fix(agent99): dispatch actionable alerts with receipts

docs/operations/agent99-enterprise-ai-automation-work-items.snapshot.json

@@ -1,6 +1,6 @@
 {
   "schema_version": "agent99_enterprise_ai_automation_work_items_v1",
-  "generated_at": "2026-07-10T19:40:00+08:00",
+  "generated_at": "2026-07-11T10:27:00+08:00",
   "status": "in_progress_p0_alert_routing_grouping_dedupe",
   "scope_complete": false,
   "north_star": "Make Windows host 192.168.0.99 the policy-controlled on-prem AI execution node for all hosts, VMs, services, products, sites, tools, packages, alerts and recovery workflows, with external monitoring when host 99 is unavailable.",
@@ -19,30 +19,51 @@
   "current_p0": {
     "id": "AG99-P0-002",
     "title": "Alert taxonomy, routing, grouping and dedupe",
-    "reason": "Explicit event kind and suggested mode are not consistently authoritative, and forwarding before grouping can create wrong-domain or duplicate Agent99 work.",
-    "implementation_status": "local_package_and_contract_green_runtime_deploy_pending",
-    "next_action": "Finish the current Gitea source receipt, atomically deploy the tested Agent99 package to Windows 99 when an authenticated no-secret management channel is available, then replay backup, provider, reboot, PostgreSQL, host and K3s events against the runtime manifest."
+    "reason": "Direct Alertmanager events and actionable Telegram cards did not share one receipt and resolution contract; cold-start could be misclassified as Kubernetes and failed CI/CD cards were not mirrored to Agent99.",
+    "implementation_status": "source_candidate_and_contract_green_runtime_deploy_pending",
+    "next_action": "Finish the Gitea source receipt and single-writer CD coordination, atomically deploy the tested package to Windows 99, verify the runtime manifest, then replay exact ACTION REQUIRED cold-start and CI/CD failure cards plus the existing alert corpus."
   },
   "current_runtime_truth": {
-    "observed_at": "2026-07-10T15:04:27+08:00",
+    "observed_at": "2026-07-11T10:27:41+08:00",
     "awoooi_health_http_status": 200,
     "awoooi_health_state": "degraded",
+    "awoooi_runtime_build_commit_sha": "4a4b95e5135e8cf71d807a54700a962f5490088f",
+    "gitea_main_observed_sha": "f102616f0",
+    "gitea_main_deployed_source_sha": "4a4b95e5135e8cf71d807a54700a962f5490088f",
+    "runtime_work_items_generated_at": "2026-07-10T19:40:00+08:00",
+    "windows99_agent99_manifest_observed_at": "2026-07-11T09:51:00+08:00",
+    "windows99_agent99_source_revision": "e478ee128dfa046a9ddad872d3f60446e4477486e",
+    "windows99_agent99_runtime_matched": true,
+    "windows99_agent99_runtime_file_count": 14,
+    "windows99_agent99_runtime_mismatch_count": 0,
     "reboot_slo_status": "blocked_reboot_auto_recovery_slo_not_ready",
-    "reboot_slo_readiness_percent": 60,
-    "reboot_slo_primary_blocker": "reboot_event_required_host_unreachable",
+    "reboot_slo_readiness_percent": 33,
+    "reboot_slo_primary_blocker": "fresh_all_host_reboot_event_missing",
     "reboot_slo_active_blockers": [
       "all_required_hosts_not_in_10_minute_reboot_window",
-      "fresh_all_host_reboot_event_missing",
       "host_boot_observation_older_than_target_window",
-      "host_unreachable_after_reboot",
       "host_uptime_unknown",
-      "reboot_event_required_host_unreachable",
       "windows99_vmware_guest_power_not_ready",
       "edge_public_maintenance_live_config_drift",
       "windows99_vmware_vmx_missing",
       "windows99_update_no_auto_reboot_policy_not_ready",
       "privileged_edge_apply_channel_unavailable",
-      "windows99_vmware_autostart_config_not_ready"
+      "stockplatform_freshness_blocked",
+      "stockplatform_ingestion_blocked",
+      "post_start_blocked_not_zero",
+      "service_green_not_1",
+      "fresh_all_host_reboot_event_missing",
+      "windows99_vmware_autostart_config_not_ready",
+      "product_data_green_not_1",
+      "stockplatform_freshness_core_market_index_daily_tw_stale",
+      "stockplatform_freshness_core_price_daily_missing",
+      "stockplatform_freshness_core_chips_daily_missing",
+      "stockplatform_freshness_core_margin_short_daily_missing",
+      "stockplatform_freshness_ai_recommendations_stale",
+      "stockplatform_ingestion_core.price_daily_incomplete",
+      "stockplatform_ingestion_core.chips_daily_incomplete",
+      "stockplatform_ingestion_core.margin_short_daily_incomplete",
+      "stockplatform_ingestion_core.market_index_daily.tw_incomplete"
