AgentSkillsCN

incident-response

事件管理、值班流程和运行手册执行。

SKILL.md
--- frontmatter
name: incident-response
description: Incident management, on-call procedures, and runbook execution.

Incident Response

Severity Levels

LevelDescriptionResponse Time
P1Service down15 min
P2Major degradation30 min
P3Minor impact4 hours
P4No impactNext business day

Incident Flow

code
Alert → Acknowledge → Assess → Mitigate → Resolve → Postmortem
          │             │         │
          └── Page ─────┴── Communicate

On-Call Checklist

  1. •Acknowledge alert within SLA
  2. •Assess impact and severity
  3. •Communicate status to stakeholders
  4. •Mitigate - Stop the bleeding
  5. •Investigate root cause
  6. •Resolve underlying issue
  7. •Document in postmortem

Communication Template

code
🔴 INCIDENT: [Brief description]
Impact: [Who/what is affected]
Status: [Investigating/Mitigating/Resolved]
ETA: [Expected resolution time]
Updates: [Channel/page]

Common Runbooks

High CPU

  1. •Identify process: top -c
  2. •Check for runaway processes
  3. •Scale horizontally if needed
  4. •Investigate root cause

Out of Disk

  1. •Check usage: df -h
  2. •Find large files: du -sh /* | sort -h
  3. •Clear logs/temp files
  4. •Add storage or archive

Database Slow

  1. •Check connections: SHOW PROCESSLIST
  2. •Identify slow queries
  3. •Kill blocking queries if needed
  4. •Scale or optimize

Escalation Path

code
On-Call Engineer (15 min)
    ↓
Team Lead (30 min)
    ↓
Engineering Manager (1 hour)
    ↓
VP Engineering (2 hours)