{
  "schema_version": "host_runaway_aiops_loop_readiness_v1",
  "generated_at": "2026-06-18T15:08:00+08:00",
  "program_status": {
    "overall_completion_percent": 100,
    "current_priority": "P3",
    "current_task_id": "P3-009",
    "next_task_id": "P3-010",
    "read_only_mode": true,
    "runtime_authority": "host_runaway_aiops_loop_readiness_only_no_host_write",
    "status_note": "110 CPU runaway AIOps loop 已完成監控、告警、AI event packet、PlayBook / KM contract 與 production readback；runtime remediation 仍需 owner approval / maintenance window / evidence ref / dry-run / post-check。"
  },
  "source_refs": [
    "scripts/ops/host-runaway-process-exporter.py",
    "scripts/ops/host-runaway-process-remediation.py",
    "ops/monitoring/alerts-unified.yml",
    "apps/api/src/services/telegram_gateway.py",
    "docs/runbooks/HOST-RUNAWAY-PROCESS-AIOPS-PLAYBOOK.md",
    "docs/runbooks/FULL-STACK-COLD-START-SOP.md",
    "docs/LOGBOOK.md"
  ],
  "rollups": {
    "loop_stage_count": 6,
    "alert_lane_count": 2,
    "asset_writeback_contract_count": 5,
    "source_ref_count": 7,
    "live_readback_metric_count": 8,
    "blocked_runtime_action_count": 12,
    "runtime_remediation_authorized_count": 0,
    "telegram_send_count": 0,
    "gateway_queue_write_count": 0,
    "bot_api_call_count": 0,
    "host_write_count": 0,
    "process_termination_count": 0,
    "docker_restart_count": 0,
    "systemd_restart_count": 0,
    "nginx_reload_count": 0,
    "firewall_change_count": 0,
    "kubectl_action_count": 0,
    "production_write_count": 0
  },
  "loop_stages": [
    {
      "stage_id": "read_only_host_textfile_exporter",
      "display_name": "110 textfile exporter",
      "owner_agent": "openclaw",
      "status": "production_readback_complete",
      "completion_percent": 100,
      "evidence_refs": [
        "scripts/ops/host-runaway-process-exporter.py",
        "/home/wooo/node_exporter_textfiles/host_runaway_process.prom"
      ],
      "next_action": "保持每 2 分鐘 cron scrape，若 monitor missing / stale 則先修 observability，不宣稱 AI Ops 綠燈。",
      "blocked_runtime_actions": [
        "process_termination",
        "docker_restart"
      ]
    },
    {
      "stage_id": "prometheus_alert_rules",
      "display_name": "Prometheus alert lanes",
      "owner_agent": "hermes",
      "status": "deployed",
      "completion_percent": 100,
      "evidence_refs": [
        "ops/monitoring/alerts-unified.yml",
        "deploy-alerts #3147"
      ],
      "next_action": "用 HostOrphanBrowserSmokeHighCpu 分流 orphan smoke，用 HostCiRunnerLoadSaturation 分流合法 CI load。",
      "blocked_runtime_actions": [
        "auto_silence",
        "auto_restart"
      ]
    },
    {
      "stage_id": "telegram_ai_event_packet",
      "display_name": "Telegram / AI event packet",
      "owner_agent": "hermes",
      "status": "production_deployed",
      "completion_percent": 100,
      "evidence_refs": [
        "apps/api/src/services/telegram_gateway.py",
        "f358a0f6",
        "2d278568"
      ],
      "next_action": "告警實際 firing 時產生專屬 lane 與 dry-run / owner / evidence gate，不只輸出泛用 CPU 文本。",
      "blocked_runtime_actions": [
        "telegram_direct_send_without_gate",
        "bot_api_call_without_receipt"
      ]
    },
    {
      "stage_id": "playbook_contract",
      "display_name": "Runaway PlayBook",
      "owner_agent": "openclaw",
      "status": "ready",
      "completion_percent": 100,
      "evidence_refs": [
        "docs/runbooks/HOST-RUNAWAY-PROCESS-AIOPS-PLAYBOOK.md",
        "docs/runbooks/FULL-STACK-COLD-START-SOP.md"
      ],
      "next_action": "下一次事故依照 monitor -> alert -> triage -> dry-run -> approval -> post-check 順序執行。",
      "blocked_runtime_actions": [
        "skip_dry_run",
        "skip_post_check"
      ]
    },
    {
      "stage_id": "km_verifier_writeback_contract",
      "display_name": "KM / Verifier writeback contract",
      "owner_agent": "nemotron",
      "status": "contract_ready",
      "completion_percent": 100,
      "evidence_refs": [
        "docs/LOGBOOK.md",
        "docs/workplans/2026-06-04-reboot-cold-start-backup-recovery-workplan.md"
      ],
      "next_action": "真實事故需回寫 KM 條目、PlayBook 信任證據、Verifier post-check 與 recurrence guard。",
      "blocked_runtime_actions": [
        "claim_closed_without_km",
        "claim_closed_without_verifier"
      ]
    },
    {
      "stage_id": "gated_remediation_helper",
      "display_name": "Gated remediation helper",
      "owner_agent": "openclaw",
      "status": "dry_run_ready_apply_blocked",
      "completion_percent": 100,
      "evidence_refs": [
        "scripts/ops/host-runaway-process-remediation.py",
        "scripts/ops/tests/test_host_runaway_process_exporter.py"
      ],
      "next_action": "只有在 owner approval、maintenance window、evidence ref、rule、dry-run 與 post-check 都齊備後，才允許最小 SIGTERM。",
