chore(cd): deploy 80c36ba [skip ci]

【根因】INC-20260507-99ADF2 飛輪斷流，566+ stuck incidents（30秒漲 1）核心
原因：NO_ACTION 路徑 (approval_execution.py:251) 提前 return True，跳過
line 482-495 已有的 resolve_incident 呼叫，incident 永遠卡 INVESTIGATING。

【修法】
- approval_execution.py NO_ACTION 分支補 resolve_incident 呼叫 + 成功/失敗
  log，背景 log 加 path="no_action" 用於 prod 量化修法生效率（debugger
  全鏈分析 + critic 1st/2nd 審查必修 #1）。
- incident_service.py resolve_incident 在 line 1106 加 RESOLVED 冪等 guard，
  早於所有副作用（status mutation / Redis / DB / postmortem / KB / KM /
  disposition），順帶修 success path line 482-495 重觸 postmortem 的潛在
  老風險（critic 必修 #2）。

【遵守 Codex 5/6 設計（feedback_respect_codex_design_intent.md）】
- 不動 flywheel_stats_service.py / heartbeat_report_service.py /
  auto_repair_service.record_auto_repair() / metrics_repository UPPER(status)。
- resolve_incident 不寫 auto_repair_executions 表（Codex 5/6 source of
  truth），不污染 24h KPI 計算。

【Test 覆蓋】
- test_approval_execution_no_action.py：NO_ACTION → resolve 被呼叫一次 +
  resolve raise 時仍 return True（NO_ACTION 不能因 resolve 失敗退化成 False，
  否則污染 auto_execute KPI line 207-208 註解契約）。
- test_incident_service_resolve_idempotency.py：RESOLVED → return existing +
  save_to_working_memory 不被呼叫；not_found → return None。

【驗收條件（部署後 24h）】
1. grep `path="no_action"` 中 incident_resolved_after_no_action_execution
   數量 vs background_execution_noop 數量，1:1 才算修復成功。
2. awoooi_flywheel_incidents_stuck 從每 30 秒漲 1 變平緩。
3. SRE 群 24h 內若湧入 >20 份 NO_ACTION postmortem 觸發 follow-up 評估
   resolution_type="no_action" 跳過 postmortem（critic Minor #3 方案 B）。

Refs: INC-20260507-99ADF2, debugger root cause #1 (鏈 A), critic 1st 必修
#1 #2, critic 2nd 必修 #1 #2 #3

Co-Authored-By: Codex (aider) <noreply@anthropic.com>
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

apps/api/src/main.py

@@ -80,6 +80,7 @@ from src.core.feature_flags import aiops_flags  # ADR-080: AI 自主化飛輪 fe
 from src.core.http_client import close_all_http_clients, init_all_http_clients
 from src.core.logging import get_logger, setup_logging
 from src.core.redis_client import close_redis_pool, init_redis_pool
+from src.services.flywheel_stats_service import get_flywheel_stats_service
 from src.core.sse import get_publisher
 from src.core.telemetry import setup_telemetry, shutdown_telemetry
 
@@ -1005,10 +1006,17 @@ app.include_router(platform_v1.router, prefix="/api/v1/platform", tags=["AwoooP
 @app.get("/metrics", include_in_schema=False)
 async def prometheus_metrics() -> Response:
     """Prometheus metrics endpoint for alerting"""
-    return Response(
-        content=generate_latest(),
-        media_type=CONTENT_TYPE_LATEST,
-    )
+    content = generate_latest().decode("utf-8")
+    # 2026-05-07 ogt + Claude Sonnet 4.6 — INC-20260507-99ADF2 修復
+    # 飛輪指標（awoooi_flywheel_*）原本只在 /api/v1/stats/flywheel/metrics 暴露，
+    # 110 Prom awoooi-api job scrape /metrics 時抓不到 → FlywheelExecutionRateMissing 永久 firing
+    # 修法：在此串入飛輪指標，讓既有 scrape job 無需新增 job 即可抓到
+    try:
+        flywheel_metrics = await get_flywheel_stats_service().compute()
+        content += flywheel_metrics.to_prometheus_lines()
+    except Exception:
+        logger.warning("prometheus_metrics_flywheel_error")
+    return Response(content=content, media_type=CONTENT_TYPE_LATEST)
 
