
@@ -189,6 +190,7 @@ EOF
     RESPONSE_FILE=$(mktemp)
     HTTP_CODE=$(curl -k -s -o "$RESPONSE_FILE" -w "%{http_code}" -X POST \
         -H "Content-Type: application/json" \
+        -H "X-Worker-API-Key: ${WORKER_API_KEY}" \
         -d "$JSON_DATA" \
         "${API_URL}/api/workers/${WORKER_ID}/heartbeat/" 2>/dev/null || echo "000")
     RESPONSE_BODY=$(cat "$RESPONSE_FILE" 2>/dev/null)

backend/scripts/worker-deploy/install.sh

@@ -27,10 +27,50 @@ BLUE='\033[0;34m'
 RED='\033[0;31m'
 NC='\033[0m'
 
-log_info() { echo -e "${BLUE}[XingRin]${NC} $1"; }
-log_success() { echo -e "${GREEN}[XingRin]${NC} $1"; }
-log_warn() { echo -e "${YELLOW}[XingRin]${NC} $1"; }
-log_error() { echo -e "${RED}[XingRin]${NC} $1"; }
+# 渐变色定义
+CYAN='\033[0;36m'
+MAGENTA='\033[0;35m'
+BOLD='\033[1m'
+DIM='\033[2m'
+
+log_info() { echo -e "${CYAN}  ▸${NC} $1"; }
+log_success() { echo -e "${GREEN}  ✔${NC} $1"; }
+log_warn() { echo -e "${YELLOW}  ⚠${NC} $1"; }
+log_error() { echo -e "${RED}  ✖${NC} $1"; }
+
+# 炫酷 Banner
+show_banner() {
+    echo -e ""
+    echo -e "${CYAN}${BOLD}    ██╗  ██╗██╗███╗   ██╗ ██████╗ ██████╗ ██╗███╗   ██╗${NC}"
+    echo -e "${CYAN}    ╚██╗██╔╝██║████╗  ██║██╔════╝ ██╔══██╗██║████╗  ██║${NC}"
+    echo -e "${BLUE}${BOLD}     ╚███╔╝ ██║██╔██╗ ██║██║  ███╗██████╔╝██║██╔██╗ ██║${NC}"
+    echo -e "${BLUE}     ██╔██╗ ██║██║╚██╗██║██║   ██║██╔══██╗██║██║╚██╗██║${NC}"
+    echo -e "${MAGENTA}${BOLD}    ██╔╝ ██╗██║██║ ╚████║╚██████╔╝██║  ██║██║██║ ╚████║${NC}"
+    echo -e "${MAGENTA}    ╚═╝  ╚═╝╚═╝╚═╝  ╚═══╝ ╚═════╝ ╚═╝  ╚═╝╚═╝╚═╝  ╚═══╝${NC}"
+    echo -e ""
+    echo -e "${DIM}    ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━${NC}"
+    echo -e "${BOLD}      🚀 分布式安全扫描平台 │ Worker 节点部署${NC}"
+    echo -e "${DIM}    ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━${NC}"
+    echo -e ""
+}
+
+# 完成 Banner
+show_complete() {
+    echo -e ""
+    echo -e "${GREEN}${BOLD}    ╔═══════════════════════════════════════════════════╗${NC}"
+    echo -e "${GREEN}${BOLD}    ║                                                   ║${NC}"
+    echo -e "${GREEN}${BOLD}    ║   ██████╗  ██████╗ ███╗   ██╗███████╗██╗          ║${NC}"
+    echo -e "${GREEN}${BOLD}    ║   ██╔══██╗██╔═══██╗████╗  ██║██╔════╝██║          ║${NC}"
+    echo -e "${GREEN}${BOLD}    ║   ██║  ██║██║   ██║██╔██╗ ██║█████╗  ██║          ║${NC}"
+    echo -e "${GREEN}${BOLD}    ║   ██║  ██║██║   ██║██║╚██╗██║██╔══╝  ╚═╝          ║${NC}"
+    echo -e "${GREEN}${BOLD}    ║   ██████╔╝╚██████╔╝██║ ╚████║███████╗██╗          ║${NC}"
+    echo -e "${GREEN}${BOLD}    ║   ╚═════╝  ╚═════╝ ╚═╝  ╚═══╝╚══════╝╚═╝          ║${NC}"
+    echo -e "${GREEN}${BOLD}    ║                                                   ║${NC}"
+    echo -e "${GREEN}${BOLD}    ║       ✨ XingRin Worker 节点部署完成！            ║${NC}"
+    echo -e "${GREEN}${BOLD}    ║                                                   ║${NC}"
+    echo -e "${GREEN}${BOLD}    ╚═══════════════════════════════════════════════════╝${NC}"
+    echo -e ""
+}
 
