CrashLoopBackOff Debugging Prompt

You are a senior Kubernetes SRE who has root-caused hundreds of CrashLoopBackOff incidents. You know the difference between symptom and cause.

A pod is in CrashLoopBackOff. Your job:

1. Determine **which of these specific scenarios** is happening:
   - **Application crash on startup** — the container exits non-zero before getting healthy. Look in `--previous` logs for the exception.
   - **Missing or wrong config** — ConfigMap/Secret value missing, malformed YAML, wrong env var, or volume mount path wrong.
   - **Init container failure** — main container never starts because init failed. Check `kubectl describe` for init container status.
   - **Liveness probe killing it** — container starts, probe fails, kubelet kills it, restart, repeat. Check probe config vs actual readiness time.
   - **OOMKilled in a loop** — limit too low or memory leak. Container exits with code 137. Check `lastState.terminated.reason`.
   - **Permission denied / read-only filesystem** — security context too strict for what the app needs to write.
   - **Image entrypoint wrong** — `CrashLoopBackOff` with `exit code 127` (command not found) or `126` (not executable).
2. Quote the exact log line or describe output that supports your diagnosis.
3. Recommend the targeted fix (config change, manifest patch, image rebuild). Show me the YAML diff.
4. If you need additional data (e.g., previous container logs, the actual entrypoint, file permissions inside the image), ask for it before guessing.

Pod:
- Name: [POD NAME]
- Namespace: [NS]
- Image: [IMAGE:TAG]
- Restart count: [N]

`kubectl describe pod`:
```
[PASTE — make sure 'Last State' and 'Events' sections are included]
```

Current container logs:
```
[PASTE]
```

PREVIOUS container logs (the one that crashed):
```
[PASTE — kubectl logs <pod> --previous]
```

Pod spec fragment (containers, env, volumeMounts):
```yaml
[PASTE]
```