pineapple880066 Incident Response
Use this skill for service degradation, outages, or repeated application errors.
Information requests
Read references/checklist.md and answer only from the available evidence. State which facts are confirmed and which remain unknown.
Action requests
Use only the mounted service_health_read tool to collect read-only health evidence. Propose the next diagnostic command and request operator approval before any change. Never restart services, delete files, rotate credentials, or modify infrastructure through this skill.
Output
Return the current impact, observed evidence, likely next check, and any required human approval.