ci: Validate并行化-拆分为CodeQuality/TypeCheck/Migration三个并行job

.gitea/workflows/ci-failure-monitor.yml

@@ -1,78 +0,0 @@
-name: CI Failure Monitor
-
-on:
-  schedule:
-    - cron: '0 */6 * * *'  # 每6小时检查一次
-  workflow_dispatch:
-    inputs:
-      days:
-        description: '统计最近N天的失败'
-        required: false
-        default: '7'
-      fail_threshold:
-        description: '失败次数阈值'
-        required: false
-        default: '3'
-      fail_rate_threshold:
-        description: '失败率阈值(%)'
-        required: false
-        default: '30'
-
-permissions:
-  contents: read
-
-jobs:
-  monitor:
-    name: CI重复失败检测
-    runs-on: ci-l2
-    timeout-minutes: 10
-
-    steps:
-    - name: Checkout code
-      shell: sh
-      env:
-        GITHUB_TOKEN: ${{ github.token }}
-      run: |
-        curl -sH "Authorization: token $GITHUB_TOKEN" \
-          "${GITHUB_API_URL}/repos/${GITHUB_REPOSITORY}/raw/scripts/ci/step_checkout.sh?ref=${GITHUB_SHA}" \
-          | bash
-
-    - name: Record job start time
-      shell: sh
-      run: bash scripts/ci/step_timer_start.sh
-
-    - name: Run failure detection
-      shell: sh
-      env:
-        GITEA_API_TOKEN: ${{ secrets.REVIEW_GITEA_TOKEN }}
-        GITEA_URL: https://git.xiaoxiajianji.com
-        GITEA_REPO: xiaoxia/xiaoxia-saas
-        CI_NOTIFY_WEBHOOK: ${{ secrets.CI_NOTIFY_WEBHOOK }}
-        FAIL_CHECK_DAYS: ${{ inputs.days || 7 }}
-        FAIL_THRESHOLD: ${{ inputs.fail_threshold || 3 }}
-        FAIL_RATE_THRESHOLD: ${{ inputs.fail_rate_threshold || 30 }}
-      run: |
-        set +e
-        python3 scripts/ci/ci_repeated_failure_detector.py
-        EXIT_CODE=$?
-        echo "检测完成，退出码: $EXIT_CODE"
-        # 0=无异常, 1=有警告, 2=有严重问题
-        # 监控脚本永远不fail，避免告警风暴
-        exit 0
-
-    - name: Job duration summary
-      if: always()
-      shell: sh
-      run: bash scripts/ci/step_timer_end.sh
-
-    - name: Report CI trace
-      if: always()
-      shell: sh
-      env:
-        AGENTLOOP_LICENSE_KEY: ${{ secrets.AGENTLOOP_LICENSE_KEY }}
-      run: |
-        STATUS="ok"
-        [ ${{ job.status }} = "success" ] || STATUS="error"
-        START_TIME=""
-        [ -f /tmp/ci_job_start_time ] && START_TIME=$(cat /tmp/ci_job_start_time)
-        python3 scripts/ci/ci_trace_report.py --service xiaoxia-saas-ci --status $STATUS --start-time "$START_TIME" || true

.gitea/workflows/ci-pipeline.yml

@@ -76,6 +76,103 @@ jobs:
         [ -f /tmp/ci_job_start_time ] && START_TIME=$(cat /tmp/ci_job_start_time)
         python3 scripts/ci/ci_trace_report.py --service xiaoxia-saas-ci --status $STATUS --start-time "$START_TIME" || true
 
+  validate:
+    needs: check-frontend-only
+    if: always() && needs.check-frontend-only.outputs.skip_backend != 'true'
+    name: Validate Code Quality And Tests
+    runs-on: ci-l2
+    timeout-minutes: 10
+    permissions:
+      contents: write
+    env:
+      DATABASE_URL: postgresql+psycopg://postgres:postgres@host.docker.internal:5432/xiaoxia_saas
+      USE_IN_MEMORY_DB: 'false'
+      CI_USE_SHARED_PG: 'true'
+    steps:
+    - name: Checkout code
+      shell: sh
+      env:
+        GITHUB_TOKEN: ${{ github.token }}
+      run: |
+        curl -sH "Authorization: token $GITHUB_TOKEN"           "${GITHUB_API_URL}/repos/${GITHUB_REPOSITORY}/raw/scripts/ci/step_checkout.sh?ref=${GITHUB_SHA}"           | bash
+    - name: Record job start time
+      shell: sh
+      run: bash scripts/ci/step_timer_start.sh
+    - name: Install dependencies
+      shell: sh
+      run: |
+        set -eu
+        # pip install 带重试（网络不稳定时自动重试）
+        for i in 1 2 3; do
+          python3 -m pip install -q -r requirements-base.txt && break
+          echo "pip install requirements-base.txt 失败，重试 $i/3..."
+          [ $i -eq 3 ] && exit 1
+          sleep 5
+        done
+        for i in 1 2 3; do
+          python3 -m pip install -q -r requirements.txt && break
+          echo "pip install requirements.txt 失败，重试 $i/3..."
+          [ $i -eq 3 ] && exit 1
+          sleep 5
+        done
+        for i in 1 2 3; do
+          python3 -m pip install -q -r requirements-dev.txt && break
+          echo "pip install requirements-dev.txt 失败，重试 $i/3..."
+          [ $i -eq 3 ] && exit 1
+          sleep 5
+        done
+        for i in 1 2 3; do
+          python3 -m pip install --no-binary :all: black==26.5.1 isort==8.0.1 && break
+          echo "pip install black/isort 失败，重试 $i/3..."
+          [ $i -eq 3 ] && exit 1
+          sleep 5
+        done
+    - name: Run all quality checks
+      shell: bash
+      env:
+        GITHUB_TOKEN: ${{ github.token }}
+      run: bash scripts/ci/run_validate.sh
+    - name: Auto-fix formatting (black + isort)
+      if: failure()
+      shell: sh
+      env:
+        GITHUB_TOKEN: ${{ github.token }}
+      run: python3 scripts/ci/auto_fix_formatting.py
+    - name: CI failure notification
+      if: failure()
+      shell: sh
+      env:
+        GITEA_TOKEN: ${{ secrets.GITEA_TOKEN }}
+        CI_WEBHOOK_URL: ${{ secrets.CI_WEBHOOK_URL }}
+      run: |
+        set +e
+        FAILED_JOB="Validate Code Quality And Tests" python3 scripts/ci_notify_failure.py
+    - name: Job duration summary
+      if: always()
+      shell: sh
+      run: bash scripts/ci/step_timer_end.sh
+    - name: Notify on failure
+      continue-on-error: true
+      if: failure()
+      shell: sh
+      env:
+        CI_NOTIFY_WEBHOOK: ${{ secrets.CI_NOTIFY_WEBHOOK }}
+      run: |
+        set +e
+        NOTIFY_MODE=failure JOB_NAME="Validate Code Quality And Tests" python3 scripts/ci_notify.py
+    - name: Report CI trace
+      if: always()
+      shell: sh
+      env:
+        AGENTLOOP_LICENSE_KEY: ${{ secrets.AGENTLOOP_LICENSE_KEY }}
+      run: |
+        STATUS="ok"
+        [ ${{ job.status }} = "success" ] || STATUS="error"
+        START_TIME=""
+        [ -f /tmp/ci_job_start_time ] && START_TIME=$(cat /tmp/ci_job_start_time)
+        python3 scripts/ci/ci_trace_report.py --service xiaoxia-saas-ci --status $STATUS --start-time "$START_TIME" || true
+
+
   validate-code-quality:
     name: Validate - Code Quality
     runs-on: ci-l2
@@ -621,65 +718,6 @@ jobs:
           echo "Docker login failed ($i/3), retrying in 5s..."
           sleep 5
         done
-    - name: Pre-build worker base images (fallback if not exist)
-      if: matrix.service == 'worker'
-      id: prebuild
-      shell: sh
-      run: |
-        set -eu
-        REGISTRY="git.xiaoxiajianji.com/xiaoxia-saas"
-        BASE_BUILDER="${REGISTRY}/worker-base-builder:latest"
-        BASE_RUNTIME="${REGISTRY}/worker-base-runtime:latest"
-        
-        # 尝试拉取基础镜像
-        echo "检查基础镜像..."
-        if docker pull "$BASE_BUILDER" 2>/dev/null && docker pull "$BASE_RUNTIME" 2>/dev/null; then
-          echo "基础镜像已存在，使用远程镜像"
-          echo "fallback=false" >> $GITHUB_OUTPUT
-        else
-          echo "基础镜像不存在，本地构建（fallback模式）..."
-          
-          # 构建builder基础镜像
-          echo "构建 worker-base-builder..."
-          # 用buildx docker-container驱动构建（兼容DooD模式：普通docker build看不到容器内文件）
-          BUILDER_NAME="ci-pr-builder-${GITHUB_RUN_ID:-local}"
-          if ! docker buildx inspect "$BUILDER_NAME" > /dev/null 2>&1; then
-            docker buildx create --use --name "$BUILDER_NAME" --driver docker-container
-          else
-            docker buildx use "$BUILDER_NAME"
-          fi
-          docker buildx inspect --bootstrap > /dev/null 2>&1
-          
-          # 构建builder基础镜像（带重试，buildx容器偶发不稳定）
-          echo "构建 worker-base-builder..."
-          for attempt in 1 2 3; do
-            if docker buildx build --load -f infra/docker/worker-base-builder.Dockerfile -t "$BASE_BUILDER" .; then
-              echo "worker-base-builder 构建成功"
-              break
-            fi
-            echo "worker-base-builder 构建失败，重试 $attempt/3..."
-            docker buildx rm "$BUILDER_NAME" 2>/dev/null || true
-            docker buildx create --use --name "$BUILDER_NAME" --driver docker-container
-            sleep 3
-          done
-          
-          # 构建runtime基础镜像
-          echo "构建 worker-base-runtime..."
-          for attempt in 1 2 3; do
-            if docker buildx build --load -f infra/docker/worker-base-runtime.Dockerfile -t "$BASE_RUNTIME" .; then
-              echo "worker-base-runtime 构建成功"
-              break
-            fi
-            echo "worker-base-runtime 构建失败，重试 $attempt/3..."
-            docker buildx rm "$BUILDER_NAME" 2>/dev/null || true
-            docker buildx create --use --name "$BUILDER_NAME" --driver docker-container
-            sleep 3
-          done
-          
-          echo "fallback=true" >> $GITHUB_OUTPUT
-          echo "基础镜像本地构建完成"
-        fi
-
     - name: Build PR image (verify only, no push)
       shell: sh
       run: |
@@ -693,18 +731,6 @@ jobs:
           EXTRA_BUILD_ARGS="$EXTRA_BUILD_ARGS NGINX_CONF=infra/docker/nginx-staging.conf"
         fi
 