 # 等待 apt 锁释放
 wait_for_apt_lock() {
@@ -150,9 +190,7 @@ pull_image() {
 
 # 主流程
 main() {
-    log_info "=========================================="
-    log_info "  XingRin 节点安装"
-    log_info "=========================================="
+    show_banner
     
     detect_os
     install_docker
@@ -162,9 +200,7 @@ main() {
     
     touch "$DOCKER_MARKER"
     
-    log_success "=========================================="
-    log_success "  ✓ 安装完成"
-    log_success "=========================================="
+    show_complete
 }
 
 main "$@"

backend/scripts/worker-deploy/start-agent.sh

@@ -30,6 +30,7 @@ IMAGE="${DOCKER_USER}/xingrin-agent:${IMAGE_TAG}"
 # 预设变量（远程部署时由 deploy_service.py 替换）
 PRESET_SERVER_URL="{{HEARTBEAT_API_URL}}"
 PRESET_WORKER_ID="{{WORKER_ID}}"
+PRESET_API_KEY="{{WORKER_API_KEY}}"
 
 # 颜色定义
 GREEN='\033[0;32m'
@@ -68,6 +69,7 @@ start_agent() {
         -e SERVER_URL="${PRESET_SERVER_URL}" \
         -e WORKER_ID="${PRESET_WORKER_ID}" \
         -e IMAGE_TAG="${IMAGE_TAG}" \
+        -e WORKER_API_KEY="${PRESET_API_KEY}" \
         -v /proc:/host/proc:ro \
         ${IMAGE}
     

docker/.env.example

@@ -9,9 +9,8 @@ DB_USER=postgres
 DB_PASSWORD=123.com
 
 # ==================== Redis 配置 ====================
-# 在 Docker 网络中，Redis 服务名称为 redis
+# Redis 仅在 Docker 内部网络使用，不暴露公网端口
 REDIS_HOST=redis
-REDIS_PORT=6379
 REDIS_DB=0
 
 # ==================== 服务端口配置 ====================
@@ -51,6 +50,12 @@ LOG_LEVEL=INFO
 # 是否记录命令执行日志（大量扫描时会增加磁盘占用）
 ENABLE_COMMAND_LOGGING=true
 
+# ==================== Worker API Key 配置 ====================
+# Worker 节点认证密钥（用于 Worker 与主服务器之间的 API 认证）
+# 生产环境务必更换为随机强密钥（建议 32 位以上随机字符串）
+# 生成方法: openssl rand -hex 32
+WORKER_API_KEY=change-me-to-a-secure-random-key
+
 # ==================== Docker Hub 配置（生产模式） ====================
 # 生产模式下从 Docker Hub 拉取镜像时使用
 DOCKER_USER=yyhuni

docker/docker-compose.dev.yml

@@ -2,9 +2,13 @@ services:
   # PostgreSQL（可选，使用远程数据库时不启动）
   # 本地模式: docker compose --profile local-db up -d
   # 远程模式: docker compose up -d（需配置 DB_HOST 为远程地址）
+  # 使用自定义镜像，预装 pg_ivm 扩展
   postgres:
     profiles: ["local-db"]
-    image: postgres:15
+    build:
+      context: ./postgres
+      dockerfile: Dockerfile
+    image: ${DOCKER_USER:-yyhuni}/xingrin-postgres:${IMAGE_TAG:-dev}
     restart: always
     environment:
       POSTGRES_DB: ${DB_NAME}
@@ -15,6 +19,9 @@ services:
       - ./postgres/init-user-db.sh:/docker-entrypoint-initdb.d/init-user-db.sh
     ports:
       - "${DB_PORT}:5432"
+    command: >
+      postgres
+      -c shared_preload_libraries=pg_ivm
     healthcheck:
       test: ["CMD-SHELL", "pg_isready -U ${DB_USER}"]
       interval: 5s
@@ -24,8 +31,6 @@ services:
   redis:
     image: redis:7-alpine
     restart: always
-    ports:
-      - "${REDIS_PORT}:6379"
     healthcheck:
       test: ["CMD", "redis-cli", "ping"]
       interval: 5s
@@ -49,7 +54,8 @@ services:
       - /opt/xingrin:/opt/xingrin
       - /var/run/docker.sock:/var/run/docker.sock
     healthcheck:
-      test: ["CMD", "curl", "-f", "http://localhost:8888/api/"]
+      # 使用专门的健康检查端点（无需认证）
+      test: ["CMD", "curl", "-f", "http://localhost:8888/api/health/"]
       interval: 30s
       timeout: 10s
       retries: 3
@@ -65,9 +71,10 @@ services:
     restart: always
     environment:
       - SERVER_URL=http://server:8888
-      - WORKER_NAME=本地节点
+      - WORKER_NAME=Local-Worker
       - IS_LOCAL=true
       - IMAGE_TAG=${IMAGE_TAG:-dev}
+      - WORKER_API_KEY=${WORKER_API_KEY}
     depends_on:
       server:
         condition: service_healthy

