k8s-troubleshooter

Features and purpose

Use this skill when users report Kubernetes cluster issues, pod failures, or need incident response. Comprehensive troubleshooting for diagnosing cluster, workload, networking, storage, and Helm issues. Invoke for: pods not starting (Pending, CrashLoopBackOff, ImagePull), service connectivity problems, DNS resolution failures, storage/PVC issues, node health problems, CNI/Calico networking, Helm release failures, or cluster-wide performance degradation. Provides systematic diagnostic workflows, standardized investigation reports with severity-based depth (Executive Triage Cards for rapid decision-making), and incident response playbooks with phased triage (baseline → inspect → correlate → deep dive). All commands are read-only by default for production safety.

Quick start

Follow the repository instructions for this skill and your agent. Review the files and dependencies before adding it to your workflow.