-        # Worker fallback模式：基础镜像本地已构建，用普通docker build绕过buildx
-        if [ "${{ matrix.service }}" = "worker" ] && [ "${{ steps.prebuild.outputs.fallback }}" = "true" ]; then
-          echo "Fallback模式：用普通docker build（基础镜像本地已构建）"
-          BUILD_ARG_STR=""
-          for arg in $EXTRA_BUILD_ARGS; do
-            BUILD_ARG_STR="$BUILD_ARG_STR --build-arg $arg"
-          done
-          docker build -f ${{ matrix.dockerfile }} -t "${IMAGE_TAG}" $BUILD_ARG_STR .
-          echo "Fallback PR Build successful"
-          exit 0
-        fi
-
         NO_CACHE_FLAG=""
         for i in 1 2 3; do
           echo "PR Build attempt $i/3"
@@ -1553,4 +1579,5 @@ jobs:
         [ ${{ job.status }} = "success" ] || STATUS="error"
         START_TIME=""
         [ -f /tmp/ci_job_start_time ] && START_TIME=$(cat /tmp/ci_job_start_time)
-        python3 scripts/ci/ci_trace_report.py --service xiaoxia-saas-ci --status $STATUS --start-time "$START_TIME" || true
+        python3 scripts/ci/ci_trace_report.py --service xiaoxia-saas-ci --status $STATUS --start-time "$START_TIME" || true
+

.gitea/workflows/pr-automation.yml

@@ -54,9 +54,7 @@ jobs:
             CONTEXTS=("CI/CD Pipeline / Frontend Lint (pull_request)")
           else
             CONTEXTS=(
-              "CI/CD Pipeline / Validate - Code Quality (pull_request)"
-              "CI/CD Pipeline / Validate - Type Check (mypy) (pull_request)"
-              "CI/CD Pipeline / Validate - Migration (alembic) (pull_request)"
+              "CI/CD Pipeline / Validate Code Quality And Tests (pull_request)"
               "CI/CD Pipeline / Frontend Lint (pull_request)"
             )
           fi
@@ -235,9 +233,7 @@ jobs:
             echo "纯前端改动，只检查Frontend Lint"
           else
             CONTEXTS=(
-              "CI/CD Pipeline / Validate - Code Quality (pull_request)"
-              "CI/CD Pipeline / Validate - Type Check (mypy) (pull_request)"
-              "CI/CD Pipeline / Validate - Migration (alembic) (pull_request)"
+              "CI/CD Pipeline / Validate Code Quality And Tests (pull_request)"
               "CI/CD Pipeline / Frontend Lint (pull_request)"
               "CI/CD Pipeline / PR Build API Image (pull_request)"
               "CI/CD Pipeline / PR Build Worker Image (pull_request)"

.gitea/workflows/worker-base-image.yml

@@ -1,103 +0,0 @@
-name: Worker Base Image Build
-
-on:
-  push:
-    branches:
-      - develop
-      - main
-    paths:
-      - 'requirements-base.txt'
-      - 'requirements-worker.txt'
-      - 'infra/docker/worker-base-builder.Dockerfile'
-      - 'infra/docker/worker-base-runtime.Dockerfile'
-  workflow_dispatch:  # 支持手动触发
-
-jobs:
-  build-worker-base:
-    name: Build Worker Base Images
-    runs-on: runtime-builder
-    timeout-minutes: 30
-    strategy:
-      fail-fast: false
-      matrix:
-        include:
-        - name: builder
-          dockerfile: infra/docker/worker-base-builder.Dockerfile
-          image_name: worker-base-builder
-          cache_name: worker-base-builder-cache
-        - name: runtime
-          dockerfile: infra/docker/worker-base-runtime.Dockerfile
-          image_name: worker-base-runtime
-          cache_name: worker-base-runtime-cache
-    steps:
-    - name: Checkout code
-      shell: sh
-      env:
-        GITHUB_TOKEN: ${{ github.token }}
-      run: |
-        curl -sH "Authorization: token $GITHUB_TOKEN"           "${GITHUB_API_URL}/repos/${GITHUB_REPOSITORY}/raw/scripts/ci/step_checkout.sh?ref=${GITHUB_SHA}"           | bash
-
-    - name: Docker login to Registry
-      shell: sh
-      env:
-        ACR_USERNAME: ${{ secrets.ACR_USERNAME }}
-        ACR_PASSWORD: ${{ secrets.ACR_PASSWORD }}
-        GITEA_REGISTRY_USER: xiaoxia
-        GITEA_REGISTRY_TOKEN: ${{ secrets.REGISTRY_TOKEN }}
-      run: |
-        set -eu
-        for i in 1 2 3; do
-          echo "=== Docker login 尝试 $i/3 ==="
-          if printf '%s' "${ACR_PASSWORD}" | docker login xiaoxia-registry.cn-hangzhou.cr.aliyuncs.com -u "${ACR_USERNAME}" --password-stdin              && docker login git.xiaoxiajianji.com -u "${GITEA_REGISTRY_USER}" -p "${GITEA_REGISTRY_TOKEN}"; then
-            echo "✅ Docker login successful"
-            break
-          fi
-          echo "❌ Docker login 失败（尝试 $i/3），5s 后重试..."
-          sleep 5
-        done
-
-    - name: Setup buildx builder
-      shell: sh
-      run: |
-        set -eu
-        BUILDER_NAME="ci-builder-${GITHUB_RUN_ID}-${{ matrix.name }}"
-        if ! docker buildx inspect "$BUILDER_NAME" > /dev/null 2>&1; then
-          docker buildx create --use --name "$BUILDER_NAME" --driver docker-container
-          echo "Created $BUILDER_NAME"
-        else
-          docker buildx use "$BUILDER_NAME"
-          echo "Using existing $BUILDER_NAME"
-        fi
-        docker buildx inspect --bootstrap
-
-    - name: Build and push base image
-      shell: sh
-      run: |
-        set -eu
-        REGISTRY="xiaoxia-registry.cn-hangzhou.cr.aliyuncs.com/xiaoxiakeji"
-        IMAGE_TAG="${REGISTRY}/${{ matrix.image_name }}:latest"
-        SAFE_REF_NAME=$(echo "${GITHUB_REF_NAME}" | tr '/' '-')
-        CACHE_REF="${REGISTRY}/${{ matrix.cache_name }}:${SAFE_REF_NAME}"
-        
-        echo "=== Building ${{ matrix.name }} base image ==="
-        echo "Image: ${IMAGE_TAG}"
-        echo "Cache: ${CACHE_REF}"
-        
-        # 用通用构建脚本
-        bash scripts/ci/docker_build_push.sh           ${{ matrix.dockerfile }}           "${IMAGE_TAG}"           "${CACHE_REF}"
-        
-        # 同时推送到 Gitea Packages 作为备份（可选）
-        GITEA_IMAGE="git.xiaoxiajianji.com/xiaoxia-saas/${{ matrix.image_name }}:latest"
-        docker tag "${IMAGE_TAG}" "${GITEA_IMAGE}"
-        docker push "${GITEA_IMAGE}" || echo "Gitea Packages push failed (non-fatal)"
-        
-        echo ""
-        echo "✅ ${{ matrix.name }} base image built and pushed"
-
-    - name: Cleanup buildx builder
-      if: always()
-      shell: sh
-      run: |
-        docker buildx rm "ci-builder-${GITHUB_RUN_ID}-${{ matrix.name }}" 2>/dev/null || true
-        docker buildx prune -f 2>/dev/null || true
-        echo "Builder cleanup done"

infra/docker/worker-base-builder.Dockerfile

@@ -1,43 +0,0 @@
-# ============================================================
-# Worker Builder 基础镜像
-# 预编译：编译工具 + 基础依赖 + Worker大包
-# 当 requirements-base.txt 或 requirements-worker.txt 变更时重新构建
-# 业务构建从此镜像开始，只需要安装业务依赖，节省15+分钟
-# ============================================================
-
-FROM git.xiaoxiajianji.com/xiaoxia/base/python:3.12-slim
-
-# 使用阿里云镜像加速
-RUN sed -i 's|deb.debian.org|mirrors.aliyun.com|g' /etc/apt/sources.list.d/debian.sources 2>/dev/null || \
-    sed -i 's|deb.debian.org|mirrors.aliyun.com|g' /etc/apt/sources.list 2>/dev/null || true
-
-# 安装编译工具
-RUN apt-get update && apt-get install -y --no-install-recommends \
-    gcc \
-    g++ \
-    python3-dev \
-    binutils \
-    && rm -rf /var/lib/apt/lists/*
-
-# 创建 venv
-RUN python -m venv /opt/venv
-ENV PATH="/opt/venv/bin:$PATH"
-
-WORKDIR /tmp
-
-# 基础依赖（变化极少）
-COPY requirements-base.txt /tmp/requirements-base.txt
-RUN --mount=type=cache,target=/root/.cache/pip,sharing=locked \
-    pip install --no-cache-dir -i https://mirrors.aliyun.com/pypi/simple/ --trusted-host mirrors.aliyun.com \
-    -r /tmp/requirements-base.txt \
-    && rm /tmp/requirements-base.txt
-
-# Worker 大包（变化少）
-COPY requirements-worker.txt /tmp/requirements-worker.txt
-RUN --mount=type=cache,target=/root/.cache/pip,sharing=locked \
-    pip install --no-cache-dir -i https://mirrors.aliyun.com/pypi/simple/ --trusted-host mirrors.aliyun.com \
-    -r /tmp/requirements-worker.txt \
-    && rm /tmp/requirements-worker.txt
-
-# 预先做一次 strip（基础层瘦身，业务层增量）
-RUN find /opt/venv -name "*.so" -type f -exec strip --strip-all {} \; 2>/dev/null || true

