| name | deployment-correlation-triage |
| description | Quick triage skill to correlate incidents with recent deployments from Jenkins CI/CD. Use during initial incident assessment to determine if a recent code or infrastructure change is the likely trigger. This skill should run as part of any initial incident investigation. |
Deployment Correlation Triage
Use this skill in the initial triage phase of any incident investigation to
determine whether a recent deployment is correlated with the incident.
Step 1: Identify affected services
From the incident alert or alarm, extract:
- Affected service name(s)
- AWS account and AWS Region where the incident is occurring
- Incident start timestamp
Step 2: Check recent deployments
For each affected service, invoke the get_recent_deployments MCP tool with:
service_name: the affected service
hours: 4
Step 3: Classify the correlation
| Deployment timing relative to incident | Correlation | Action |
|---|
| Within 1 hour | VERY HIGH | Recommend immediate rollback; treat the deployment as the primary hypothesis |
| Within 2 hours | HIGH | Recommend rollback; investigate for confirmation |
| Within 4 hours | MEDIUM | Keep the deployment as an active hypothesis but investigate other causes in parallel |
| No deployment | LOW | Focus on infrastructure, capacity, or external-dependency causes |
Step 4: Document and escalate
Include in the triage summary:
- Number of deployments found per service
- Correlation level for each
- Recommended investigation direction (rollback, RCA focus areas)
Escalate to Incident RCA with these findings attached.