     ],
     "claims_forbidden": [
       "all_services_recovered_within_10_minutes",
@@ -304,22 +325,29 @@
       "title": "Alert taxonomy, routing, grouping and dedupe",
       "owner_lane": "awoooi_alert_router_agent99_ingress",
       "dependencies": ["AG99-P0-001"],
-      "scope_refs": ["all:alerts", "platform:AWOOOI", "host:99"],
-      "problem": "Source text contaminates routing and the bridge forwards before grouping, creating wrong modes and duplicate incidents.",
-      "trigger": "Any normalized alert or Telegram monitoring event entering the AWOOOI-to-Agent99 bridge.",
+      "scope_refs": ["all:alerts", "platform:AWOOOI", "host:99", "surface:Telegram"],
+      "problem": "Source text can contaminate routing, actionable Telegram cards can bypass Agent99, transport responses can be mistaken for accepted work, and wrong modes or duplicate incidents can result.",
+      "trigger": "Any normalized Alertmanager event, AI automation card, ACTION REQUIRED incident card, or failed AWOOOI CI/CD card entering the AWOOOI-to-Agent99 bridge.",
       "ai_role": "Normalized incident classifier",
       "action_mode": "controlled_apply",
-      "evidence_sources": ["Alertmanager payload", "suggestedMode", "event kind", "fingerprint", "five-minute grouping window"],
-      "controlled_actions": ["prioritize explicit kind and suggested mode", "group before bridge", "single-flight by incident fingerprint"],
+      "evidence_sources": ["Alertmanager payload", "Telegram outbound source envelope", "relay acceptance receipt", "suggestedMode", "event kind", "fingerprint", "five-minute grouping window"],
+      "controlled_actions": ["prioritize explicit kind and suggested mode", "group before bridge", "single-flight by incident fingerprint", "mirror actionable cards", "retain sanitized dispatch receipt", "propagate source-event resolution policy"],
       "prohibited_actions": ["route from source name alone", "emit duplicate same-state incidents"],
       "rollback": "Restore previous router while keeping dedupe store and queued events intact.",
-      "verifier": "Replay routing corpus and assert one active incident and expected mode per fingerprint.",
+      "verifier": "Replay exact Alertmanager and Telegram card corpus; assert one active incident, expected mode, accepted relay receipt, runtime evidence and legal source-event closure per fingerprint.",
       "callback": "Publish normalized route decision and dedupe reason to AWOOOI and AwoooP.",
-      "acceptance": "Known backup, provider, reboot, PostgreSQL, host and K3s samples route to the correct lane with no duplicate same-state dispatch.",
-      "completion_percent": 80,
-      "implementation_progress": "Local Agent99 contract and synthetic replay are green; Windows 99 runtime deployment and runtime replay remain pending.",
-      "evidence_refs": ["local Agent99 PowerShell contract 19/19", "local Agent99 synthetic replay 8/8", "isolated atomic deploy test passed"],
-      "next_action": "Deploy the tested package atomically to Windows 99, verify the runtime manifest, and rerun the routing/dedupe corpus through the live ingress."
+      "acceptance": "Known ACTION REQUIRED cold-start, CI/CD failure, backup, provider, reboot, PostgreSQL, host and K3s samples route to the correct lane with no duplicate same-state dispatch; every accepted dispatch has a durable receipt and runtime verifier outcome.",
+      "completion_percent": 85,
+      "completion_dimensions": {
+        "source_contract_percent": 100,
+        "test_contract_percent": 100,
+        "gitea_cd_percent": 0,
+        "windows99_runtime_percent": 0,
+        "live_replay_percent": 0
+      },
+      "implementation_progress": "Source dispatcher, receipt, taxonomy and resolution-policy contracts are green; Gitea CD, Windows 99 runtime deployment and exact live replay remain pending.",
+      "evidence_refs": ["131 focused tests passed after Gitea main rebase", "276 expanded all-Telegram and Agent99/RepairCandidate tests passed after Telegram single-owner integration", "repository static governance suite passed", "synthetic relay HTTP 202 ok=true inboxTriggered=true", "runtime work-items endpoint still serves 2026-07-10 snapshot"],
+      "next_action": "Finish the Gitea source receipt and single-writer coordination, deploy atomically to Windows 99, verify the runtime manifest, and rerun the exact actionable-card and Alertmanager corpus through live ingress."
     },
     {
       "id": "AG99-P0-003",