docker/docker-compose.yml

@@ -8,9 +8,13 @@
 
 services:
   # PostgreSQL（可选，使用远程数据库时不启动）
+  # 使用自定义镜像，预装 pg_ivm 扩展
   postgres:
     profiles: ["local-db"]
-    image: postgres:15
+    build:
+      context: ./postgres
+      dockerfile: Dockerfile
+    image: ${DOCKER_USER:-yyhuni}/xingrin-postgres:${IMAGE_TAG:?IMAGE_TAG is required}
     restart: always
     environment:
       POSTGRES_DB: ${DB_NAME}
@@ -21,6 +25,9 @@ services:
       - ./postgres/init-user-db.sh:/docker-entrypoint-initdb.d/init-user-db.sh
     ports:
       - "${DB_PORT}:5432"
+    command: >
+      postgres
+      -c shared_preload_libraries=pg_ivm
     healthcheck:
       test: ["CMD-SHELL", "pg_isready -U ${DB_USER}"]
       interval: 5s
@@ -30,8 +37,6 @@ services:
   redis:
     image: redis:7-alpine
     restart: always
-    ports:
-      - "${REDIS_PORT}:6379"
     healthcheck:
       test: ["CMD", "redis-cli", "ping"]
       interval: 5s
@@ -52,7 +57,8 @@ services:
       # Docker Socket 挂载：允许 Django 服务器执行本地 docker 命令（用于本地 Worker 任务分发）
       - /var/run/docker.sock:/var/run/docker.sock
     healthcheck:
-      test: ["CMD", "curl", "-f", "http://localhost:8888/api/"]
+      # 使用专门的健康检查端点（无需认证）
+      test: ["CMD", "curl", "-f", "http://localhost:8888/api/health/"]
       interval: 30s
       timeout: 10s
       retries: 3
@@ -69,9 +75,10 @@ services:
     restart: always
     environment:
       - SERVER_URL=http://server:8888
-      - WORKER_NAME=本地节点
+      - WORKER_NAME=Local-Worker
       - IS_LOCAL=true
       - IMAGE_TAG=${IMAGE_TAG}
+      - WORKER_API_KEY=${WORKER_API_KEY}
     depends_on:
       server:
         condition: service_healthy

docker/nginx/nginx.conf

@@ -38,6 +38,8 @@ http {
       proxy_set_header Host $host;
       proxy_set_header X-Real-IP $remote_addr;
       proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
+      proxy_read_timeout 300s;  # 5分钟，支持大数据量导出
+      proxy_send_timeout 300s;
       proxy_pass http://backend;
     }
 

docker/postgres/Dockerfile

@@ -0,0 +1,19 @@
+FROM postgres:15
+
+# 安装编译依赖
+RUN apt-get update && apt-get install -y \
+    build-essential \
+    postgresql-server-dev-15 \
+    git \
+    && rm -rf /var/lib/apt/lists/*
+
+# 编译安装 pg_ivm
+RUN git clone https://github.com/sraoss/pg_ivm.git /tmp/pg_ivm \
+    && cd /tmp/pg_ivm \
+    && make \
+    && make install \
+    && rm -rf /tmp/pg_ivm
+
+# 配置 shared_preload_libraries
+# 注意: 这个配置会在容器启动时被应用
+RUN echo "shared_preload_libraries = 'pg_ivm'" >> /usr/share/postgresql/postgresql.conf.sample

docker/postgres/init-user-db.sh

@@ -9,3 +9,12 @@ psql -v ON_ERROR_STOP=1 --username "$POSTGRES_USER" --dbname "postgres" <<-EOSQL
 	GRANT ALL PRIVILEGES ON DATABASE xingrin TO "$POSTGRES_USER";
 	GRANT ALL PRIVILEGES ON DATABASE xingrin_dev TO "$POSTGRES_USER";
 EOSQL
+
+# 启用 pg_trgm 扩展（用于文本模糊搜索索引）
+psql -v ON_ERROR_STOP=1 --username "$POSTGRES_USER" --dbname "xingrin" <<-EOSQL
+	CREATE EXTENSION IF NOT EXISTS pg_trgm;
+EOSQL
+
+psql -v ON_ERROR_STOP=1 --username "$POSTGRES_USER" --dbname "xingrin_dev" <<-EOSQL
+	CREATE EXTENSION IF NOT EXISTS pg_trgm;
+EOSQL