infra/docker/worker-base-runtime.Dockerfile

@@ -1,17 +0,0 @@
-# ============================================================
-# Worker Runtime 基础镜像
-# 预安装：ffmpeg + 运行时依赖
-# 变化极少，业务构建从此镜像开始
-# ============================================================
-
-FROM git.xiaoxiajianji.com/xiaoxia/base/python:3.12-slim
-
-# 使用阿里云镜像加速
-RUN sed -i 's|deb.debian.org|mirrors.aliyun.com|g' /etc/apt/sources.list.d/debian.sources 2>/dev/null || \
-    sed -i 's|deb.debian.org|mirrors.aliyun.com|g' /etc/apt/sources.list 2>/dev/null || true
-
-# 运行时依赖：ffmpeg + opencv需要的libglib
-RUN apt-get update && apt-get install -y --no-install-recommends \
-    ffmpeg \
-    libglib2.0-0 \
-    && rm -rf /var/lib/apt/lists/*

infra/docker/worker.Dockerfile

@@ -1,43 +1,88 @@
 # ============================================================
-# Worker Dockerfile - 分层缓存优化版
-# 优化：基础依赖 + Worker大包预构建为基础镜像，业务构建仅叠加业务依赖
-# 基础镜像：worker-base-builder / worker-base-runtime
-# 预计节省：依赖不变时构建时间从23min降至5min以内
+# Worker Dockerfile - 优化版（多阶段构建 + 镜像瘦身 + cache mount加速）
+# 优化项：
+#   1. 多阶段构建：builder 阶段安装编译依赖，runtime 阶段只保留运行时
+#   2. ffmpeg 通过 apt 安装（阿里云镜像加速，几秒完成，稳定可靠）
+#   3. Python 依赖瘦身：strip .so 调试符号 + 清理测试文件 + 清理缓存
+#   4. pip cache mount：加速依赖下载（跨构建共享pip wheel缓存）
 # ============================================================
 
 # ==================== Builder 阶段 ====================
-# 从预构建的builder基础镜像开始，已经包含：
-#   - 编译工具 (gcc/g++/python3-dev/binutils)
-#   - requirements-base.txt 全部依赖
-#   - requirements-worker.txt 全部依赖 (numpy/scipy/opencv)
-#   - 预strip的.so文件
-FROM xiaoxia-registry.cn-hangzhou.cr.aliyuncs.com/xiaoxiakeji/worker-base-builder:latest AS builder
+FROM git.xiaoxiajianji.com/xiaoxia/base/python:3.12-slim AS builder
 
-ENV PATH="/opt/venv/bin:$PATH"
+# 使用阿里云镜像加速
+RUN sed -i 's|deb.debian.org|mirrors.aliyun.com|g' /etc/apt/sources.list.d/debian.sources 2>/dev/null || \
+    sed -i 's|deb.debian.org|mirrors.aliyun.com|g' /etc/apt/sources.list 2>/dev/null || true
 
+# 安装编译工具（仅 builder 需要）
+RUN apt-get update && apt-get install -y --no-install-recommends \
+    gcc \
+    g++ \
+    python3-dev \
+    binutils \
+    && rm -rf /var/lib/apt/lists/*
+
+
+# ---- 安装 Python 依赖 ----
 WORKDIR /tmp
 
-# ---- 安装业务依赖（变化频繁，单独一层）----
+# 创建 venv
+RUN python -m venv /opt/venv
+ENV PATH="/opt/venv/bin:$PATH"
+
+# 基础依赖
+COPY requirements-base.txt /tmp/requirements-base.txt
+RUN --mount=type=cache,target=/root/.cache/pip,sharing=locked \
+    pip install --no-cache-dir -i https://mirrors.aliyun.com/pypi/simple/ --trusted-host mirrors.aliyun.com \
+    -r /tmp/requirements-base.txt \
+    && rm /tmp/requirements-base.txt
+
+# Worker 专属大包
+COPY requirements-worker.txt /tmp/requirements-worker.txt
+RUN --mount=type=cache,target=/root/.cache/pip,sharing=locked \
+    pip install --no-cache-dir -i https://mirrors.aliyun.com/pypi/simple/ --trusted-host mirrors.aliyun.com \
+    -r /tmp/requirements-worker.txt \
+    && rm /tmp/requirements-worker.txt
+
+# 业务依赖
 COPY requirements.txt /tmp/requirements.txt
 RUN --mount=type=cache,target=/root/.cache/pip,sharing=locked \
     pip install --no-cache-dir -i https://mirrors.aliyun.com/pypi/simple/ --trusted-host mirrors.aliyun.com \
     -r /tmp/requirements.txt \
     && rm /tmp/requirements.txt
 
-# ---- 增量瘦身（只处理新增的业务依赖）----
+# ---- Python 依赖瘦身 ----
+# 1. strip .so 文件的调试符号（节省约 80-100MB）
 RUN find /opt/venv -name "*.so" -type f -exec strip --strip-all {} \; 2>/dev/null || true
+
+# 2. 清理测试文件（节省约 20MB）
+RUN find /opt/venv -type d -name "tests" -exec rm -rf {} + 2>/dev/null; \
+    find /opt/venv -type d -name "test" -exec rm -rf {} + 2>/dev/null; \
+    find /opt/venv -name "test_*.py" -delete 2>/dev/null || true
+
+# 3. 清理 .pyc 缓存和 __pycache__（节省约 10MB，运行时按需生成）
 RUN find /opt/venv -type d -name "__pycache__" -exec rm -rf {} + 2>/dev/null; \
     find /opt/venv -name "*.pyc" -delete 2>/dev/null || true
 
+# 4. 清理 dist-info 中的文档
+RUN find /opt/venv -name "*.dist-info" -type d -exec sh -c 'rm -f "$1"/DESCRIPTION.rst "$1"/INSTALLER "$1"/LICENSE* "$1"/WHEEL "$1"/entry_points.txt' _ {} \; 2>/dev/null || true
+
 # ==================== Runtime 阶段 ====================
-# 从预构建的runtime基础镜像开始，已经包含：
-#   - ffmpeg
-#   - libglib2.0-0
-FROM xiaoxia-registry.cn-hangzhou.cr.aliyuncs.com/xiaoxiakeji/worker-base-runtime:latest AS runtime
+FROM git.xiaoxiajianji.com/xiaoxia/base/python:3.12-slim AS runtime
 
 # 构建参数：版本号
 ARG APP_VERSION=dev
 
+# 使用阿里云镜像加速
+RUN sed -i 's|deb.debian.org|mirrors.aliyun.com|g' /etc/apt/sources.list.d/debian.sources 2>/dev/null || \
+    sed -i 's|deb.debian.org|mirrors.aliyun.com|g' /etc/apt/sources.list 2>/dev/null || true
+
+# 安装最小运行时依赖（opencv-python-headless 需要 libglib2.0-0）
+RUN apt-get update && apt-get install -y --no-install-recommends \
+    ffmpeg \
+    libglib2.0-0 \
+    && rm -rf /var/lib/apt/lists/*
+
 # 从 builder 复制 Python 虚拟环境
 COPY --from=builder /opt/venv /opt/venv
 

requirements-dev.txt

@@ -11,5 +11,4 @@ pytest==8.3.3
 pytest-asyncio==0.24.0
 pytest-cov==6.0.0
 pytest-timeout==2.3.1
-pytest-xdist==3.6.1
 diff-cover==8.0.3

scripts/ci/auto_fix_formatting.py

@@ -280,7 +280,7 @@ def main():
 
     # 提交修复
     run("git add -A")
-    run('git commit -m "style: auto-format with black + isort + prettier"')
+    run('git commit -m "style: auto-format with black + isort + prettier [ci skip]"')
 
