wooo/awoooi
1
1
Fork 0
Code Issues 2 Pull Requests 1 Actions 2 Packages Projects Releases Wiki Activity

fix(agent99): terminalize unscoped cold-start recovery

Browse Source
This commit is contained in:
Your Name
2026-07-22 14:17:58 +08:00
parent 3a6f8badb7
commit 9fce2cd00f
8 changed files with 466 additions and 64 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
docs/operations/sre-ai-agent-conversation-commitments.snapshot.json
View File

@@ -49,7 +49,7 @@
{"id":"AIA-CONV-030","category":"telegram","status":"source_implemented_runtime_pending","title":"對話可啟動受控調查/修復並建立 Codex 開發 work item","linked_work_items":["AIA-SRE-001","AIA-SRE-015","AIA-SRE-017"],"terminal_condition":"討論結果轉為去重 work item；production action 仍走 typed executor/verifier。","source_evidence":["verified controlled_action_request ingress gate","exact canonical asset/domain resolution or asset_identity_unresolved drift item","fingerprint-deduplicated durable internal work-item receipt","deterministic no-provider no-runtime-mutation Telegram response"]},
{"id":"AIA-CONV-031","category":"telegram","status":"source_implemented_runtime_pending","title":"同 fingerprint 去重、聚合、抑噪及 resolved/recovered 更新","linked_work_items":["AIA-SRE-017"],"terminal_condition":"同事件只維護 canonical lifecycle，恢復卡有 destination-bound provider acknowledgement。","source_evidence":["existing firing fingerprint convergence and parent-child alert storm aggregation","single-row durable resolved lifecycle reservation with bounded retry","duplicate resolved webhook suppression before incident or Telegram side effects","recovery card accepted only with destination-bound provider acknowledgement"]},
{"id":"AIA-CONV-032","category":"named_incident","status":"in_progress","title":"INC-20260711-11C751 cold-start-gate 補 PlayBook、transport、rollback 與 verifier","linked_work_items":["AIA-SRE-004","AIA-SRE-006","AIA-SRE-015"],"terminal_condition":"namespace identity 修正後完成一次 bounded same-run repair 或明確 no-write terminal。"},
{"id":"AIA-CONV-032","category":"named_incident","status":"source_implemented_runtime_pending","title":"INC-20260711-11C751 cold-start-gate 補 PlayBook、transport、rollback 與 verifier","linked_work_items":["AIA-SRE-004","AIA-SRE-006","AIA-SRE-015"],"terminal_condition":"namespace identity 修正後完成一次 bounded same-run repair 或明確 no-write terminal。","source_evidence":["source namespace separated from normalized host_recovery execution identity","exact inventory scope absence blocks Recover before single-flight claim or Agent99 transport","durable alert operation receipt records the explicit no-write terminal","consumer-validated source no-write verifier receipt with runtime closure false"]},
{"id":"AIA-CONV-033","category":"named_incident","status":"in_progress","title":"INC-20260711-D037E5 修復 stale candidate 阻擋、自動排隊與 durable closure","linked_work_items":["AIA-SRE-007","AIA-SRE-015"],"terminal_condition":"新候選可 claim，apply 不重複，verifier、Telegram 與 learning receipts 同 run 完成。"},
{"id":"AIA-CONV-034","category":"named_incident","status":"in_progress","title":"AWOOOI CPU 高負載與 P99 上升完成 RCA、bounded repair 與 verifier","linked_work_items":["AIA-SRE-004","AIA-SRE-006","AIA-SRE-010"],"terminal_condition":"相關 metrics/logs/changes 對齊 canonical asset，修復後 latency 與資源 verifier 關閉。"},
{"id":"AIA-CONV-035","category":"named_incident","status":"in_progress","title":"Backup/restore escrow 補 freshness、escrow metadata、restore drill 與 DR scorecard","linked_work_items":["AIA-SRE-011","AIA-SRE-016"],"terminal_condition":"read-only evidence 齊全；缺 escrow/restore receipt 時保持 blocked 而非假綠。"},
@@ -98,8 +98,8 @@
"active_or_completed_commitments": 68,
"by_status": {
"analysis_or_governance_complete": 6,
"source_implemented_runtime_pending": 29,
"in_progress": 31,
"source_implemented_runtime_pending": 30,
"in_progress": 30,
"not_started_or_no_current_evidence": 2,
"superseded": 2
},