What problem does it solve? Production outages often trigger chaotic debugging while the system is still down, turning minor incidents into major ones. This Skill enforces a strict incident response order — triage severity first, contain the blast radius before investigating, root-cause only after the system is stable, and document everything in a postmortem. ## Core Features & Use Cases - Severity Triage: Classifies incidents as P1 (full outage), P2 (partial degradation), or P3 (minor issue) with defined containment time targets for each level. - Containment-First Workflow: Blocks root-cause analysis and code changes until containment is verified via health checks, supporting strategies like rollback, feature flags, traffic shifting, and rate limiting. - Postmortem Generation: Produces a complete incident report with timeline, root cause, contributing factors, and a prevention actions table saved to .rune/incidents/. - Use Case: Your login service starts returning 503 errors after a deploy. Invoke the skill to classify it as P2, roll back to the previous commit, verify recovery via watchdog, run a security check, and generate a postmortem identifying the connection pool exhaustion root cause. ## Quick Start Respond to the production outage where users report login failures by triaging severity, containing the issue, and generating a postmortem report.
