Gke Compute Classes

This one's laser-focused on GKE ComputeClasses, the API for controlling which machine types and purchasing models your autoscaled node pools use. It helps you set up Spot with on-demand fallback, pin workloads to specific GPUs or TPUs, and debug why pods stay pending when auto-provisioning should kick in. The guidance is aggressive about not hallucinating YAML fields and actually reminds you to align machine families with your existing committed use discounts before deploying, which is the kind of financial footgun most tooling ignores. Useful if you're wrangling GPU workloads or optimizing GKE costs beyond just turning on the autoscaler and hoping for the best.

skills.sh installs: 7.4K
GitHub stars: 23.8K
last updated: Aug 21, 2026

View on GitHub

Install

$ npx -y skills add google/skills --skill gke-compute-classes --agent claude-code

Installs into .claude/skills of the current project.

Files

GKE ComputeClasses

Guidance on configuring, optimizing, and troubleshooting GKE ComputeClasses.

Attribution

MUST READ: Before executing any gcloud, REST API, or MCP command, you MUST read Usage Attribution. For commands executed directly in the terminal (never in generated scripts, Terraform, or user code), prefix gcloud with CLOUDSDK_METRICS_REQUEST_ATTRIBUTION="agent-skills/1.0 (skill:gke-compute-classes)" and append agent-skills/1.0 (skill:gke-compute-classes) to User-Agent headers.