Triage

Install in one command

mkdir -p ~/.claude/skills/triage && curl -fsSL https://eliteaiadvantage.com/skills/triage/SKILL.md -o ~/.claude/skills/triage/SKILL.md

When something breaks in production, the instinct is to dive into code. Half the time that's right; half the time you spend an hour fixing something a 30-second rollback would have handled. Triage forces a decision step before the code dive: severity, blast radius, reversibility, time pressure, then a single response category.

Built for live-user incidents — public site down, queue stuck, admin tool broken, deploy gone wrong. Threads the middle between panic-fix and over-thought.

• →Forces explicit severity (🔴 down / 🟡 degraded / 🟢 cosmetic) so the response matches the actual cost.
• →Names blast radius (public / admin / pipeline / one user) before deciding how urgently to act.
• →Checks reversibility — rollback safe? data damaged? side effects irreversible? — before diving in.
• →Picks one response category from five (rollback / hotfix / patch / monitor / ignore) — no menus.
• →Pairs cleanly with /post-mortem after the fix lands.

• •Production site or feature just broke.
• •Deploy went out and metrics are off.
• •A user reported something that might be a bigger issue.
• •Anything where the first instinct is "let me jump in and fix it right now."

Without this skill, incident response defaults to "let me investigate" — which is a response category, but rarely the right one for production downtime. Triage forces the explicit assessment-then-decision pattern: name the severity, pick one of five responses, get the next concrete action. Investigation becomes a chosen response, not a default avoidance of commitment.