From 057001d866afd1f7766624b37d5adc655e32c6bb Mon Sep 17 00:00:00 2001 From: ogt Date: Wed, 15 Jul 2026 17:53:03 +0800 Subject: [PATCH] docs(agent99): record Host111 recovery receipt --- ...recovery-readback-2026-07-15.snapshot.json | 34 ++++++++++++++++--- docs/runbooks/FULL-STACK-COLD-START-SOP.md | 11 ++++++ 2 files changed, 41 insertions(+), 4 deletions(-) diff --git a/docs/operations/agent99-host112-host111-recovery-readback-2026-07-15.snapshot.json b/docs/operations/agent99-host112-host111-recovery-readback-2026-07-15.snapshot.json index 1d160a52b..e925a2e01 100644 --- a/docs/operations/agent99-host112-host111-recovery-readback-2026-07-15.snapshot.json +++ b/docs/operations/agent99-host112-host111-recovery-readback-2026-07-15.snapshot.json @@ -1,6 +1,6 @@ { "schema_version": "agent99_host_recovery_incident_readback_v1", - "generated_at": "2026-07-15T16:49:24+08:00", + "generated_at": "2026-07-15T17:50:00+08:00", "trace_id": "agent99-host112-kernel-memory-20260715-1531", "run_id": "AG99-H112-KMEM-20260715-1531", "work_item_id": "AG99-P0-HOST112-KERNEL-MEMORY", @@ -40,14 +40,40 @@ }, "host111": { "asset_type": "external_macos_host", - "terminal": "wake_sent_not_verified", - "reachable": false, + "terminal": "reachable_readback_verified_pressure_warning", + "reachable": true, "wol_attempts_observed": 3, "packets_per_attempt": 40, "failure_backoff_minutes": 30, "latest_runtime_evidence_ref": "agent99-Recover-20260715-164157.json", "telegram_message_id": "39483", - "safe_next_action": "retain incident, suppress duplicate WOL, verify power/network and exact host identity before recovery close" + "connectivity_recovered_at": "2026-07-15T17:50:00+08:00", + "controlled_readback_apply": { + "trace_id": "agent99-host111-recovered-20260715-1750", + "run_id": "AG99-H111-RECOVERED-20260715-1750", + "work_item_id": "AG99-P0-HOST111-RECOVERY", + "terminal": "verified_ready", + "script_sha256": "8490bbf3a1b784b348865791555dc3cb037f4ecd11b0d08e0778bbbf02e70d4c", + "direct_readback_ready": true, + "restricted_control_path_ready": true, + "rollback_available": true, + "private_key_value_read": false, + "secret_value_read": false + }, + "post_verifier": { + "boot_id": "darwin_1784043389", + "startup_enabled": "enabled", + "startup_active": "active", + "logical_cores": 8, + "load1_decay_sample": 17.05, + "cpu_host_percent_samples": [33.14, 27.0], + "mem_available_percent_samples": [78, 77], + "swap_used_percent": 64.66, + "disk_used_percent": 10, + "pressure_terminal": "warning_not_critical" + }, + "agent99_runtime_reconciliation_pending": true, + "safe_next_action": "deploy current Agent99 main, rerun scheduled Perf with darwin current-CPU schema, then close the historical WOL incident only after lifecycle writeback" }, "source_change": { "source_verified_at": "2026-07-15T17:30:53+08:00", diff --git a/docs/runbooks/FULL-STACK-COLD-START-SOP.md b/docs/runbooks/FULL-STACK-COLD-START-SOP.md index 96caeb354..921482739 100644 --- a/docs/runbooks/FULL-STACK-COLD-START-SOP.md +++ b/docs/runbooks/FULL-STACK-COLD-START-SOP.md @@ -2743,6 +2743,17 @@ Bonjour `.80` 位址均失聯。Agent99 已自動 Detect 並執行 allowlisted W - Host111 是外部 Mac,不是 Windows99 VMware guest;沒有 exact VMX 時禁止 VM power action、刪 `.lck`、殺 `vmware-vmx` 或把它標成已恢復。 +2026-07-15 17:50 Host111 `.111:22` 已重新可達,但不能只用 port open 關閉事件。run +`AG99-H111-RECOVERED-20260715-1750` 受控更新遠端 Darwin readback 與 110 restricted +forced-command public-key binding,保留 rollback backup;direct 與 control-path verifier +均回 READY,script SHA-256 為 +`8490bbf3a1b784b348865791555dc3cb037f4ecd11b0d08e0778bbbf02e70d4c`,Ollama startup +為 active。兩次 current CPU host samples 為 `33.14% / 27.00%`,屬 warning;memory +available `78% / 77%`、disk `10%`,舊 load1 從 `70.81` 衰減至 `17.05`。在新版 +Agent99 runtime 尚未部署前,舊 runtime 仍會用 load1/core 誤判 critical,因此 Host111 +只能標 `reachable_readback_verified_pressure_warning`;必須再以新版 scheduled Perf、 +lifecycle recovery writeback 與 Telegram receipt 關閉歷史 WOL incident。 + 機器可讀 incident evidence: `docs/operations/agent99-host112-host111-recovery-readback-2026-07-15.snapshot.json`。