Skip to content
Get packs

GKE ComputeClasses Helper

Lets an agent configure, tune and troubleshoot GKE ComputeClass node pools for Spot, GPU/TPU and machine-family targeting

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

What it does

This skill gives an AI agent detailed guidance for Google Kubernetes Engine (GKE) ComputeClasses, which are rules that tell the cluster what kinds of machines to create for workloads. Once installed, the agent can draft example ComputeClass YAML (for Spot VMs with on-demand fallback, specific GPUs/TPUs, or machine families), explain scheduling rules such as taints and tolerations, and help debug pending pods. It also ships a set of example YAML templates and a shell script that live-tails cluster autoscaler events using gcloud and jq.

Great for

  • Platform or DevOps engineers running GKE who need ComputeClass YAML drafted or reviewed
  • Teams trying to cap a committed-use machine family and spill excess demand to other families
  • Engineers debugging pods stuck in Pending or noScaleUp on GKE

Skip it if

  • Cluster-level Node Auto Provisioning or general GKE cluster creation, which the skill itself says are out of scope
  • Production deployments without review, since the templates are labelled EXAMPLE TEMPLATE - DO NOT DEPLOY and machine-family choices depend on your own commitments and zone availability
  • Users without GKE experience, as the guidance assumes Kubernetes and gcloud knowledge

Try asking

  • “Write a ComputeClass that prefers Spot VMs and falls back to on-demand nodes for my batch jobs”
  • “Why is my GPU pod stuck Pending with noScaleUp on GKE?”
  • “Cap N4 usage at my 100-core CUD and spill excess load to N4D and C4”
  • “Show me a live tail of autoscaler scale-up failures for my cluster”
Step-by-step guide: how to use GKE ComputeClasses HelperInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 15 original files, the Apache-2.0 licence and an attribution note.

    gke-compute-classes.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip gke-compute-classes.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip gke-compute-classes.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

  • Replace placeholders such as <YOUR-ZONE-HERE> and align machine families with your existing CUDs or reservations before applying anything
  • Review the bash script and run it with --errors-only first to limit output while you test
  • Ask the agent to explain any YAML field it proposes and check it against current GKE documentation, since the skill references upstream code and version-specific behaviour

What's inside

  • assets/balanced-reserved-zonal-compute-class.yaml
  • assets/capacity-quota-spillover.yaml
  • assets/computeclass-rbac-editor.yaml
  • assets/dynamic-rwo-storageclass.yaml
  • assets/equal-priority-tiebreak-compute-class.yaml
  • assets/genai-inference-g4-compute-class.yaml
  • assets/kafka-broker-compute-class.yaml
  • assets/log-autoscaler-events.sh
  • assets/manual-pool-tiebreak-compute-class.yaml
  • assets/postgres-primary-compute-class.yaml
  • assets/redis-compute-class.yaml
  • assets/restrict-computeclass-usage-vap.yaml
  • assets/shared-l4-inference-compute-class.yaml
  • assets/spark-proximal-compute-class.yaml
  • 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.