     # 推送（head_branch已从ensure_git_repo获取）
     print(f"\nPR来源分支: {head_branch}")

scripts/ci/ci_failure_diagnosis.py

@@ -1,447 +0,0 @@
-#!/usr/bin/env python3
-"""CI失败诊断增强脚本：自动分类失败原因 + 提取关键错误 + 给出修复建议。
-# Trigger CI after auto-format fix
-
-支持的失败类型：
-1. Lint/格式问题 (ruff/black/eslint/prettier)
-2. 单元测试失败
-3. Docker构建失败
-4. 依赖安装失败 (pip/npm)
-5. 超时
-6. 缓存问题
-7. 数据库/迁移问题
-8. 网络问题
-9. 其他
-
-用法：
-    python3 scripts/ci/ci_failure_diagnosis.py [--job-name "Job Name"] [--log-file /path/to/log]
-
-如果不传--log-file，会尝试从Gitea API获取失败job的日志。
-"""
-
-import json
-import os
-import re
-import sys
-import urllib.request
-from dataclasses import dataclass, field
-from typing import List, Optional
-
-
-@dataclass
-class FailureDiagnosis:
-    """失败诊断结果"""
-
-    category: str  # 失败分类
-    category_cn: str  # 中文分类名
-    severity: str  # 严重程度: high / medium / low
-    summary: str  # 一句话摘要
-    error_lines: List[str] = field(default_factory=list)  # 关键错误行
-    suggestions: List[str] = field(default_factory=list)  # 修复建议
-    auto_fixable: bool = False  # 是否可以自动修复
-    related_docs: str = ""  # 相关文档链接
-
-
-# ============================================================
-# 失败模式定义
-# ============================================================
-
-FAILURE_PATTERNS = [
-    # ===== Lint / 格式问题 =====
-    {
-        "pattern": r"(ruff|black|isort)\b.*(error|failed|Error)",
-        "category": "lint_python",
-        "category_cn": "Python代码质量检查",
-        "severity": "low",
-        "summary_contains": ["ruff", "black", "isort"],
-        "suggestions": [
-            "本地运行 `black . && isort . && ruff check --fix .` 自动修复",
-            "使用 `scripts/agent-commit.sh` 提交（自动格式化）",
-            "如确认无误，可加 `# noqa: xxx` 忽略特定规则",
-        ],
-        "auto_fixable": True,
-    },
-    {
-        "pattern": r"ESLint|prettier|eslint",
-        "category": "lint_frontend",
-        "category_cn": "前端代码检查",
-        "severity": "low",
-        "summary_contains": ["eslint", "prettier"],
-        "suggestions": [
-            "本地运行 `cd apps/web && npm run lint:fix` 自动修复",
-            "Prettier问题: `cd apps/web && npx prettier --write .`",
-        ],
-        "auto_fixable": True,
-    },
-    {
-        "pattern": r"F\d{3}|E\d{3}|W\d{3}.*ruff|ruff.*F\d{3}",
-        "category": "lint_python",
-        "category_cn": "Python代码质量检查",
-        "severity": "low",
-        "suggestions": [
-            "F401: 删除未使用的import",
-            "F841: 删除未使用的变量或加下划线前缀",
-            "E501: 行超长，加 `# noqa: E501`",
-            "F811: 删重复import",
-            "运行 `ruff check --fix .` 自动修复大部分问题",
-        ],
-        "auto_fixable": True,
-    },
-    # ===== 单元测试失败 =====
-    {
-        "pattern": r"FAILED|assert.*Error|AssertionError",
-        "category": "unit_test",
-        "category_cn": "单元测试失败",
-        "severity": "high",
-        "suggestions": [
-            "检查相关测试文件，确认是代码问题还是测试用例问题",
-            "本地运行对应测试：`pytest path/to/test.py -v`",
-            "如测试依赖外部服务，检查mock是否正确",
-        ],
-        "auto_fixable": False,
-    },
-    {
-        "pattern": r"pytest.*failed|\d+ failed.*\d+ passed",
-        "category": "unit_test",
-        "category_cn": "单元测试失败",
-        "severity": "high",
-        "suggestions": [
-            "查看上方日志中的FAILED测试用例",
-            "检查失败断言的期望值 vs 实际值",
-            "新代码影响了现有测试行为，确认是预期内变更吗？",
-        ],
-        "auto_fixable": False,
-    },
-    # ===== Docker 构建失败 =====
-    {
-        "pattern": r"Dockerfile.*not found|docker build.*failed|ERROR: failed to solve",
-        "category": "docker_build",
-        "category_cn": "Docker构建失败",
-        "severity": "high",
-        "suggestions": [
-            "检查Dockerfile语法是否正确",
-            "检查引用的基础镜像是否存在",
-            "本地运行 `docker build -f path/to/Dockerfile .` 复现",
-        ],
-        "auto_fixable": False,
-    },
-    {
-        "pattern": r"manifest.*not found|no such image|image.*not found",
-        "category": "docker_build",
-        "category_cn": "镜像不存在",
-        "severity": "medium",
-        "suggestions": [
-            "检查基础镜像名称和tag是否正确",
-            "确认镜像仓库可访问，登录是否有效",
-            "如为新基础镜像，需先手动构建一次基础镜像",
-        ],
-        "auto_fixable": False,
-    },
-    {
-        "pattern": r"ETXTBSY|text file busy",
-        "category": "docker_build",
-        "category_cn": "文件锁冲突（ETXTBSY）",
-        "severity": "low",
-        "summary": "esbuild并发构建冲突，重试即可",
-        "suggestions": ["偶发问题，点击Rerun重新运行即可", "如频繁出现，检查是否有多个job并发写入同一文件"],
-        "auto_fixable": True,
-    },
-    # ===== 依赖安装失败 =====
-    {
-        "pattern": r"pip install.*error|Could not find a version|No matching distribution",
-        "category": "dependency",
-        "category_cn": "pip依赖安装失败",
-        "severity": "medium",
-        "suggestions": [
-            "检查requirements.txt中的版本号是否正确",
-            "如为新版本刚发布，可能源还没同步，稍后重试",
-            "检查网络连接，可尝试切换pip镜像源",
-        ],
-        "auto_fixable": False,
-    },
-    {
-        "pattern": r"npm.*ERR|npm install.*failed|E404|ECONNREFUSED.*npm",
-        "category": "dependency",
-        "category_cn": "npm依赖安装失败",
-        "severity": "medium",
-        "suggestions": [
-            "检查package.json中的版本号是否存在",
-            "网络问题：检查npm registry是否可访问",
-            "国内网络建议配置npmmirror镜像源",
-        ],
-        "auto_fixable": False,
-    },
-    {
-        "pattern": r"Connection refused|timed out|network.*unreachable",
-        "category": "network",
-        "category_cn": "网络问题",
-        "severity": "medium",
-        "summary": "网络连接失败，可能是源站问题或DNS问题",
-        "suggestions": [
-            "点击Rerun重试，网络问题通常是临时的",
-            "如持续失败，检查对应服务是否正常",
-            "检查Runner网络配置",
-        ],
-        "auto_fixable": True,
-    },
-    # ===== 超时 =====
-    {
-        "pattern": r"timeout|timed out|exceeded.*time limit|job.*cancelled.*timeout",
-        "category": "timeout",
-        "category_cn": "执行超时",
-        "severity": "medium",
-        "suggestions": [
-            "如首次出现：重试一次，可能是临时性能波动",
-            "频繁出现：检查构建是否变慢了，最近是否加了新依赖",
-            "可适当增加timeout-minutes配置",
-        ],
-        "auto_fixable": False,
-    },
-    # ===== 数据库/迁移 =====
-    {
-        "pattern": r"alembic.*error|migration.*failed|relation.*does not exist|column.*does not exist",
-        "category": "migration",
-        "category_cn": "数据库迁移失败",
-        "severity": "high",
-        "suggestions": [
-            "检查迁移脚本是否正确，down_revision是否对",
-            "确认数据库中是否有脏数据或残留表",
-            "迁移脚本合并冲突时，重新生成迁移文件",
-        ],
-        "auto_fixable": False,
-    },
-    # ===== 缓存问题 =====
-    {
-        "pattern": r"cache.*corrupt|cache.*invalid|snapshot.*not found|failed to compute cache key",
-        "category": "cache",
-        "category_cn": "缓存损坏",
-        "severity": "low",
-        "suggestions": ["构建系统会自动清理损坏缓存并重试，通常无需干预", "如持续失败，手动清理Runner上的缓存目录"],
-        "auto_fixable": True,
-    },
-    # ===== Checkout 失败 =====
-    {
-        "pattern": r"Could not resolve host|fatal:.*repository|SSL.*problem",
-        "category": "checkout",
-        "category_cn": "代码拉取失败",
-        "severity": "low",
