Skip to content
Get packs

GKE HPA Scaling Troubleshooter

Diagnoses why a Kubernetes autoscaler on GKE is stuck, erroring, or not scaling, and proposes fixes.

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

What it does

When a Kubernetes workload on Google Kubernetes Engine (GKE) fails to scale as expected, this skill guides the AI agent through a structured investigation of its HorizontalPodAutoscaler (HPA), the object that adds or removes copies of an app based on load. The agent reads HPA status and events, matches the error to a known failure pattern (such as unknown metrics, missing resource requests, or a broken custom-metrics pipeline), and recommends a specific correction packaged as a reviewable configuration change. It is designed to be read-only during diagnosis, and if the cluster cannot be reached it gives the user the exact commands to run instead.

Great for

  • Platform or DevOps engineers whose GKE HPA shows <unknown> metrics or FailedGet* events
  • Developers whose app does not scale up or down as expected under load
  • Teams troubleshooting custom or external metrics feeding an autoscaler

Skip it if

  • Writing new HPA or VPA configurations or following scaling best practices; the skill's own description points to a separate gke-workload-scaling skill for this
  • Cluster Autoscaler or node-pool sizing problems, which are explicitly out of scope
  • Beginners without access to a cluster, since the most useful steps need kubectl and gcloud and live cluster data

Try asking

  • “My HPA on the checkout deployment shows <unknown> for its CPU metric. Help me find out why.”
  • “Our HPA reports FailedGetExternalMetric events. Walk me through the diagnosis.”
  • “This HPA in namespace prod never scales down even though load is low. What could cause that?”
  • “My workload with minReplicas 0 will not scale up from zero. Investigate.”
Step-by-step guide: how to use GKE HPA Scaling TroubleshooterInstall, 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-scaling-troubleshooting.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

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

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip gke-workload-scaling-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 agent the HPA name, namespace, cluster name, location and project up front so it does not have to guess.
  • Review every proposed manifest change before applying it; the skill frames fixes as GitOps changes for human review.
  • If the agent cannot reach your cluster, run the diagnostic commands it lists and paste the output back into the conversation.

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.