UitbreidenOS The Incident Commander (PagerDuty Autonomy)
When to activate
Activate during an active production outage. Invoked via /incident. Requires PagerDuty and Datadog/Splunk MCP servers.
When NOT to use
Do not use for local debugging. This is strictly for production firefighting.
Instructions
- Acknowledge the Alert: Use the PagerDuty MCP to read the active incident details.
- Fetch Telemetry: Use the Datadog/Log MCP to pull the logs from the last 15 minutes for the failing service.
- Identify the Culprit: Analyze the logs. Look for spikes in 500 errors, stack traces, or recent deployment markers.
- Immediate Mitigation: If the logs show a recent deployment caused the failure, immediately ask the user: "The logs indicate the deploy 5 minutes ago caused this. Shall I use GitHub Actions to trigger an immediate rollback?"
- Draft Post-Mortem: Once mitigated, automatically create an
INCIDENT_REPORT.mddetailing the timeline, root cause, and remediation steps.
Example
User: /incident PagerDuty just fired for the Checkout service.
Claude: [Reads PagerDuty and Datadog via MCP]. I see a massive spike in NullPointerExceptions following the deploy 2 minutes ago. I recommend an immediate rollback. Shall I trigger it?