-        "suggestions": ["临时网络问题，点击Rerun重试", "如持续失败，检查Gitea服务状态"],
-        "auto_fixable": True,
-    },
-]
-
-
-def analyze_log(log_text: str, job_name: str = "") -> FailureDiagnosis:
-    """分析日志，返回诊断结果"""
-
-    lines = log_text.strip().split("\n")
-
-    # 收集所有匹配的模式
-    matched = []
-    error_lines = []
-
-    for line in lines:
-        line_stripped = line.strip()
-        # 收集ERROR/FAILED/Failed等错误行（最多20行）
-        if re.search(r"(ERROR|FAILED|Error|error:|FAIL:|Traceback)", line_stripped):
-            if len(error_lines) < 20:
-                error_lines.append(line_stripped)
-
-        for pattern_info in FAILURE_PATTERNS:
-            if re.search(pattern_info["pattern"], line_stripped, re.IGNORECASE):
-                matched.append(pattern_info)
-                break  # 一行只匹配一个模式
-
-    if not matched:
-        # 未识别的失败类型
-        return FailureDiagnosis(
-            category="unknown",
-            category_cn="未知错误",
-            severity="medium",
-            summary="未识别的失败类型，需要人工查看日志",
-            error_lines=error_lines[:10],
-            suggestions=[
-                "点击'查看失败日志'查看完整日志",
-                "如为偶发问题，可先重试一次",
-                "常见原因：环境问题、配置问题、新增逻辑引入的bug",
-            ],
-            auto_fixable=False,
-        )
-
-    # 选最严重、最具体的那个
-    severity_order = {"high": 3, "medium": 2, "low": 1}
-    matched.sort(key=lambda x: severity_order.get(x["severity"], 0), reverse=True)
-    best_match = matched[0]
-
-    # 生成摘要
-    if "summary" in best_match:
-        summary = best_match["summary"]
-    else:
-        summary = f"{best_match['category_cn']}检查失败"
-        if job_name:
-            summary = f"[{job_name}] {summary}"
-
-    # 从error_lines中过滤出与该分类相关的
-    relevant_errors = error_lines[:10]
-
-    return FailureDiagnosis(
-        category=best_match["category"],
-        category_cn=best_match["category_cn"],
-        severity=best_match["severity"],
-        summary=summary,
-        error_lines=relevant_errors,
-        suggestions=best_match["suggestions"],
-        auto_fixable=best_match.get("auto_fixable", False),
-    )
-
-
-def fetch_failed_job_log(run_id: str, job_id: str, token: str, repo: str) -> Optional[str]:
-    """从Gitea API获取失败job的日志"""
-    api_base = f"https://git.xiaoxiajianji.com/api/v1/repos/{repo}"
-
-    # 尝试获取job的日志
-    url = f"{api_base}/actions/runs/{run_id}/jobs/{job_id}/log"
-    req = urllib.request.Request(url)
-    req.add_header("Authorization", f"token {token}")
-
-    try:
-        with urllib.request.urlopen(req, timeout=15) as resp:
-            return resp.read().decode("utf-8", errors="replace")
-    except Exception as e:
-        print(f"获取日志失败: {e}", file=sys.stderr)
-        return None
-
-
-def format_diagnosis_markdown(d: FailureDiagnosis, job_name: str = "", run_url: str = "") -> str:
-    """将诊断结果格式化为飞书卡片markdown"""
-
-    severity_emoji = {"high": "🔴", "medium": "🟡", "low": "🟢"}
-    emoji = severity_emoji.get(d.severity, "⚪")
-
-    lines = []
-    lines.append(f"**分类**: {emoji} {d.category_cn}")
-    lines.append(f"**问题**: {d.summary}")
-
-    if d.error_lines:
-        lines.append("")
-        lines.append("**关键错误行**:")
-        for err in d.error_lines[:5]:
-            # 截断过长的行
-            if len(err) > 150:
-                err = err[:147] + "..."
-            lines.append(f"  `{err}`")
-
-    lines.append("")
-    lines.append("**修复建议**:")
-    for i, s in enumerate(d.suggestions[:5], 1):
-        lines.append(f"  {i}. {s}")
-
-    if d.auto_fixable:
-        lines.append("")
-        lines.append("💡 **可自动修复**：如格式问题，可尝试点击Rerun让auto-fix自动处理")
-
-    if run_url:
-        lines.append("")
-        lines.append(f"[查看完整日志]({run_url})")
-
-    return "\n".join(lines)
-
-
-def main():
-    job_name = os.environ.get("FAILED_JOB", "")
-    run_id = os.environ.get("GITHUB_RUN_ID", "")
-    repo = os.environ.get("GITHUB_REPOSITORY", "xiaoxia/xiaoxia-saas")
-    token = os.environ.get("GITHUB_TOKEN", "")
-
-    # 1. 尝试获取日志
-    log_text = ""
-
-    # 优先从环境变量或文件读取
-    log_file = os.environ.get("CI_LOG_FILE", "")
-    if log_file and os.path.exists(log_file):
-        with open(log_file) as f:
-            log_text = f.read()
-    elif run_id and token:
-        # 尝试从API获取（需要job_id，这里简化处理）
-        pass
-
-    # 如果没有日志，用job_name做粗略分类
-    if not log_text:
-        # 基于job名做初始判断
-        if any(k in job_name.lower() for k in ["validate", "lint", "quality"]):
-            d = FailureDiagnosis(
-                category="lint_general",
-                category_cn="代码质量检查",
-                severity="low",
-                summary=f"{job_name} 检查失败（日志不可用，基于job名初步诊断）",
-                suggestions=["点击查看日志获取具体错误信息", "格式类问题通常可自动修复"],
-                auto_fixable=True,
-            )
-        elif "build" in job_name.lower():
-            d = FailureDiagnosis(
-                category="build_general",
-                category_cn="构建失败",
-                severity="high",
-                summary=f"{job_name} 构建失败（日志不可用）",
-                suggestions=["点击查看日志获取具体构建错误", "常见原因：Dockerfile错误、依赖安装失败、网络问题"],
-                auto_fixable=False,
-            )
-        elif "test" in job_name.lower():
-            d = FailureDiagnosis(
-                category="test_general",
-                category_cn="测试失败",
-                severity="high",
-                summary=f"{job_name} 测试失败（日志不可用）",
-                suggestions=["点击查看日志获取具体失败的测试用例", "检查最近代码改动是否影响了测试"],
-                auto_fixable=False,
-            )
-        elif "deploy" in job_name.lower():
-            d = FailureDiagnosis(
-                category="deploy_general",
-                category_cn="部署失败",
-                severity="high",
-                summary=f"{job_name} 部署失败（日志不可用）",
-                suggestions=["检查目标服务器状态和网络", "检查镜像是否正确推送", "查看服务器上的容器日志"],
-                auto_fixable=False,
-            )
-        else:
-            d = FailureDiagnosis(
-                category="unknown",
-                category_cn="未知错误",
-                severity="medium",
-                summary=f"{job_name} 失败",
-                suggestions=["点击查看日志获取详细信息"],
-                auto_fixable=False,
-            )
-    else:
-        d = analyze_log(log_text, job_name)
-
-    # 输出诊断结果
-    run_url = f"https://git.xiaoxiajianji.com/{repo}/actions/runs/{run_id}" if run_id else ""
-
-    print("=" * 60)
-    print("  CI 失败诊断报告")
-    print("=" * 60)
-    print()
-    print(format_diagnosis_markdown(d, job_name, run_url))
-    print()
-    print("=" * 60)
-
-    # 将诊断结果写入文件（供通知脚本读取）
-    output_file = os.environ.get("DIAGNOSIS_OUTPUT", "/tmp/ci_diagnosis.json")
-    result = {
-        "category": d.category,
-        "category_cn": d.category_cn,
-        "severity": d.severity,
-        "summary": d.summary,
-        "error_lines": d.error_lines,
-        "suggestions": d.suggestions,
-        "auto_fixable": d.auto_fixable,
-    }
-    with open(output_file, "w") as f:
-        json.dump(result, f, ensure_ascii=False, indent=2)
-    print(f"\n诊断结果已保存到: {output_file}")
-
-
-if __name__ == "__main__":
-    main()

