clawpod-app Incident Response
Assess production incidents and provide structured rollback recommendations.
When to Use
- Production service shows elevated error rates or latency
- User reports a customer-facing issue after a deploy
- Monitoring alerts fire for a deployed service
Steps
- Identify the affected service and recent deployment history
- Gather current metrics: error rate, latency (P50/P95/P99), CPU, memory
- Compare against pre-deploy baseline metrics
- Assess severity using threshold analysis
- Recommend: no action / monitor / rollback with clear reasoning
Output Format
## Incident Assessment -- [Service]
**Trigger**: [What happened]
**Last Deploy**: [time] by [author] -- commit [sha]
### Current Metrics vs Baseline
| Metric | Current | Baseline | Delta | Status |
|--------|---------|----------|-------|--------|
| Error Rate | X% | Y% | +Z% | OK/WARN/CRITICAL |
| P99 Latency | Xms | Yms | +Zms | OK/WARN/CRITICAL |
### Recommendation
[Rollback / Monitor / No action needed] -- [reasoning]
### If Rollback Needed
1. [Step-by-step rollback procedure]
2. [Verification steps]