feat(adr-081): Phase 1 感官縱深 — 8D 情報蒐集 + 執行後驗證
成品: - IncidentEvidence DB model(8D 感官 + pre/post 執行狀態) - EvidenceSnapshot dataclass(build_summary → LLM 上下文) - SanitizationService(Prompt Injection 0-tolerance,12 pattern) - MCPToolRegistry(動態工具登記,suggest_tools 不寫死告警類型) - PreDecisionInvestigator(8D 並行感官,P99 < 8s,Redis 30s 快取) - PostExecutionVerifier(warmup 10s → 後狀態評估 success/degraded/failed) - decision_manager + approval_execution 接線(feature flag 守衛) Gate 1 修復:D4/D5/D7/D8 補 sanitize_dict_values;移除裸 "error" failure signal 防 error_rate key 誤判;evidence_snapshot rowcount 零行警告。 測試:130 passed(+111 新增) Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
@@ -270,6 +270,17 @@ class ApprovalExecutionService:
|
||||
)
|
||||
)
|
||||
|
||||
# ADR-081 Phase 1: 執行後驗證 (fire-and-forget)
|
||||
# PostExecutionVerifier 等待 K8s 收斂後抓取後狀態,補填 EvidenceSnapshot
|
||||
from src.core.feature_flags import aiops_flags
|
||||
if aiops_flags.is_sub_flag_enabled("AIOPS_P1_POST_EXECUTION_VERIFIER"):
|
||||
asyncio.create_task(
|
||||
self._run_post_execution_verify(
|
||||
approval=approval,
|
||||
action_taken=f"{operation_type.value}:{resource_name}",
|
||||
)
|
||||
)
|
||||
|
||||
# 2026-04-07 Claude Code: Sprint 4 B3 — 記錄人工批准處置類型
|
||||
try:
|
||||
anomaly_key = await self._get_anomaly_key_from_approval(approval)
|
||||
@@ -487,6 +498,63 @@ class ApprovalExecutionService:
|
||||
self._write_execution_result_to_km(approval, success, error_message)
|
||||
)
|
||||
|
||||
async def _run_post_execution_verify(
|
||||
self,
|
||||
approval: "ApprovalRequest",
|
||||
action_taken: str,
|
||||
) -> None:
|
||||
"""
|
||||
ADR-081 Phase 1: 執行後驗證 (fire-and-forget 包裝)
|
||||
|
||||
1. 從 incident_id 查 Incident
|
||||
2. 從 incident_evidence 取最新 EvidenceSnapshot
|
||||
3. 呼叫 PostExecutionVerifier.verify() 補填後狀態 + 驗證結果
|
||||
4. 結果傳給 learning_service 更新 Playbook trust_score(Phase 3)
|
||||
"""
|
||||
if not approval.incident_id:
|
||||
return
|
||||
|
||||
try:
|
||||
from src.services.incident_service import get_incident_service
|
||||
from src.services.post_execution_verifier import get_post_execution_verifier
|
||||
from src.services.evidence_snapshot import EvidenceSnapshot
|
||||
|
||||
incident_svc = get_incident_service()
|
||||
incident = await incident_svc.get_incident(approval.incident_id)
|
||||
if incident is None:
|
||||
logger.warning(
|
||||
"post_verify_incident_not_found",
|
||||
approval_id=str(approval.id),
|
||||
incident_id=approval.incident_id,
|
||||
)
|
||||
return
|
||||
|
||||
# 取最新 EvidenceSnapshot(若 Phase 1 flag 有啟動才會有)
|
||||
snapshot = await EvidenceSnapshot.get_latest_snapshot(approval.incident_id)
|
||||
|
||||
verifier = get_post_execution_verifier()
|
||||
verification_result = await verifier.verify(
|
||||
incident=incident,
|
||||
snapshot=snapshot,
|
||||
action_taken=action_taken,
|
||||
)
|
||||
|
||||
logger.info(
|
||||
"post_verify_complete",
|
||||
approval_id=str(approval.id),
|
||||
incident_id=approval.incident_id,
|
||||
result=verification_result,
|
||||
action=action_taken,
|
||||
)
|
||||
|
||||
except Exception as _e:
|
||||
# 驗證失敗不影響執行結果
|
||||
logger.warning(
|
||||
"post_verify_failed",
|
||||
approval_id=str(approval.id),
|
||||
error=str(_e),
|
||||
)
|
||||
|
||||
async def _write_execution_result_to_km(
|
||||
self,
|
||||
approval: "ApprovalRequest",
|
||||
|
||||
Reference in New Issue
Block a user