enterprise-agent-ops
長時間運行 agent 系統的營運控制、監控與變更管理
中文翻譯
--- name: enterprise-agent-ops description: 長時間運行或雲端託管的 agent 系統的營運控制:runtime 生命週期(啟動、暫停、停止、重啟)、可觀測性(log、metric、trace)、最小權限的安全範圍與 kill switch,以及帶有稽核 log 與成功率/成本指標的 rollout/rollback 變更管理。在 PM2、systemd 或 container 上跑 production agent 叢集,需要監控、事件應變或部署 gate 時使用。 metadata: origin: ECC --- # Enterprise Agent Ops 用在雲端託管或持續運行的 agent 系統,這類系統需要單一 CLI session 以外的營運控制。 ## 營運面向 1. runtime 生命週期(啟動、暫停、停止、重啟) 2. 可觀測性(log、metric、trace) 3. 安全控制(範圍、權限、kill switch) 4. 變更管理(rollout、rollback、稽核) ## 基本控制 - 不可變的部署產物 - 最小權限的憑證 - 環境層級的 secret 注入 - 硬性的 timeout 與重試預算 - 高風險動作的稽核 log ## 要追蹤的指標 - 成功率 - 每個任務的平均重試次數 - 恢復時間 - 每個成功任務的成本 - 失敗類別分佈 ## 事件處理模式 失敗暴增時: 1. 凍結新的 rollout 2. 擷取有代表性的 trace 3. 隔離出問題的路徑 4. 用最小的安全變更修補 5. 跑 regression 加安全檢查 6. 逐步恢復 ## 部署整合 這個 skill 可以搭配: - PM2 工作流 - systemd 服務 - container 編排器 - CI/CD gate
英文原文
--- name: enterprise-agent-ops description: Operational controls for long-lived or cloud-hosted agent systems — runtime lifecycle (start, pause, stop, restart), observability (logs, metrics, traces), least-privilege safety scopes and kill switches, and rollout/rollback change management with audit logs and success/cost metrics. Use when running production agent fleets on PM2, systemd, or containers that need monitoring, incident response, or deployment gates. metadata: origin: ECC --- # Enterprise Agent Ops Use this skill for cloud-hosted or continuously running agent systems that need operational controls beyond single CLI sessions. ## Operational Domains 1. runtime lifecycle (start, pause, stop, restart) 2. observability (logs, metrics, traces) 3. safety controls (scopes, permissions, kill switches) 4. change management (rollout, rollback, audit) ## Baseline Controls - immutable deployment artifacts - least-privilege credentials - environment-level secret injection - hard timeout and retry budgets - audit log for high-risk actions ## Metrics to Track - success rate - mean retries per task - time to recovery - cost per successful task - failure class distribution ## Incident Pattern When failure spikes: 1. freeze new rollout 2. capture representative traces 3. isolate failing route 4. patch with smallest safe change 5. run regression + security checks 6. resume gradually ## Deployment Integrations This skill pairs with: - PM2 workflows - systemd services - container orchestrators - CI/CD gates