Claude Code Kubernetes HPA Skill for Metric and Replica Checks
Set scaling bounds from an observed workload and verify that an HPA can read its metric and adjust replica counts.
中文Install in Claude Code
Use a permitted test cluster with kubectl configured, a scalable Deployment, and metrics-server. Configure CPU requests on the relevant containers when testing CPU utilization.
- Get the complete `plugins/devops/auto-scaling-configurator/skills/configuring-auto-scaling-policies` directory from the pinned GitHub checkout.
- Copy it under the target project's `.claude/skills/`. Confirm `.claude/skills/configuring-auto-scaling-policies/SKILL.md`, open Claude Code at the project root, and call `/configuring-auto-scaling-policies`.
Example task
Set a scratch service to a range of two to five replicas with a CPU utilization target. Observe it during controlled load and recovery.What to check
The report records metric availability, scaling events and observed replica counts within the configured bounds.
Original SKILL.md on GitHub · Updated
Frequently asked question
Why might the CPU target show as unknown?
Check the resource metrics API, the containers' CPU requests and the HPA conditions, then confirm the observation window covers enough samples.