      "blocked_runtime_actions": [
        "sigkill",
        "restart_service"
      ]
    }
  ],
  "alert_lanes": [
    {
      "alertname": "HostOrphanBrowserSmokeHighCpu",
      "lane_id": "orphan_browser_smoke_runaway_process",
      "classification": "host_resource_runaway_process",
      "action_policy": "triage_packet_then_dry_run_then_gated_sigterm",
      "dry_run_allowed": true,
      "apply_allowed_without_owner_gate": false,
      "runtime_write_gate": 0,
      "next_action": "建立 triage packet，列出 process group、CPU、age、source command、dry-run target 與 post-check。"
    },
    {
      "alertname": "HostCiRunnerLoadSaturation",
      "lane_id": "ci_runner_load_saturation",
      "classification": "host_resource_capacity",
      "action_policy": "capacity_triage_no_process_remediation",
      "dry_run_allowed": false,
      "apply_allowed_without_owner_gate": false,
      "runtime_write_gate": 0,
      "next_action": "關聯 Gitea Actions run、runner queue、load/core、swap trend；合法 CI 不走 process kill。"
    }
  ],
  "asset_writeback_contract": [
    {
      "asset_id": "knowledge_base_incident_summary",
      "display_name": "Knowledge Base incident summary",
      "required_on_real_incident": true,
      "live_write_enabled": false,
      "required_fields": [
        "incident_id",
        "root_cause",
        "evidence_refs",
        "post_check_result"
      ]
    },
    {
      "asset_id": "playbook_trust_evidence",
      "display_name": "PlayBook trust evidence",
      "required_on_real_incident": true,
      "live_write_enabled": false,
      "required_fields": [
        "playbook_id",
        "matched_rule",
        "success_or_failure",
        "negative_reinforcement_note"
      ]
    },
    {
      "asset_id": "awooop_work_item_truth_chain",
      "display_name": "AwoooP Work Item truth-chain",
      "required_on_real_incident": true,
      "live_write_enabled": false,
      "required_fields": [
        "owner",
        "status",
        "approval_ref",
        "maintenance_window"
      ]
    },
    {
      "asset_id": "verifier_post_check",
      "display_name": "Verifier post-check",
      "required_on_real_incident": true,
      "live_write_enabled": false,
      "required_fields": [
        "before_metrics",
        "after_metrics",
        "alert_state",
        "service_health"
      ]
    },
    {
      "asset_id": "recurrence_guard",
      "display_name": "Recurrence guard",
      "required_on_real_incident": true,
      "live_write_enabled": false,
      "required_fields": [
        "prometheus_metric",
        "alert_lane",
        "owner_followup",
        "next_review_at"
      ]
    }
  ],
  "live_readback": {
    "host_label": "110",
    "monitor_up": 1,
    "orphan_browser_group_count": 0,
    "active_ci_container_count": 2,
    "load5_per_core_upper_observed": 0.81,
    "swap_used_ratio_upper_observed": 1.0,
    "remediation_authorized_count": 0,
    "alerts_firing_count": 0,
    "deploy_marker": "2d278568",
    "runtime_revision": "f358a0f6c3e614e407dedb6eee89bf10b2bc8173",
    "argocd_sync": "Synced",
    "argocd_health": "Healthy",
    "production_route_count": 3,
    "forbidden_public_hit_count": 0
  },
  "remediation_gate": {
    "dry_run_required": true,
    "owner_approval_required": true,
    "maintenance_window_required": true,
    "evidence_ref_required": true,
    "post_check_required": true,
    "allowed_signal_after_gate": "SIGTERM",
    "process_termination_authorized": false,
    "disallowed_actions": [
      "SIGKILL",
      "docker restart",
      "systemctl restart",
      "nginx reload",
      "firewall change",
      "kubectl action",
      "secret read",
      "production write"
    ]
  },
  "activation_boundaries": {
    "read_only_readback_allowed": true,
    "ai_triage_packet_allowed": true,
    "dry_run_generation_allowed": true,
    "runtime_remediation_enabled": false,
    "process_termination_authorized": false,
    "telegram_send_enabled": false,
    "gateway_queue_write_enabled": false,
    "bot_api_call_enabled": false,
    "host_write_enabled": false,
    "docker_restart_enabled": false,
    "systemd_restart_enabled": false,
    "nginx_reload_enabled": false,
    "firewall_change_enabled": false,
    "kubectl_action_enabled": false,
    "production_write_enabled": false,
    "secret_read_enabled": false
  },
  "next_steps": [
    {
      "step_id": "real_alert_receipt_fixture",
      "description": "下一次真實 firing 或脫敏 fixture 應驗證 Telegram card、AwoooP Work Item、KM / PlayBook / Verifier 欄位一致。",
      "runtime_write_allowed": false
    },
    {
      "step_id": "owner_approved_apply_drill",
      "description": "只有在人工批准維護窗口後，才可用非生產 fixture 演練 gated SIGTERM 與 post-check。",
      "runtime_write_allowed": false
    }
  ]
}