scripts/ci/ci_repeated_failure_detector.py

@@ -1,417 +0,0 @@
-#!/usr/bin/env python3
-"""
-CI重复失败检测脚本
-- 扫描最近N天的CI失败
-- 按job名称分组统计失败率
-- 识别高失败率job（系统性故障）
-- 飞书通知告警
-"""
-
-import json
-import os
-import sys
-import time
-import urllib.error
-import urllib.request
-from collections import defaultdict
-from datetime import datetime, timedelta, timezone
-
-
-def get_env(name, default=None, required=False):
-    val = os.environ.get(name, default)
-    if required and not val:
-        print(f"❌ 缺少环境变量: {name}")
-        sys.exit(1)
-    return val
-
-
-GITEA_URL = get_env("GITEA_URL", "https://git.xiaoxiajianji.com")
-GITEA_TOKEN = get_env("GITEA_API_TOKEN", required=False) or get_env("GITHUB_TOKEN", "")
-REPO = get_env("GITEA_REPO", "xiaoxia/xiaoxia-saas")
-DAYS = int(get_env("FAIL_CHECK_DAYS", "7"))
-FAIL_THRESHOLD = int(get_env("FAIL_THRESHOLD", 3))  # 失败次数阈值
-FAIL_RATE_THRESHOLD = float(get_env("FAIL_RATE_THRESHOLD", "30"))  # 失败率阈值%
-CONSECUTIVE_FAIL_THRESHOLD = int(get_env("CONSECUTIVE_FAIL_THRESHOLD", "3"))  # 连续失败阈值
-WEBHOOK = get_env("CI_NOTIFY_WEBHOOK", "")
-
-
-def api_get(path):
-    """调用Gitea API"""
-    url = f"{GITEA_URL}/api/v1{path}"
-    req = urllib.request.Request(url)
-    if GITEA_TOKEN:
-        req.add_header("Authorization", f"token {GITEA_TOKEN}")
-    try:
-        with urllib.request.urlopen(req, timeout=30) as resp:
-            return json.loads(resp.read())
-    except urllib.error.HTTPError as e:
-        print(f"  HTTP {e.code}: {path}")
-        return None
-    except Exception as e:
-        print(f"  错误: {e}")
-        return None
-
-
-def fetch_recent_runs(days=7, per_page=50, max_pages=10):
-    """获取最近N天的runs"""
-    since = (datetime.now(timezone.utc) - timedelta(days=days)).isoformat()
-    all_runs = []
-
-    for page in range(1, max_pages + 1):
-        path = f"/repos/{REPO}/actions/runs?page={page}&limit={per_page}"
-        data = api_get(path)
-        if not data:
-            break
-
-        runs = data.get("workflow_runs", data.get("runs", []))
-        if not runs:
-            break
-
-        # 检查时间范围（Gitea用started_at，格式2026-07-22T10:58:10+08:00）
-        oldest = None
-        for r in runs:
-            started = r.get("started_at", r.get("created_at", ""))
-            if started and started >= since:
-                all_runs.append(r)
-            else:
-                oldest = started
-
-        if oldest and oldest < since:
-            break
-
-        if len(runs) < per_page:
-            break
-
-    return all_runs
-
-
-def fetch_run_jobs(run_id):
-    """获取run的所有jobs"""
-    path = f"/repos/{REPO}/actions/runs/{run_id}/jobs"
-    data = api_get(path)
-    if not data:
-        return []
-    return data.get("jobs", [])
-
-
-def analyze_failures(runs):
-    """
-    分析失败情况
-
-    返回:
-    - job_stats: {job_name: {total, success, failure, skipped, failure_rate, failures: [...]}}
-    - consecutive_failures: {job_name: current_streak, max_streak, last_status}
-    """
-    job_stats = defaultdict(
-        lambda: {
-            "total": 0,
-            "success": 0,
-            "failure": 0,
-            "error": 0,
-            "skipped": 0,
-            "cancelled": 0,
-            "failures": [],
-        }
-    )
-
-    # 按时间正序排列（旧→新）用于连续失败计算
-    sorted_runs = sorted(runs, key=lambda r: r.get("started_at", r.get("created_at", "")))
-
-    # 连续失败跟踪 {job_name: streak}
-    consecutive = defaultdict(lambda: {"current": 0, "max": 0, "last_run": None})
-
-    for run in sorted_runs:
-        run_id = run.get("id")
-        run_status = run.get("status", "")
-        run_conclusion = run.get("conclusion", "")
-        run_started = run.get("started_at", run.get("created_at", ""))
-        event = run.get("event", "")
-
-        # 只统计pull_request和push事件的CI
-        if event not in ("pull_request", "push"):
-            continue
-
-        jobs = fetch_run_jobs(run_id)
-
-        for job in jobs:
-            name = job.get("name", "")
-            status = job.get("status", "")
-            conclusion = job.get("conclusion", "")
-
-            # 跳过非CI核心job（如AI Code Review、Preview等）
-            skip_prefixes = ("AI Code Review", "Preview", "PR Automation", "Auto")
-            if any(name.startswith(p) for p in skip_prefixes):
-                continue
-
-            stats = job_stats[name]
-            stats["total"] += 1
-
-            if conclusion == "success":
-                stats["success"] += 1
-                consecutive[name]["current"] = 0
-            elif conclusion == "failure":
-                stats["failure"] += 1
-                stats["failures"].append(
-                    {
-                        "run_id": run_id,
-                        "time": run_started,
-                        "event": event,
-                    }
-                )
-                consecutive[name]["current"] += 1
-                if consecutive[name]["current"] > consecutive[name]["max"]:
-                    consecutive[name]["max"] = consecutive[name]["current"]
-                consecutive[name]["last_run"] = run_id
-            elif conclusion == "error":
-                stats["error"] += 1
-                # error也算失败的一种
-                consecutive[name]["current"] += 1
-                if consecutive[name]["current"] > consecutive[name]["max"]:
-                    consecutive[name]["max"] = consecutive[name]["current"]
-            elif conclusion == "skipped":
-                stats["skipped"] += 1
-                # skipped不算也不打断连续失败
-            elif conclusion == "cancelled":
-                stats["cancelled"] += 1
-                # cancelled不算失败也不打断
-
-    # 计算失败率
-    for name, stats in job_stats.items():
-        total_actual = stats["total"] - stats["skipped"] - stats["cancelled"]
-        if total_actual > 0:
-            stats["failure_rate"] = round((stats["failure"] + stats["error"]) / total_actual * 100, 1)
-        else:
-            stats["failure_rate"] = 0.0
-
-    return dict(job_stats), dict(consecutive)
-
-
-def find_high_failures(job_stats, consecutive):
-    """
-    找出高风险job
-
-    告警级别：
-    - critical: 连续失败 >= CONSECUTIVE_FAIL_THRESHOLD，或 失败率>=50%且失败次数>=5
-    - warning: 失败率>=FAIL_RATE_THRESHOLD且失败次数>=FAIL_THRESHOLD
-    - info: 失败次数>=2
-    """
-    critical = []
-    warning = []
-    info = []
-
-    for name, stats in job_stats.items():
-        fail_count = stats["failure"] + stats["error"]
-        rate = stats["failure_rate"]
-        streak = consecutive.get(name, {}).get("current", 0)
-        max_streak = consecutive.get(name, {}).get("max", 0)
-
-        issue = {
-            "name": name,
-            "fail_count": fail_count,
-            "total": stats["total"],
-            "failure_rate": rate,
-            "current_streak": streak,
-            "max_streak": max_streak,
-            "recent_failures": stats["failures"][-5:],  # 最近5次
-        }
-
-        if streak >= CONSECUTIVE_FAIL_THRESHOLD or (rate >= 50 and fail_count >= 5):
-            critical.append(issue)
-        elif rate >= FAIL_RATE_THRESHOLD and fail_count >= FAIL_THRESHOLD:
-            warning.append(issue)
-        elif fail_count >= 2:
-            info.append(issue)
-
-    # 按失败次数倒序
-    critical.sort(key=lambda x: x["fail_count"], reverse=True)
-    warning.sort(key=lambda x: x["fail_count"], reverse=True)
-    info.sort(key=lambda x: x["fail_count"], reverse=True)
-
-    return critical, warning, info
-
-
-def generate_report(critical, warning, info, days, total_runs):
-    """生成Markdown报告"""
-    lines = []
-    lines.append("# CI重复失败检测报告")
-    lines.append("")
-    lines.append(f"**统计周期**: 最近{days}天")
-    lines.append(f"**扫描Runs**: {total_runs}个")
-    lines.append(f"**生成时间**: {datetime.now(timezone.utc).strftime('%Y-%m-%d %H:%M UTC')}")
-    lines.append("")
-
-    lines.append(f"## 概览")
-    lines.append("")
-    lines.append(f"| 级别 | 数量 |")
-    lines.append(f"|------|------|")
-    lines.append(f"| 🔴 严重 (连续失败≥{CONSECUTIVE_FAIL_THRESHOLD}次 或 失败率≥50%) | {len(critical)} |")
-    lines.append(f"| 🟡 警告 (失败率≥{FAIL_RATE_THRESHOLD}% 且 失败≥{FAIL_THRESHOLD}次) | {len(warning)} |")
-    lines.append(f"| 🔵 关注 (失败≥2次) | {len(info)} |")
-    lines.append("")
-
-    if critical:
-        lines.append("## 🔴 严重问题")
-        lines.append("")
-        for item in critical:
-            lines.append(f"### {item['name']}")
-            lines.append("")
-            lines.append(f"- 失败次数: **{item['fail_count']}** / {item['total']} 次运行")
-            lines.append(f"- 失败率: **{item['failure_rate']}%**")
-            lines.append(f"- 当前连续失败: **{item['current_streak']}** 次 (历史最高: {item['max_streak']} 次)")
-            lines.append("")
-            if item["recent_failures"]:
-                lines.append("最近失败:")
-                lines.append("")
-                for f in item["recent_failures"]:
-                    lines.append(f"- [{f['time'][:16]}] run #{f['run_id']} ({f['event']})")
-                lines.append("")
-
-    if warning:
-        lines.append("## 🟡 警告")
-        lines.append("")
-        for item in warning:
-            lines.append(
-                f"- **{item['name']}**: {item['fail_count']}次失败 / {item['total']}次运行 ({item['failure_rate']}%)"
-            )
-        lines.append("")
-
-    if info:
-        lines.append("## 🔵 关注列表")
-        lines.append("")
-        lines.append("| Job名称 | 失败次数 | 总次数 | 失败率 | 当前连续 |")
-        lines.append("|---------|----------|--------|--------|----------|")
-        for item in info[:20]:  # 最多显示20个
-            lines.append(
-                f"| {item['name']} | {item['fail_count']} | {item['total']} | {item['failure_rate']}% | {item['current_streak']} |"
-            )
-        lines.append("")
-
-    return "\n".join(lines)
-
-
-def send_feishu_notification(critical, warning, info, days):
-    """发送飞书通知"""
-    if not WEBHOOK:
-        print("  ⚠️  未配置WEBHOOK，跳过飞书通知")
-        return False
-
-    total_issues = len(critical) + len(warning) + len(info)
-    if total_issues == 0:
-        print("  ✅ 无异常，不发送通知")
-        return True
-
-    level = "🔴 严重告警" if critical else "🟡 警告" if warning else "🔵 关注"
-
-    title = f"CI重复失败检测 - {level}"
-    text = f"统计周期: 最近{days}天\n\n"
-
-    if critical:
-        text += "【严重问题】\n"
-        for item in critical[:5]:
-            text += f"• {item['name']}\n"
-            text += f"  失败 {item['fail_count']}/{item['total']} ({item['failure_rate']}%) 连续{item['current_streak']}次\n"
-        if len(critical) > 5:
-            text += f"  ...还有{len(critical)-5}个\n"
-        text += "\n"
-
-    if warning:
-        text += "【警告】\n"
-        for item in warning[:5]:
-            text += f"• {item['name']}: {item['fail_count']}次失败 ({item['failure_rate']}%)\n"
-        if len(warning) > 5:
-            text += f"  ...还有{len(warning)-5}个\n"
-        text += "\n"
-
-    if info and not critical and not warning:
-        text += "【关注列表】\n"
-        for item in info[:10]:
-            text += f"• {item['name']}: {item['fail_count']}次失败\n"
-        text += "\n"
-
-    text += f"共发现 {total_issues} 个异常job"
-
-    payload = {"msg_type": "text", "content": {"text": f"{title}\n\n{text}"}}
-
-    data = json.dumps(payload).encode()
-    req = urllib.request.Request(WEBHOOK, data=data, headers={"Content-Type": "application/json"})
-
-    try:
-        with urllib.request.urlopen(req, timeout=10) as resp:
-            result = json.loads(resp.read())
-            if result.get("code") == 0 or result.get("StatusCode") == 0:
-                print("  ✅ 飞书通知已发送")
-                return True
-            else:
-                print(f"  ⚠️  飞书返回: {result}")
-                return False
-    except Exception as e:
-        print(f"  ❌ 飞书通知失败: {e}")
-        return False
-
-
-def main():
-    print(f"=== CI重复失败检测 ===")
-    print(f"统计周期: 最近{DAYS}天")
-    print(f"仓库: {REPO}")
-    print()
-
-    print("1. 获取最近的Runs...")
-    runs = fetch_recent_runs(days=DAYS)
-    print(f"   找到 {len(runs)} 个runs")
-
-    if not runs:
-        print("⚠️  没有找到runs，退出")
-        return
-
-    print()
-    print("2. 分析job失败情况（可能需要点时间）...")
-    job_stats, consecutive = analyze_failures(runs)
-    print(f"   共统计 {len(job_stats)} 个job")
-
-    print()
-    print("3. 识别高风险job...")
-    critical, warning, info = find_high_failures(job_stats, consecutive)
-    print(f"   🔴 严重: {len(critical)}")
-    print(f"   🟡 警告: {len(warning)}")
-    print(f"   🔵 关注: {len(info)}")
-
-    print()
-    print("4. 生成报告...")
-    report = generate_report(critical, warning, info, DAYS, len(runs))
-
-    # 保存报告
-    report_path = os.environ.get("REPORT_PATH", f"/tmp/ci_failure_report_{int(time.time())}.md")
-    with open(report_path, "w") as f:
-        f.write(report)
-    print(f"   报告已保存: {report_path}")
-
-    # 打印摘要
-    print()
-    print("=== 摘要 ===")
-    if critical:
-        print("🔴 严重问题:")
-        for item in critical[:5]:
-            print(
-                f"  {item['name']}: {item['fail_count']}次失败, {item['failure_rate']}%, 连续{item['current_streak']}次"
-            )
-    if warning:
-        print("🟡 警告:")
-        for item in warning[:5]:
-            print(f"  {item['name']}: {item['fail_count']}次失败, {item['failure_rate']}%")
-
-    print()
-    print("5. 发送飞书通知...")
-    send_feishu_notification(critical, warning, info, DAYS)
-
-    print()
-    print("✅ 检测完成")
-
-    # 有严重问题时退出码非零，方便workflow标记
-    if critical:
-        sys.exit(2)
-    elif warning:
-        sys.exit(1)
-
-
-if __name__ == "__main__":
-    main()

scripts/ci/run_integration_tests.sh

@@ -1,7 +1,6 @@
 #!/bin/bash
 # CI Integration Tests Job 主脚本
 # 包含：依赖安装、ffmpeg安装、Redis启动、PG启动、迁移、测试、清理、覆盖率
-# 支持 pytest-xdist 并行执行：每个 worker 使用独立数据库，预期加速 2-4 倍
 set -eu
 
 echo "=== CI Integration Tests 开始 ==="
@@ -29,13 +28,12 @@ for i in 1 2 3; do
   sleep 5
 done
 for i in 1 2 3; do
-  python3 -m pip install -q pytest-rerunfailures pytest-xdist && break
-  echo "pip install pytest-rerunfailures/pytest-xdist 失败，重试 $i/3..."
+  python3 -m pip install -q pytest-rerunfailures && break
+  echo "pip install pytest-rerunfailures 失败，重试 $i/3..."
   [ $i -eq 3 ] && exit 1
   sleep 5
 done
 pytest --version
-echo "pytest-xdist: $(python3 -c "import xdist; print(xdist.__version__)" 2>/dev/null || echo 'not installed')"
 
 # --- 安装 ffmpeg ---
 echo ""
@@ -189,8 +187,8 @@ if [ "$USE_SHARED_PG" = "true" ]; then
   echo "等待共享PG连接就绪..."
   wait_tcp_ready "$SHARED_PG_HOST" "$SHARED_PG_PORT" 5
 
