Skip to content
Skill

Gke AI troubleshooting tpu partition overlap

Diagnoses GKE TPU SliceCreationFailed errors caused by overlapping partitions or nodes

Install
Scripts
None
Platforms
Claude Code, Codex, Cursor
GitHub stars
24.4K
Last commit
2 days ago
License
Apache-2.0
Listing updated
1 day ago

Similar skills

About Gke AI troubleshooting tpu partition overlap

Gke AI troubleshooting tpu partition overlap is an open-source agent skill for cloud and infrastructure. It is available on Claude Code, Codex, Cursor, Gemini CLI, GitHub Copilot, Windsurf and Goose.

This skill walks an agent through a read-only investigation of a GKE TPU slice that fails to form with the SliceCreationFailed reason because its partitions or nodes are already held by another slice or JobSet. It first confirms the failure status, then rules out the other documented causes such as missing nodes or a partition count that does not match the topology.

Next it inspects the partition and state labels on the node pool, counts which partitions are healthy or usable, and locates any other Slice resource that shares nodes with the failed one. Cloud Monitoring queries on the partition state metric help confirm past or current associations. Fixes are routed through GKE options such as Kueue with TAS, slice reuse settings and the retry-on-failure annotation, and the agent is told never to delete or modify other slices or node pools itself.

The skill expects a GKE cluster with kubectl access, gcloud, and the Kubernetes Engine Viewer and Monitoring Viewer roles. It hands off to a sibling skill for general slice lifecycle or provisioning problems that do not involve node overlap.

When your agent uses it

Diagnose and resolve GKE TPU slice creation failures (SliceCreationFailed) caused by partition overlap conflicts where requested TPU partitions or their nodes are already used by another slice or JobSet. Use when diagnosing GKE TPU slice creation failures with SliceCreationFailed partition or node conflicts. Don't use for general slice lifecycle states or slice provisioning failures without node overlap (use gke-ai-troubleshooting-tpu-dynamic-slices-monitoring), quota exhaustion, or unresponsive TPU nodes.

From the skill's SKILL.md. Your agent reads this description to decide when to load the skill.

What it can access

From a static check of its 2 files (21 KB) on GitHub on 17 November 2026. Nothing was run. Read the files before you install a skill: it can tell your agent to run anything its tools allow.

Check another skill before you install it.

Install

Copy the skill folder to ~/.claude/skills/gke-ai-troubleshooting-tpu-partition-overlap to use it in all your projects:

tmp=$(mktemp -d) &&
git clone --depth 1 https://github.com/google/skills.git "$tmp" &&
mkdir -p ~/.claude/skills &&
rm -rf ~/.claude/skills/gke-ai-troubleshooting-tpu-partition-overlap &&
cp -R "$tmp/skills/cloud/gke-ai-troubleshooting-tpu-partition-overlap" ~/.claude/skills/gke-ai-troubleshooting-tpu-partition-overlap &&
rm -rf "$tmp"

For one project only, put the folder in .claude/skills/gke-ai-troubleshooting-tpu-partition-overlap inside it instead.

Claude Code docs on skills

Copy the skill folder to ~/.agents/skills/gke-ai-troubleshooting-tpu-partition-overlap to use it in every repository:

tmp=$(mktemp -d) &&
git clone --depth 1 https://github.com/google/skills.git "$tmp" &&
mkdir -p ~/.agents/skills &&
rm -rf ~/.agents/skills/gke-ai-troubleshooting-tpu-partition-overlap &&
cp -R "$tmp/skills/cloud/gke-ai-troubleshooting-tpu-partition-overlap" ~/.agents/skills/gke-ai-troubleshooting-tpu-partition-overlap &&
rm -rf "$tmp"

For one repository only, put the folder in .agents/skills/gke-ai-troubleshooting-tpu-partition-overlap inside it instead.

Codex docs on skills

Copy the skill folder to ~/.cursor/skills/gke-ai-troubleshooting-tpu-partition-overlap to use it in every project:

tmp=$(mktemp -d) &&
git clone --depth 1 https://github.com/google/skills.git "$tmp" &&
mkdir -p ~/.cursor/skills &&
rm -rf ~/.cursor/skills/gke-ai-troubleshooting-tpu-partition-overlap &&
cp -R "$tmp/skills/cloud/gke-ai-troubleshooting-tpu-partition-overlap" ~/.cursor/skills/gke-ai-troubleshooting-tpu-partition-overlap &&
rm -rf "$tmp"

For one project only, put the folder in .cursor/skills/gke-ai-troubleshooting-tpu-partition-overlap inside it instead.

Cursor docs on skills

Run in a terminal (add --scope workspace to install it for the current workspace only):

gemini skills install https://github.com/google/skills.git --path skills/cloud/gke-ai-troubleshooting-tpu-partition-overlap

Or copy the folder to ~/.gemini/skills/gke-ai-troubleshooting-tpu-partition-overlap yourself.

