Site health #24
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| name: Site health | |
| # 定期访问已发布站点的关键 URL，确认它们仍然可达。 | |
| # | |
| # 背景：/AgentGuide/interview/ 曾在 2026-02-26 上线后 404，直到 2026-03-09 | |
| # 才由外部贡献者提 PR #36 修复，暴露窗口约 11 天。仓库内的其它校验都是对 | |
| # 源文件做断言，没有任何东西会去访问真实站点。 | |
| # | |
| # 说明： | |
| # - 只检查 HTTP 状态码。页面返回 200 但内容损坏，这里检测不到。 | |
| # - GitHub 的定时任务不保证准时，高峰期会延迟甚至跳过；6 小时是名义间隔。 | |
| # - 定时工作流的失败通知发给最后修改过下方 cron 表达式的用户。 | |
| on: | |
| schedule: | |
| - cron: '0 */6 * * *' | |
| workflow_dispatch: | |
| permissions: | |
| contents: read | |
| jobs: | |
| probe: | |
| name: Probe published URLs | |
| runs-on: ubuntu-latest | |
| steps: | |
| - name: Probe published site | |
| run: | | |
| set -uo pipefail | |
| BASE="https://adongwanai.github.io/AgentGuide" | |
| failed=0 | |
| { | |
| echo "### 站点探测结果" | |
| echo | |
| echo "| 结果 | 状态码 | 路径 | 覆盖 |" | |
| echo "|:---:|:---:|:---|:---|" | |
| } >> "$GITHUB_STEP_SUMMARY" | |
| probe() { | |
| path="$1"; expected="$2"; label="$3" | |
| # --retry 只对网络错误和 5xx/408/429 生效，404 不会被重试 | |
| # curl 在连接失败时自身会输出 000，因此这里不再追加，只兜底空值 | |
| code="$(curl -sS -I -o /dev/null -w '%{http_code}' \ | |
| --retry 2 --retry-delay 5 --max-time 30 \ | |
| "$BASE$path" 2>/dev/null)" || true | |
| code="${code:-000}" | |
| if [ "$code" = "$expected" ]; then | |
| printf ' ok %-4s %-38s %s\n' "$code" "$path" "$label" | |
| printf '| ✅ | %s | `%s` | %s |\n' "$code" "$path" "$label" >> "$GITHUB_STEP_SUMMARY" | |
| else | |
| printf ' FAIL %-4s %-38s %s (期望 %s)\n' "$code" "$path" "$label" "$expected" | |
| printf '| ❌ | %s（期望 %s） | `%s` | %s |\n' "$code" "$expected" "$path" "$label" >> "$GITHUB_STEP_SUMMARY" | |
| failed=1 | |
| fi | |
| } | |
| echo "探测 $BASE" | |
| echo | |
| # 每个探测点对应 deploy-pages.yml 中一个独立的产物拷贝步骤， | |
| # 任一步骤失败时其余步骤仍会成功，所以必须逐个覆盖。 | |
| probe "/" 200 "首页 / Pages 服务本身" | |
| probe "/assets/site.js" 200 "cp -r assets/." | |
| probe "/data/resources.json" 200 "cp data/resources.json（首页筛选数据源）" | |
| probe "/research/" 200 "ai-research-ebook 构建与拷贝" | |
| probe "/research/docs/intro/01-overview/" 200 "research 的 BASE_PATH 与深层路由" | |
| probe "/interview/" 200 "InterviewGuide 构建与拷贝" | |
| probe "/interview/questions/q-1/" 200 "interview 的 BASE_PATH 与深层路由" | |
| # 反向断言：确认不存在的路径确实返回 404。 | |
| # 若某天引入了 catch-all 兜底规则导致所有路径都返回 200， | |
| # 上面全部探测会集体失去意义，这一条用于兜底自检。 | |
| probe "/__nonexistent-health-probe__/" 404 "404 仍能正确返回（探测器自检）" | |
| echo | |
| if [ "$failed" -ne 0 ]; then | |
| echo "站点探测失败：上面标记 FAIL 的 URL 未返回期望状态码。" | |
| exit 1 | |
| fi | |
| echo "全部探测点正常。" |