Kubernetes Health

What it does

This skill enables AI agents to query the health status of Kubernetes clusters and inspect specific resources for operational issues. It allows agents to verify cluster readiness, check pod states, and identify potential failures within the container orchestration environment.

When to use it

• Diagnosing why a deployment has failed or pods are stuck in CrashLoopBackOff.
• Verifying that critical infrastructure services are running after a cluster update or restart.
• Automating health checks before triggering automated rollback procedures.
• Investigating resource constraints like OOMKilled errors or Pending states.

Key capabilities

• Retrieves overall Kubernetes cluster health metrics.
• Inspects the status of specific pods, nodes, and services.
• Identifies common failure states such as NotReady, CrashLoopBackOff, and ImagePullBackOff.
• Provides actionable insights into resource utilization and scheduling issues.

Example prompts

• "Check the health of the production Kubernetes cluster and list any nodes that are not ready."
• "Investigate why the 'payment-service' pods are failing to start and report the specific error messages."
• "Run a diagnostic on the current cluster state and summarize any critical issues found in the last hour."

Tips & gotchas

Ensure your AI agent has appropriate RBAC permissions to read cluster resources, as health checks require access to node and pod details. This skill is most effective when combined with other Kubernetes skills that allow for remediation actions once an issue is identified.