// Generic incident investigation playbook for service outages, 5xx spikes, latency regressions, broker issues, and crash loops when no organization-specific skill exists.
| name | generic-investigation |
| description | Generic incident investigation playbook for service outages, 5xx spikes, latency regressions, broker issues, and crash loops when no organization-specific skill exists. |
| allowed-tools | Read Bash Grep |
Use this skill when a task is clearly an incident investigation but there is no more specific environment or service skill available.
Ask:
Look for:
Trace from edge to dependency:
User -> ingress/load balancer -> service -> queue/db/cache/external dependency
Split at the midpoint if the cause is unclear.
When concluding, avoid saying only "service is broken". State:
If the environment-specific details are unknown, stay generic and state assumptions explicitly.
Example topology and dependency hints for testing overlay-loaded skills with pi-sre-mode.
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.