troubleshooting

name

Troubleshooting

description

Debugging and incident management — 6-step triage, Stop-the-Line, competing hypotheses for dev-time bugs plus incident detection, response, and blameless postmortems for production issues. USE WHEN debugging, fix bug, error recovery, troubleshooting, reproduce bug, root cause, stack trace, test failing, build broken, incident, outage, SEV1, SEV2, postmortem, blameless review, incident response, 5 whys.

Troubleshooting

Unified skill for when things break — in development or in production.

Workflow Routing

Request Pattern	Route To
Debugging, fix bug, error recovery, troubleshooting, reproduce bug, root cause, stack trace, test failing, build broken	Debugging/SKILL.md
Incident, outage, SEV1, SEV2, postmortem, blameless review, incident response, war room, 5 whys, incident timeline	Incident/SKILL.md

Examples

Example 1: Debug a failing test

User: "This test is failing intermittently"
→ Routes to Debugging
→ Non-reproducible bug playbook: isolate, instrument, competing hypotheses

Example 2: Production incident

User: "The API is returning 500s in production"
→ Routes to Incident
→ New incident workflow: triage, mitigate, communicate, resolve

Example 3: Post-incident review

User: "Write a postmortem for yesterday's outage"
→ Routes to Incident/Postmortem
→ Blameless review: timeline, contributing factors, action items