-  # 创建主数据库（xdist 模式下各 worker 会创建自己的数据库，主库作为 fallback）
-  echo "创建主测试数据库: $CI_DB_NAME"
+  # 创建独立数据库
+  echo "创建测试数据库: $CI_DB_NAME"
   PGPASSWORD="$SHARED_PG_PASSWORD" python3 -c "
 import psycopg2
 conn = psycopg2.connect(host='$SHARED_PG_HOST', port=$SHARED_PG_PORT, user='$SHARED_PG_USER', password='$SHARED_PG_PASSWORD', dbname='postgres')
@@ -240,37 +238,30 @@ else
   echo "TCP connectivity to PostgreSQL confirmed on port $PG_PORT"
 fi
 
-# --- 执行迁移（主数据库，xdist worker 会各自创建自己的库并迁移） ---
+# --- 执行迁移 ---
 echo ""
-echo "=== 执行 Alembic 迁移（主数据库） ==="
+echo "=== 执行 Alembic 迁移 ==="
 PYTHONPATH="$PWD/apps/api:$PWD" python3 -m alembic upgrade head
 echo "✅ 迁移完成"
 
-# --- 运行集成测试（pytest-xdist 并行） ---
+# --- 运行集成测试 ---
 echo ""
-echo "=== 运行集成测试（pytest-xdist 并行模式） ==="
-echo "CPU 核数: $(nproc 2>/dev/null || echo 'unknown')"
-
-# 集成测试使用 pytest-xdist 并行加速（coverage 由单元测试负责，并行模式下 coverage 不稳定）
-# -n 2: 限制2个worker，避免DooD模式下读到宿主机全部核数导致OOM
-#   集成测试涉及ffmpeg编码+PG多库，内存开销大，2worker较稳妥
-#   （auto模式下读到宿主机4核8线程=8worker，OOM直接杀worker进程）
-# --dist loadfile: 同一测试文件分配到同一 worker（共享 fixture 更高效）
-# --maxfail=1: 遇到失败停止调度新测试（并行模式下等价于 -x）
-PYTHONPATH="$PWD/apps/api:$PWD" python3 -m pytest tests/integration \
-  -q --timeout=60 --maxfail=1 --reruns 2 --reruns-delay 1 \
-  -m "not performance" \
-  -n 2 --dist loadfile \
-  -p no:cacheprovider
-
+echo "=== 运行集成测试 ==="
+PYTHONPATH="$PWD/apps/api:$PWD" python3 -m coverage run \
+  --source=apps/api/app,packages \
+  --omit="*/migrations/*,*/tests/*,*/test_*.py,*/site-packages/*" \
+  --branch \
+  -m pytest tests/integration -q --timeout=60 -x --reruns 2 --reruns-delay 1 -m "not performance"
+python3 -m coverage report --show-missing
+python3 -m coverage xml -o coverage.xml
+python3 -m coverage report --fail-under=40 > /dev/null
 echo "✅ 集成测试通过"
 
-# --- API 性能基线测试（仅告警，串行执行） ---
+# --- API 性能基线测试（仅告警） ---
 echo ""
 echo "=== API 性能基线测试（仅告警） ==="
 set +e
 PERF_OUTPUT=$(mktemp)
-# 性能测试单独串行运行（不参与并行，避免资源竞争影响测量结果）
 PYTHONPATH="$PWD/apps/api:$PWD" python3 -m pytest tests/integration/test_api_performance.py \
   -v --timeout=120 -p no:cacheprovider 2>&1 | tee "$PERF_OUTPUT"
 echo ""
@@ -293,29 +284,14 @@ set -e
 echo ""
 echo "=== 清理 ==="
 if [ "$USE_SHARED_PG" = "true" ]; then
-  # 清理共享PG上的测试数据库（主库 + 可能残留的 worker 库）
-  echo "清理共享PG测试数据库..."
-
-  # 清理所有以 CI_DB_NAME 开头的数据库（主库 + worker 库）
+  # 清理共享PG上的测试数据库
+  echo "清理共享PG测试数据库: $CI_DB_NAME"
   PGPASSWORD="${SHARED_PG_PASSWORD}" python3 -c "
 import psycopg2
 conn = psycopg2.connect(host='${SHARED_PG_HOST}', port=${SHARED_PG_PORT}, user='${SHARED_PG_USER}', password='${SHARED_PG_PASSWORD}', dbname='postgres')
 conn.autocommit = True
 cur = conn.cursor()
-
-# 查找所有需要清理的数据库（主库 + worker 库）
-cur.execute(\"SELECT datname FROM pg_database WHERE datname LIKE '$CI_DB_NAME%'\")
-dbs = [row[0] for row in cur.fetchall()]
-
-for db in dbs:
-    try:
-        # 强制断开所有连接
-        cur.execute(f\"SELECT pg_terminate_backend(pid) FROM pg_stat_activity WHERE datname = '{db}' AND pid <> pg_backend_pid()\")
-        cur.execute(f'DROP DATABASE IF EXISTS \"{db}\" WITH (FORCE)')
-        print(f'  已清理: {db}')
-    except Exception as e:
-        print(f'  警告: 清理 {db} 失败: {e}')
-
+cur.execute(f'DROP DATABASE IF EXISTS \"$CI_DB_NAME\" WITH (FORCE)')
 cur.close()
 conn.close()
 " 2>/dev/null || echo "WARN: 数据库清理失败（可能已被清理）"

scripts/ci/step_timer_start.sh

@@ -2,5 +2,3 @@
 # CI 公共步骤：Job 开始计时
 echo "JOB_START_TIME=$(date +%s)" >> $GITHUB_ENV
 echo "Job started at $(date)"
-# trigger CI run for PR validation
-# trigger CI - worker dood fallback fix test

scripts/ci_notify_failure.py

@@ -1,57 +1,17 @@
 #!/usr/bin/env python3
-"""发送 CI 失败通知到飞书/项目群 webhook（增强版：带失败诊断）。
-
-诊断功能：自动分析失败原因，给出分类和修复建议。
-"""
+"""发送 CI 失败通知到飞书/项目群 webhook。"""
 
 import json
 import os
-import subprocess
 import sys
 import urllib.request
 
 
-def run_diagnosis() -> dict:
-    """运行失败诊断脚本，返回诊断结果"""
-    diag_script = os.path.join(os.path.dirname(os.path.abspath(__file__)), "ci/ci_failure_diagnosis.py")
-    if not os.path.exists(diag_script):
-        diag_script = "scripts/ci/ci_failure_diagnosis.py"
-
-    result = {
-        "category": "unknown",
-        "category_cn": "未知",
-        "severity": "medium",
-        "summary": "",
-        "error_lines": [],
-        "suggestions": [],
-        "auto_fixable": False,
-    }
-
-    try:
-        # 运行诊断脚本
-        env = os.environ.copy()
-        env["DIAGNOSIS_OUTPUT"] = "/tmp/ci_diagnosis_result.json"
-
-        proc = subprocess.run([sys.executable, diag_script], capture_output=True, text=True, timeout=30, env=env)
-
-        # 尝试读取结果文件
-        output_file = "/tmp/ci_diagnosis_result.json"
-        if os.path.exists(output_file):
-            with open(output_file) as f:
-                result = json.load(f)
-        elif proc.stdout:
-            # 从stdout解析
-            pass
-    except Exception as e:
-        print(f"诊断脚本执行失败: {e}", file=sys.stderr)
-
-    return result
-
-
 def main() -> int:
     webhook = os.environ.get("CI_NOTIFY_WEBHOOK", "")
     if not webhook:
         print("未配置 CI_NOTIFY_WEBHOOK，跳过通知")
+        print("如需启用，请在仓库 Settings -> Secrets and variables -> Actions 中添加 CI_NOTIFY_WEBHOOK")
         return 0
 
     failed_job = os.environ.get("FAILED_JOB", "Unknown Job")
@@ -61,86 +21,6 @@ def main() -> int:
     run_id = os.environ.get("GITHUB_RUN_ID", "unknown")
     repo = os.environ.get("GITHUB_REPOSITORY", "unknown")
     run_url = f"https://git.xiaoxiajianji.com/{repo}/actions/runs/{run_id}"
-    pr_number = os.environ.get("PR_NUMBER", "")
-
-    # 运行诊断
-    diagnosis = run_diagnosis()
-
-    # 构建卡片内容
-    severity_color = {"high": "red", "medium": "orange", "low": "blue"}
-    card_status = severity_color.get(diagnosis.get("severity", "medium"), "red")
-
-    # 标题
-    title = f"❌ CI失败 - {diagnosis.get('category_cn', '未知')}"
-
-    # 诊断部分
-    diag_lines = []
-    diag_lines.append(f"**任务**: {failed_job}")
-    diag_lines.append(f"**分类**: {diagnosis.get('category_cn', '未知')}")
-    if diagnosis.get("summary"):
-        diag_lines.append(f"**问题**: {diagnosis['summary']}")
-
-    # 错误行
-    error_lines = diagnosis.get("error_lines", [])
-    if error_lines:
-        diag_lines.append("")
-        diag_lines.append("**关键错误**:")
-        for err in error_lines[:3]:
-            if len(err) > 100:
-                err = err[:97] + "..."
-            diag_lines.append(f"`{err}`")
-
-    # 修复建议
-    suggestions = diagnosis.get("suggestions", [])
-    if suggestions:
-        diag_lines.append("")
-        diag_lines.append("**修复建议**:")
-        for i, s in enumerate(suggestions[:3], 1):
-            diag_lines.append(f"{i}. {s}")
-
-    if diagnosis.get("auto_fixable"):
-        diag_lines.append("")
-        diag_lines.append("💡 *可自动修复的问题，试试Rerun*")
-
-    # 基本信息
-    info_lines = [
-        f"**分支**: {branch}",
-        f"**提交**: `{commit}`",
-        f"**提交者**: {actor}",
-    ]
-    if pr_number:
-        info_lines.append(f"**PR**: #{pr_number}")
-
-    elements = [
-        {
-            "tag": "div",
-            "text": {
-                "tag": "lark_md",
-                "content": "\n".join(diag_lines),
-            },
-        },
-        {
-            "tag": "hr",
-        },
-        {
-            "tag": "div",
-            "text": {
-                "tag": "lark_md",
-                "content": "\n".join(info_lines),
-            },
-        },
-        {
-            "tag": "action",
-            "actions": [
-                {
-                    "tag": "button",
-                    "text": {"tag": "plain_text", "content": "查看失败日志"},
-                    "url": run_url,
-                    "type": "danger",
-                },
-            ],
-        },
-    ]
 
     payload = {
         "msg_type": "interactive",
@@ -148,11 +28,36 @@ def main() -> int:
             "header": {
                 "title": {
                     "tag": "plain_text",
-                    "content": title,
+                    "content": "❌ CI 构建失败",
                 },
-                "status": card_status,
+                "status": "red",
             },
-            "elements": elements,
+            "elements": [
+                {
+                    "tag": "div",
+                    "text": {
+                        "tag": "lark_md",
+                        "content": (
+                            f"**任务**: {failed_job}\n"
+                            f"**分支**: {branch}\n"
+                            f"**提交**: {commit}\n"
+                            f"**提交者**: {actor}\n"
+                            f"**Run ID**: {run_id}"
+                        ),
+                    },
+                },
+                {
+                    "tag": "action",
+                    "actions": [
+                        {
+                            "tag": "button",
+                            "text": {"tag": "plain_text", "content": "查看失败日志"},
+                            "url": run_url,
+                            "type": "danger",
+                        }
+                    ],
+                },
+            ],
         },
     }
 