 
 # =============================================================================

apps/api/src/services/approval_execution.py

@@ -222,6 +222,7 @@ class ApprovalExecutionService:
                     approval_id=str(approval.id),
                     action=approval.action,
                     reason="NO_ACTION - 純調查/觀察類,不執行破壞動作",
+                    path="no_action",
                 )
                 # 標為 SUCCESS (觀察/調查本身就是成功完成)
                 await service.update_execution_status(approval.id, success=True)
@@ -248,6 +249,29 @@ class ApprovalExecutionService:
                     duration_ms=int((time.time() - _aol_started_ms) * 1000),
                     output={"reason": "NO_ACTION", "action": approval.action[:200]},
                 )
+                # F2 (2026-05-07 ogt + Claude Sonnet 4.6 + Codex):
+                # NO_ACTION 路徑要把 incident 推到 RESOLVED，否則 incident 永遠卡
+                # INVESTIGATING（FlywheelExecutionRateMissing 死告警 + 566 stuck 增長根因 #1）。
+                # resolve_incident 內已加 RESOLVED 冪等 guard，重複 resolve 會 idempotent
+                # return existing incident 不會重觸發 postmortem。
+                if approval.incident_id:
+                    try:
+                        from src.services.incident_service import get_incident_service
+
+                        await get_incident_service().resolve_incident(approval.incident_id)
+                        logger.info(
+                            "incident_resolved_after_no_action_execution",
+                            incident_id=approval.incident_id,
+                            approval_id=str(approval.id),
+                            path="no_action",
+                        )
+                    except Exception as _resolve_e:
+                        logger.warning(
+                            "incident_resolve_after_no_action_execution_failed",
+                            incident_id=approval.incident_id,
+                            approval_id=str(approval.id),
+                            error=str(_resolve_e),
+                        )
                 return True  # NO_ACTION 視為成功完成
 
             # 真解析失敗 (非 NO_ACTION)

apps/api/src/services/incident_service.py

@@ -1103,6 +1103,18 @@ class IncidentService:
             logger.warning("incident_not_found_for_resolve", incident_id=incident_id)
             return None
 
+        # 1.5 F2 (2026-05-07 ogt + Codex + Claude Sonnet 4.6) — 冪等保護：
+        # 已經 RESOLVED 的 incident 直接 return existing，避免後續所有副作用
+        # 重複觸發（postmortem / KB extract / KM convert / disposition / Telegram）。
+        # F2 NO_ACTION 路徑會頻繁呼叫 resolve_incident，必須擋在 status mutation 之前。
+        if incident.status == IncidentStatus.RESOLVED:
+            logger.info(
+                "incident_resolve_skipped_already_resolved",
+                incident_id=incident_id,
+                resolution_type=resolution_type,
+            )
+            return incident
+
         # 2. 更新狀態
         incident.status = IncidentStatus.RESOLVED
         incident.resolved_at = now_taipei()

apps/api/tests/test_approval_execution_no_action.py

@@ -0,0 +1,101 @@
+from types import SimpleNamespace
+
+from unittest.mock import AsyncMock
+
+import pytest
+
+from src.services.approval_execution import ApprovalExecutionService
+
+
+@pytest.mark.asyncio
+async def test_no_action_execution_resolves_incident_once(monkeypatch):
+    # Arrange
+    approval = SimpleNamespace(
+        id="approval-noaction-1",
+        action="NO_ACTION: 先做觀察",
+        incident_id="INC-TEST-001",
+    )
+    incident_service = SimpleNamespace(resolve_incident=AsyncMock())
+
+    monkeypatch.setattr(
+        "src.services.approval_execution.get_approval_service",
+        lambda: SimpleNamespace(update_execution_status=AsyncMock()),
+    )
+    monkeypatch.setattr(
+        "src.services.approval_execution.get_timeline_service",
+        lambda: SimpleNamespace(add_event=AsyncMock()),
+    )
+    monkeypatch.setattr(
+        "src.services.approval_execution.parse_operation_from_action",
+        lambda _: SimpleNamespace(
+            operation_type=None, resource_name=None, namespace=None
+        ),
+    )
+    monkeypatch.setattr(
+        "src.services.incident_service.get_incident_service",
+        lambda: incident_service,
+    )
+    monkeypatch.setattr(
+        "src.services.approval_execution.ApprovalExecutionService._push_execution_result_to_alert",
+        AsyncMock(return_value=None),
+    )
+    monkeypatch.setattr(
+        "src.services.approval_execution.ApprovalExecutionService._log_aol_completed",
+        AsyncMock(return_value=None),
+    )
+
+    # Act
+    result = await ApprovalExecutionService().execute_approved_action(approval)
+
+    # Assert
+    assert result is True
+    incident_service.resolve_incident.assert_awaited_once_with("INC-TEST-001")
+
+
+@pytest.mark.asyncio
+async def test_no_action_execution_returns_true_when_resolve_raises(monkeypatch):
+    """resolve_incident 拋錯時，NO_ACTION 仍須 return True。
+
+    契約：NO_ACTION 是「純觀察類」成功完成（line 207-208 註解明說避免污染
+    auto_execute KPI）。resolve 失敗只該 warning log，不該讓 result 退化成 False。
+    """
+    approval = SimpleNamespace(
+        id="approval-noaction-2",
+        action="NO_ACTION: 觀察",
+        incident_id="INC-TEST-002",
+    )
+    incident_service = SimpleNamespace(
+        resolve_incident=AsyncMock(side_effect=RuntimeError("redis down"))
+    )
+
+    monkeypatch.setattr(
+        "src.services.approval_execution.get_approval_service",
+        lambda: SimpleNamespace(update_execution_status=AsyncMock()),
+    )
+    monkeypatch.setattr(
+        "src.services.approval_execution.get_timeline_service",
+        lambda: SimpleNamespace(add_event=AsyncMock()),
+    )
+    monkeypatch.setattr(
+        "src.services.approval_execution.parse_operation_from_action",
+        lambda _: SimpleNamespace(
+            operation_type=None, resource_name=None, namespace=None
+        ),
+    )
+    monkeypatch.setattr(
+        "src.services.incident_service.get_incident_service",
+        lambda: incident_service,
+    )
+    monkeypatch.setattr(
+        "src.services.approval_execution.ApprovalExecutionService._push_execution_result_to_alert",
+        AsyncMock(return_value=None),
+    )
+    monkeypatch.setattr(
+        "src.services.approval_execution.ApprovalExecutionService._log_aol_completed",
+        AsyncMock(return_value=None),
+    )
+
+    result = await ApprovalExecutionService().execute_approved_action(approval)
+
+    assert result is True
+    incident_service.resolve_incident.assert_awaited_once_with("INC-TEST-002")

