Skip to content
Get packs

GKE Workload Troubleshooting

Diagnoses failing GKE pods (CrashLoopBackOff, OOMKilled, ImagePullBackOff, Pending) from logs and events

Source Download skill
Repository
google/skills
Stars
21,782
Licence
Apache-2.0
Files
2 files

What it does

When a application running on Google Kubernetes Engine (GKE) keeps crashing, won't start, or can't pull its container image, this skill makes the agent follow a structured checklist: check pod status, read events and logs, and work out the likely root cause. It then proposes a fix, such as a manifest change or an IAM adjustment, and explains it, but it is designed not to change live systems itself. If the cluster can't be reached, it gives you the exact kubectl commands to run and still drafts a likely fix from the symptoms you describe.

Great for

  • Developers whose pods show CrashLoopBackOff, OOMKilled, ImagePullBackOff or Pending status
  • Platform or SRE engineers who want a consistent first-pass diagnosis of workload failures
  • Teams working offline or in a sandbox who want a runbook of kubectl commands plus a suggested manifest fix

Skip it if

  • Provisioning GKE clusters or creating node pools, which the skill explicitly excludes (infrastructure work, not workload debugging)
  • Problems with non-Kubernetes Google Cloud services, which are out of scope
  • Production incidents where you should not rely on an AI's proposed fix without review, since the skill only proposes changes and a human must apply and verify them

Try asking

  • “My pod in the payments namespace keeps going into CrashLoopBackOff with exit code 137, figure out why”
  • “Why is my deployment 'api-server' stuck in ImagePullBackOff on cluster prod-east?”
  • “These pods have been Pending for 20 minutes in the default namespace, what is blocking scheduling?”
  • “Diagnose the crash loop that started around 14:30 UTC today in the orders workload”
Step-by-step guide: how to use GKE Workload TroubleshootingInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 2 original files, the Apache-2.0 licence and an attribution note.

    gke-workload-troubleshooting.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip gke-workload-troubleshooting.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip gke-workload-troubleshooting.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Give the project ID, cluster name, location and namespace up front so the agent does not have to guess from your environment defaults.
  • Include the time the problem started, or say 'five minutes ago', so the agent can center its log query window correctly.
  • Review every proposed manifest or IAM change before applying it; the skill is built to propose fixes, not apply them.

What's inside

  • references/attribution.md
  • SKILL.md

Redistributed unmodified from google/skills under the Apache-2.0 licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.