agentleFS
Sign inStart free
ArangoGutierrezArangoGutierrez/claude-toolkit/.claude/skills/k8s-debug/SKILL.md

k8s-debug

Skill4 starsChanged 3 months agoMITView on GitHub

Structured Kubernetes debugging for GPU workloads. Triggered by "pod crash", "CrashLoopBackOff", "OOMKilled", "GPU scheduling", or /k8s-debug

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Check: phase, conditions, restart count, node, QoS class.

    line 23
  2. Distinctive:

    Check events before logs

    line 71
  3. Distinctive:

    Verify node allocatable before assuming GPU issue

    line 72

≈ “Common” is an estimate: at least 1 in 5 of the 308,232 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

K8s Debug — GPU Workload Triage

Structured debugging workflow. Do NOT skip steps.

Triage Order

1. Pod Status

kubectl get pod <name> -n <ns> -o wide
kubectl describe pod <name> -n <ns>

Check: phase, conditions, restart count, node, QoS class.

2. Events (BEFORE logs)

kubectl get events -n <ns> --sort-by=.lastTimestamp --field-selector involvedObject.name=<name>

Events explain why logs may be empty.

3. Container Logs

kubectl logs <name> -n <ns> --previous
kubectl logs <name> -n <ns> -c <init-container>

4. Node Conditions

kubectl describe node <node>
kubectl get node <node> -o jsonpath='{.status.allocatable}'

5. Resource Requests vs Allocatable

kubectl describe node <node> | grep -A 20 "Allocated resources"

6. GPU Checks

See references/gpu-scheduling-checklist.md for full checklist.

kubectl get nodes -o jsonpath='{range .items[*]}{.metadata.name}{"\t"}{.status.allocatable.nvidia\.com/gpu}{"\n"}{end}'
kubectl get pods -n kube-system -l app=nvidia-device-plugin

7. DRA Checks

kubectl get resourceclaims -n <ns>
kubectl describe resourceclaim <name> -n <ns>

Common Patterns

SymptomCauseFix
CrashLoopBackOff + CUDA errorCUDA mismatchMatch runtime to driver
Pending + insufficient GPUNo capacityCheck allocatable
OOMKilledGPU mem exceededReduce batch size
Device plugin not readyPlugin crashedCheck DaemonSet

Gotchas

  • Check events before logs
  • Verify node allocatable before assuming GPU issue
  • CUDA errors often = driver/runtime mismatch, not code bugs

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free