apps/api/tests/test_incident_service_resolve_idempotency.py

@@ -0,0 +1,64 @@
+"""
+test_incident_service_resolve_idempotency
+==========================================
+
+驗證 `IncidentService.resolve_incident` 對已經 RESOLVED 的 incident 必須 idempotent：
+- 直接 return existing incident
+- 不呼叫 save_to_working_memory（避免重複 Redis write）
+- 不呼叫 incident_repository.update_status（避免重複 DB write）
+- 不觸發 postmortem / KB extract / KM convert / disposition 副作用
+
+對應 critic 必修 #2 — 沒這個單測，未來有人挪 guard 位置會悄悄破功，
+重新放大「resolve_incident 重複觸發 postmortem 洗版」的舊風險。
+"""
+
+from types import SimpleNamespace
+from unittest.mock import AsyncMock
+
+import pytest
+
+from src.models.incident import IncidentStatus
+from src.services.incident_service import IncidentService
+
+
+@pytest.mark.asyncio
+async def test_resolve_incident_skips_when_already_resolved(monkeypatch):
+    """RESOLVED 的 incident 重複 resolve 應 idempotent。"""
+    fake_incident = SimpleNamespace(
+        incident_id="INC-IDEMPO-001",
+        status=IncidentStatus.RESOLVED,
+    )
+
+    svc = IncidentService()
+
+    # Mock 入口讀取 → 回 RESOLVED incident
+    monkeypatch.setattr(
+        svc, "get_from_working_memory", AsyncMock(return_value=fake_incident)
+    )
+    # Mock 後續所有副作用 → 用 AsyncMock 監看是否被呼叫
+    save_mock = AsyncMock(return_value=True)
+    monkeypatch.setattr(svc, "save_to_working_memory", save_mock)
+
+    result = await svc.resolve_incident("INC-IDEMPO-001")
+
+    # 應 return existing incident
+    assert result is fake_incident
+    # 副作用一律不能觸發（guard 必須早於 line 1117 的 status mutation）
+    save_mock.assert_not_called()
+
+
+@pytest.mark.asyncio
+async def test_resolve_incident_returns_none_when_not_found(monkeypatch):
+    """incident 不存在時 return None。確保 guard 不影響 not-found 路徑。"""
+    svc = IncidentService()
+
+    monkeypatch.setattr(
+        svc, "get_from_working_memory", AsyncMock(return_value=None)
+    )
+    save_mock = AsyncMock(return_value=True)
+    monkeypatch.setattr(svc, "save_to_working_memory", save_mock)
+
+    result = await svc.resolve_incident("INC-NOT-EXIST")
+
+    assert result is None
+    save_mock.assert_not_called()

k8s/awoooi-prod/kustomization.yaml

@@ -40,7 +40,7 @@ resources:
 images:
 - name: 192.168.0.110:5000/library/api:IMAGE_TAG_PLACEHOLDER
   newName: 192.168.0.110:5000/awoooi/api
-  newTag: 72d86ba70bc9db6035871e22c2d4a0410e1d7cc1
+  newTag: 80c36ba801b6eae5127afec9c25aa904e1dde270
 - name: 192.168.0.110:5000/library/web:IMAGE_TAG_PLACEHOLDER
   newName: 192.168.0.110:5000/awoooi/web
-  newTag: 72d86ba70bc9db6035871e22c2d4a0410e1d7cc1
+  newTag: 80c36ba801b6eae5127afec9c25aa904e1dde270