docs/workplans/2026-07-10-agent99-enterprise-ai-automation-master-plan.md

@@ -1,6 +1,6 @@
 # Agent99 全域 AI 自動化主計畫
 
-> 建立時間：2026-07-10 15:05 CST；最近更新：2026-07-10 19:40 CST
+> 建立時間：2026-07-10 15:05 CST；最近更新：2026-07-11 10:12 CST
 > 狀態：P0 執行中，尚未達成全域 AI 自動化
 > 機器可讀總帳：`docs/operations/agent99-enterprise-ai-automation-work-items.snapshot.json`
 > 歷史需求來源：`docs/workplans/2026-07-02-commander-inserted-requirements-priority-ledger.md`
@@ -92,7 +92,7 @@
 | 順序 | ID | 工作 | 狀態 | 驗收 |
 | --- | --- | --- | --- | --- |
 | 1 | AG99-P0-001 | 修正 `exitCode=0 => resolved` 假綠，對齊 99 runtime script 與 Gitea source SHA | **完成** | Production 99 outcome self-test 通過；runtime manifest `7/7` matched、drift `0`；operator replay 為 verified resolved，alert replay 缺 source-event resolved 時維持 verifying |
-| 2 | AG99-P0-002 | 修正 alert taxonomy、wrong-domain routing、bridge-before-grouping 與重複事件 | **進行中，current P0；本機 package/contract 綠，99 runtime 待部署** | `suggestedMode`/kind 優先於 source 字串；5 分鐘先 group；同 fingerprint 只建立一個 active incident |
+| 2 | AG99-P0-002 | 修正 alert taxonomy、wrong-domain routing、bridge-before-grouping、actionable Telegram card 漏接與重複事件 | **進行中，current P0；source/test 綠，99 runtime 待部署** | Alertmanager、AI card、ACTION REQUIRED、CI/CD failure 共用 dispatcher；`suggestedMode`/kind 優先；5 分鐘先 group；同 fingerprint 只建立一個 active incident；保留 acceptance receipt |
 | 3 | AG99-P0-003 | 強化 ingress：relay token 必填、Telegram chat+sender allowlist、idempotency/single consumer | 待辦 | 未認證 LAN request 拒絕；producer `auto_repair`/policy 被保留；所有 request 可追 trace |
 | 4 | AG99-P0-004 | 新增 EvidenceCollect / RepairCandidate / MCP lane | 待辦 | 缺 evidence 的事件會建立可查 candidate，不再錯送 Perf/AwoooRepair；`INC-*` 可從收件追到 candidate |
 
@@ -150,6 +150,18 @@
 - Velero missed weekly restore dry-run 已由 controller 自動 backfill 成功；20:17 Prometheus guard 修復 bind-mount inode drift 後，source/canonical/active/runtime 四方 SHA 一致、145 rules healthy，Backup/Gitea active alerts 為 0。
 - 以上只關閉 `AG99-P0-012/014` 的 Gitea/backup 子切片；全域 backup domain、Harbor/Registry/CD DR、99 Agent runtime 與 10 分鐘 reboot SLO 仍未完成。
 