Gemini CLI docs on skills

For GitHub Copilot in VS Code, copy the skill folder to ~/.copilot/skills/gke-ai-troubleshooting-tpu-partition-overlap:

tmp=$(mktemp -d) &&
git clone --depth 1 https://github.com/google/skills.git "$tmp" &&
mkdir -p ~/.copilot/skills &&
rm -rf ~/.copilot/skills/gke-ai-troubleshooting-tpu-partition-overlap &&
cp -R "$tmp/skills/cloud/gke-ai-troubleshooting-tpu-partition-overlap" ~/.copilot/skills/gke-ai-troubleshooting-tpu-partition-overlap &&
rm -rf "$tmp"

For one project only, put the folder in .github/skills/gke-ai-troubleshooting-tpu-partition-overlap inside it instead.

VS Code docs on skills

Get the folder:

git clone --depth 1 https://github.com/google/skills.git

Then copy skills/skills/cloud/gke-ai-troubleshooting-tpu-partition-overlap into your agent's skills folder as gke-ai-troubleshooting-tpu-partition-overlap. Agents that support SKILL.md read one folder per skill.

The SKILL.md format

What it does

Confirm SliceCreationFailed status
Checks the Ready condition reason and records partition IDs, topology and labels from the Slice resource.
Rule out other causes
Checks for missing nodes or a mismatch between topology and partition count before concluding overlap.
Map partition usage on nodes
Reads partition ID and state labels across all slice shapes to see which partitions are available.
Find conflicting slices
Identifies other slices and JobSets that use the same nodes, using the Slice resources and Cloud Monitoring metrics.
Guided resolution options
Suggests Kueue and TAS scheduling, slice reuse settings, retry annotations or escalation to support.
Read-only safety rules
Limits the agent to read-only commands and leaves any cluster changes to the user.

Good for

  • Diagnose why a TPU slice in GKE stays in SliceCreationFailed
  • Find which other slice is holding the nodes my TPU partitions need
  • Check which TPU partitions are healthy and free for a new dynamic slice
  • Decide whether to rely on Kueue or retry annotations to recover a failed slice

Gke AI troubleshooting tpu partition overlap FAQ

What does this skill need to run?

It needs a GKE cluster reachable through kubectl, gcloud set up with a billing-enabled project, the Kubernetes Engine Viewer and Monitoring Viewer roles, and the Google Kubernetes Engine and Cloud Monitoring APIs enabled.

Does it change my cluster?

No. The skill is written to run only read-only diagnostic commands and to give any fix to the user to apply.

When should I use this skill instead of the general TPU monitoring one?

Use it when the failure is SliceCreationFailed and involves partition or node overlap. For general lifecycle states, provisioning failures without overlap, quota issues or unresponsive nodes, the skill points to the dynamic slices monitoring skill.

Does it delete other slices to free nodes?

No. The skill explicitly forbids deleting or changing another slice or stopping its workload, and says the owner of that slice decides when to release it.

How do I install Gke AI troubleshooting tpu partition overlap in Claude Code?

Copy the skill folder from GitHub to ~/.claude/skills/gke-ai-troubleshooting-tpu-partition-overlap to use it in all your projects, or to .claude/skills/gke-ai-troubleshooting-tpu-partition-overlap in one project.

Does Gke AI troubleshooting tpu partition overlap run scripts?

No. It is instructions only, with no scripts. This comes from a static check of the files on GitHub; nothing was run. Read the files before you install it.

Which AI agents can use Gke AI troubleshooting tpu partition overlap?

Claude Code, Codex, Cursor, Gemini CLI, GitHub Copilot, Windsurf and Goose. Each reads skills in the SKILL.md format from its own folder; the install steps on this page list them.

What are the best alternatives to Gke AI troubleshooting tpu partition overlap?

The closest alternatives to Gke AI troubleshooting tpu partition overlap are Gke AI troubleshooting tpu slice mig stuck updating, Gke AI troubleshooting tpu dynamic slices monitoring, Gke AI troubleshooting tpu reservation stockout, Gke AI troubleshooting tpu vm repair stuck and Gke AI troubleshooting tpu mxla hang.

Where this comes from

  • Description: from the skill's own SKILL.md.

More from google/skills

  1. Route platform choices to Google's own skills by fetching and filtering its skill catalog

    • Open source
    • 24.4K GitHub stars
    • Cloud and infrastructure
  2. Guardrails for gcloud commands: validate syntax with help, scope projects, dry-run first

    • Open source
    • 24.4K GitHub stars
    • Cloud and infrastructure
  3. Authenticate to Google Cloud: ADC, impersonation, service accounts and IAM roles

    • Open source
    • 24.4K GitHub stars
    • Cloud and infrastructure
  4. First-time Google Cloud setup: sign in, pick a project, link billing, with consent gates

    • Open source
    • 24.4K GitHub stars
    • Cloud and infrastructure
  5. Look up Google developer docs through the Developer Knowledge MCP server or REST fallback

    • Open source
    • 24.4K GitHub stars
    • Developer tools