| name | infra-logs |
| description | Use when checking pod logs, searching for errors, viewing recent log output, or debugging runtime issues in DEV or PROD. Use when user says "logs de", "ver logs", "qué errores hay", "check logs", "show errors", or mentions log analysis of a service. |
| argument-hint | [service-name] [environment] |
| allowed-tools | Bash, AskUserQuestion |
| model | sonnet |
Kubernetes Pod Logs
Overview
Search and analyze logs from pods in the cluster. Supports filtering by time, error patterns, and entities.
Pod Discovery
Read namespace from the project's CLAUDE.md (## Infrastructure section).
kubectl --context=<env> get pods -n <namespace> | grep <service>
Log Commands
Recent errors
kubectl --context=<env> logs <pod> -n <namespace> --since=<duration> 2>&1 \
| grep -i "error\|exception\|fail" | tail -50
Duration examples: 5m, 30m, 1h, 3h, 24h
Errors for a specific entity/operation
kubectl --context=<env> logs <pod> -n <namespace> --since=<duration> 2>&1 \
| grep -i "<entity>\|<operation>" | grep -i "error" | tail -30
Full context around errors (5 lines before/after)
kubectl --context=<env> logs <pod> -n <namespace> --since=<duration> 2>&1 \
| grep -B5 -A5 -i "error\|exception" | tail -80
Follow logs in real-time (for live debugging)
kubectl --context=<env> logs -f <pod> -n <namespace> | grep -i "<pattern>"
Logs from all pods of a deployment
kubectl --context=<env> logs -l app=<service> -n <namespace> --since=<duration> 2>&1 \
| grep -i "error" | tail -50
Previous container logs (after a crash/restart)
kubectl --context=<env> logs <pod> -n <namespace> --previous | tail -100
Output Format
Summarize findings:
- Total error count in the period
- Unique error types/messages
- Frequency patterns (burst vs steady)
- Affected entities (IDs, clients, etc.)
If no errors found, say so explicitly — don't assume the absence of evidence is evidence of absence. Suggest checking a wider time range or different pod.