+### 2026-07-11 09:35 Current P0 增量與 live truth
+
+- 已補上 `ACTION REQUIRED` 與失敗 `AWOOOI CI/CD` 卡片的 Agent99 dispatcher；cold-start 明確走 `host_recovery/Recover`，不再產生不存在的 Kubernetes deployment restart。
+- Direct Alertmanager 與 Telegram outbound mirror 都保留 sanitized relay receipt；只有 `accepted=true` 才算 dispatched，未接受會釋放 single-flight 並允許重試。
+- Source event closure 分成 `mode_verifier_can_resolve` 與 `external_source_receipt_required`；CI/CD failure 不會因 Agent99 本機命令成功而假結案。
+- Rebase 到 Telegram single-owner mainline 後 focused 回歸 `131 passed`，擴大到全部 Telegram 與 Agent99/RepairCandidate 為 `276 passed`；repo static governance checks 綠。99 relay 的 no-page smoke 回覆 HTTP 202、`ok=true`、`inboxTriggered=true`，只證明 transport/inbox 可達，不證明新 runtime 已部署。
+- Windows 99 live manifest 為 `sourceRevision=e478ee128dfa046a9ddad872d3f60446e4477486e`、`runtimeMatched=true`、`14/14` files matched、mismatch `0`。這證明現行舊 package 內部一致，也明確證明本次 P0 source 尚未部署到 99。
+- 前一張 CD `#4899` 已 Success，Gitea deploy marker/main 是 `f102616f0`，production build readback 是 `4a4b95e513`；production Agent99 work-item endpoint 仍投影 2026-07-10 舊總帳，因此 current P0 runtime closure 尚未成立。
+- `#4898` 與 `#4899` rollout 初始階段都曾短暫回裸 502，下一次 health probe 才恢復 HTTP 200，deployment 後續成功。這是 maintenance fallback / zero-raw-502 仍需關閉的 runtime evidence，不得因最終成功而刪除。
+- Production health 仍為 `degraded`：API/PostgreSQL/Redis/OpenClaw 可用，SignOz down，GCP-A Ollama timeout，GCP-B/local fallback 可用。
+- Reboot SLO live scorecard 為 `blocked_reboot_auto_recovery_slo_not_ready`、readiness 33%、24 blockers；不得宣稱全主機 10 分鐘恢復完成。
+
 ## 7. 工作項強制欄位與治理
 
 每個工作項必須先有：`id`、priority/order/status、scope、owner lane、dependencies、acceptance 與 next action。任何工作項升到 `candidate_ready`、`in_progress` 或 `executing` 前，還必須補齊 problem、trigger、evidence sources、AI role、action mode、controlled actions、prohibited actions、rollback、mode-specific verifier、callback、evidence refs 與 last update；缺一不可執行。
@@ -165,4 +177,4 @@
 
 ## 8. 立即下一步
 
-目前唯一 P0 仍是 `AG99-P0-002`。本機 Agent99 package 已完成 routing/dedupe/card/reboot identity 契約與 synthetic replay；下一步是先完成本分支 source/test/Gitea receipt，再將測試過的 Agent99 package 原子部署到 99，重跑 backup、provider、reboot、PostgreSQL、host、K3s replay corpus 並讀回 runtime manifest。99 尚未取得可用的 no-secret authenticated management channel前，狀態只能是 local green/runtime pending；不能把 `exitCode=0` 或本機測試當 production resolved。其後依序做 secure ingress、RepairCandidate、mode verifier/callback，再回到 reboot state machine live rerun。
+目前唯一 P0 仍是 `AG99-P0-002`。source/test candidate 已完成 Alertmanager 與 Telegram actionable card 共用 dispatcher、acceptance receipt、cold-start/CI-CD taxonomy、resolution policy 與 synthetic relay smoke。下一步依序是：完成 Gitea source receipt；等待單一視窗 CD 協調；原子部署測試過的 Agent99 package 到 99；驗證 runtime manifest；以 do-not-page 重播 ACTION REQUIRED cold-start、CI/CD failure、backup、provider、PostgreSQL、host、K3s corpus；讀回 Agent99 evidence、source-event closure 與 outbound envelope receipt。完成前只能標示 source green/runtime pending。其後再依序做 secure ingress、RepairCandidate、mode verifier/callback，最後回到 reboot state machine live rerun。