@@ -166,7 +71,7 @@ def main() -> int:
     try:
         with urllib.request.urlopen(req, timeout=10) as resp:
             resp.read()
-        print("通知已发送（带诊断信息）")
+        print("通知已发送")
     except Exception as e:
         print(f"通知发送失败: {e}", file=sys.stderr)
         return 1
@@ -176,5 +81,3 @@ def main() -> int:
 
 if __name__ == "__main__":
     sys.exit(main())
-
-# trigger CI - bypass [ci skip] bug

tests/integration/conftest.py

@@ -2,167 +2,18 @@
 集成测试公共 fixtures
 
 提供性能测试相关的工具、fixture 和 marker。
-支持 pytest-xdist 并行执行：每个 worker 使用独立数据库，数据完全隔离。
 """
 
 from __future__ import annotations
 
 import os
-import sys
 import time
 from contextlib import contextmanager
 from dataclasses import dataclass, field
-from pathlib import Path
 from typing import Callable, Dict, List, Optional
 
 import pytest
 
-# ── xdist 并行数据库隔离 ──────────────────────────────────────────────────
-# 每个 xdist worker 进程创建独立的数据库并执行迁移，确保测试数据完全隔离
-# 通过 PYTEST_XDIST_WORKER 环境变量识别 worker（如 gw0, gw1, ...）
-
-_WORKER_DB_NAME: Optional[str] = None
-
-
-def _get_worker_id() -> Optional[str]:
-    """获取当前 xdist worker ID，非 worker 模式返回 None"""
-    return os.environ.get("PYTEST_XDIST_WORKER")
-
-
-def _parse_database_url(url: str) -> Dict[str, str]:
-    """
-    解析 DATABASE_URL，返回各组件。
-    支持 postgresql+psycopg://user:pass@host:port/dbname 格式
-    """
-    from urllib.parse import urlparse
-
-    parsed = urlparse(url)
-    return {
-        "driver": parsed.scheme,
-        "user": parsed.username or "",
-        "password": parsed.password or "",
-        "host": parsed.hostname or "",
-        "port": str(parsed.port or 5432),
-        "dbname": parsed.path.lstrip("/") or "",
-    }
-
-
-def _create_worker_database(worker_id: str) -> str:
-    """
-    为 xdist worker 创建独立数据库并执行迁移。
-    返回新的 DATABASE_URL。
-    """
-    base_url = os.environ.get(
-        "DATABASE_URL",
-        "postgresql+psycopg://postgres:postgres@localhost:5432/xiaoxia_saas",
-    )
-    db_info = _parse_database_url(base_url)
-
-    # 生成 worker 专属数据库名
-    base_db = db_info["dbname"]
-    worker_db = f"{base_db}_{worker_id}"
-    global _WORKER_DB_NAME
-    _WORKER_DB_NAME = worker_db
-
-    # 使用 psycopg 创建数据库（连接到 postgres 库）
-    try:
-        import psycopg
-
-        conn_str = (
-            f"host={db_info['host']} port={db_info['port']} "
-            f"user={db_info['user']} password={db_info['password']} "
-            f"dbname=postgres"
-        )
-        conn = psycopg.connect(conn_str, autocommit=True)
-        cur = conn.cursor()
-
-        # 先尝试删除（防止残留）
-        cur.execute(f'DROP DATABASE IF EXISTS "{worker_db}" WITH (FORCE)')
-
-        # 创建新数据库
-        cur.execute(f'CREATE DATABASE "{worker_db}"')
-        cur.close()
-        conn.close()
-        print(f"[xdist {worker_id}] ✅ 创建数据库: {worker_db}")
-    except ImportError:
-        print(f"[xdist {worker_id}] ⚠️  psycopg 未安装，跳过数据库创建")
-        return base_url
-    except Exception as e:
-        print(f"[xdist {worker_id}] ⚠️  创建数据库失败: {e}")
-        return base_url
-
-    # 构建新的 DATABASE_URL
-    new_url = (
-        f"{db_info['driver']}://{db_info['user']}:{db_info['password']}"
-        f"@{db_info['host']}:{db_info['port']}/{worker_db}"
-    )
-
-    # 执行 alembic 迁移
-    print(f"[xdist {worker_id}] 🔄 执行 Alembic 迁移...")
-    try:
-        ROOT = Path(__file__).resolve().parents[2]
-        api_path = str(ROOT / "apps" / "api")
-        if api_path not in sys.path:
-            sys.path.insert(0, api_path)
-        if str(ROOT) not in sys.path:
-            sys.path.insert(0, str(ROOT))
-
-        from alembic import command as alembic_command
-        from alembic.config import Config as AlembicConfig
-
-        alembic_cfg = AlembicConfig(str(ROOT / "alembic.ini"))
-        alembic_cfg.set_main_option("sqlalchemy.url", new_url)
-        # 兼容不同的脚本路径配置
-        alembic_cfg.set_main_option("script_location", str(ROOT / "alembic"))
-
-        # 临时设置环境变量供 alembic env.py 使用
-        os.environ["DATABASE_URL"] = new_url
-        alembic_command.upgrade(alembic_cfg, "head")
-        print(f"[xdist {worker_id}] ✅ 迁移完成")
-    except Exception as e:
-        print(f"[xdist {worker_id}] ❌ 迁移失败: {e}")
-        raise
-
-    return new_url
-
-
-def _cleanup_worker_database(worker_id: str):
-    """清理 xdist worker 的数据库"""
-    global _WORKER_DB_NAME
-    if not _WORKER_DB_NAME:
-        return
-
-    base_url = os.environ.get(
-        "DATABASE_URL",
-        "postgresql+psycopg://postgres:postgres@localhost:5432/xiaoxia_saas",
-    )
-    db_info = _parse_database_url(base_url)
-
-    try:
-        import psycopg
-
-        conn_str = (
-            f"host={db_info['host']} port={db_info['port']} "
-            f"user={db_info['user']} password={db_info['password']} "
-            f"dbname=postgres"
-        )
-        conn = psycopg.connect(conn_str, autocommit=True)
-        cur = conn.cursor()
-        # 强制断开所有连接后删除
-        cur.execute(
-            f"SELECT pg_terminate_backend(pid) FROM pg_stat_activity "
-            f"WHERE datname = '{_WORKER_DB_NAME}' AND pid <> pg_backend_pid()"
-        )
-        cur.execute(f'DROP DATABASE IF EXISTS "{_WORKER_DB_NAME}" WITH (FORCE)')
-        cur.close()
-        conn.close()
-        print(f"[xdist {worker_id}] 🧹 已清理数据库: {_WORKER_DB_NAME}")
-    except Exception as e:
-        print(f"[xdist {worker_id}] ⚠️  清理数据库失败: {e}")
-    finally:
-        _WORKER_DB_NAME = None
-
-
 # ── 性能阈值配置 ──────────────────────────────────────────────────────────
 PERF_THRESHOLDS: Dict[str, int] = {
     "core": 500,  # 核心接口：500ms
@@ -332,41 +183,16 @@ class PerfAssert:
         return "\n".join(lines)
 
 
-# ── pytest hooks ──────────────────────────────────────────────────────────
+# ── pytest fixtures ──────────────────────────────────────────────────────
 
 
 def pytest_configure(config):
-    """
-    pytest 配置钩子。
-
-    - 注册自定义 marker
-    - xdist worker 模式下：创建独立数据库 + 执行迁移
-    """
-    # 注册自定义 marker
+    """注册自定义 marker"""
     config.addinivalue_line("markers", "performance: 标记为性能测试（可通过 -m 'not performance' 跳过）")
     config.addinivalue_line("markers", "perf_core: 核心接口性能测试（阈值 500ms）")
     config.addinivalue_line("markers", "perf_normal: 普通接口性能测试（阈值 1000ms）")
     config.addinivalue_line("markers", "perf_heavy: 重操作接口性能测试（阈值 3000ms）")
 
-    # xdist worker 模式：创建独立数据库并执行迁移
-    worker_id = _get_worker_id()
-    if worker_id:
-        # 只有当 USE_IN_MEMORY_DB 不为 true 时才创建独立数据库
-        use_in_memory = os.environ.get("USE_IN_MEMORY_DB", "true").lower() == "true"
-        if not use_in_memory:
-            print(f"[xdist {worker_id}] 🚀 worker 启动，准备独立数据库...")
-            new_db_url = _create_worker_database(worker_id)
-            os.environ["DATABASE_URL"] = new_db_url
-        else:
-            print(f"[xdist {worker_id}] ℹ️  USE_IN_MEMORY_DB=true，跳过 worker 数据库创建")
-
-
-def pytest_unconfigure(config):
-    """pytest 结束钩子：清理 xdist worker 数据库"""
-    worker_id = _get_worker_id()
-    if worker_id and _WORKER_DB_NAME:
-        _cleanup_worker_database(worker_id)
-
 
 def pytest_collection_modifyitems(config, items):
     """根据环境变量自动跳过性能测试"""
@@ -377,9 +203,6 @@ def pytest_collection_modifyitems(config, items):
                 item.add_marker(skip_perf)
 
 
-# ── pytest fixtures ──────────────────────────────────────────────────────
-
-
 @pytest.fixture
 def perf_assert():
     """