// Generic SRE investigation methodology for latency, errors, crashes, capacity issues, and dependency failures. Use for production incident investigation when you need a disciplined troubleshooting loop.
Example topology and dependency hints for testing overlay-loaded skills with pi-sre-mode.
Generic incident investigation playbook for service outages, 5xx spikes, latency regressions, broker issues, and crash loops when no organization-specific skill exists.
| name | sre-methodology |
| description | Generic SRE investigation methodology for latency, errors, crashes, capacity issues, and dependency failures. Use for production incident investigation when you need a disciplined troubleshooting loop. |
| allowed-tools | Read Bash Grep |
Use this skill to investigate incidents systematically.
Avoid command carpet-bombing. Every command should have a reason.
Always think about:
Use them together, not in isolation.
When writing findings, use this shape:
## Summary
- concise operator-facing summary
## Timeline
- event order with timestamps and timezones when relevant
## Findings
- evidence-backed statements only
## Root cause hypothesis
- current best explanation + confidence
## Actions
- mitigation
- prevention
- detection
This skill is for investigation, not remediation. Prefer inspection commands and avoid mutations unless the user explicitly changes the operating mode.