What problem does it solve? When production breaks, teams lose critical minutes to chaos: everyone debugs at once, nobody declares severity, and root-cause hunting delays the rollback that would have restored users immediately. This Skill replaces adrenaline with structure for the first fifteen minutes of an incident. ## Core Features & Use Cases - Severity Declaration Ladder: Assess impact in user terms and declare SEV1/SEV2/SEV3, with a bias toward declaring high and downgrading later. - Role Assignment: Split the Incident Commander (coordination, communication) from Ops leads (hands on keyboards) so coordination never stops. - Mitigation-First Stabilization: Check levers in order of speed — rollback recent deploys, kill feature flags, drain bad nodes — before any root-cause diagnosis. - Timeline Logging and Communication Cadence: Timestamped channel notes from minute one, plus scheduled status updates even when there is nothing new to report. - Use Case: An alert fires showing checkout errors at 40%. Use this Skill to declare SEV1 within minutes, assign an IC and ops lead, roll back the suspect deploy before diagnosing, confirm recovery in the user-facing metric, and schedule the postmortem. ## Quick Start Use the incident-triage skill to help me assess severity, assign roles, and stabilize this production outage where users are reporting checkout errors.
