Kubernetes Troubleshooter

Systematic Kubernetes incident response and debugging. Diagnoses CrashLoopBackOff, ImagePullBackOff, OOMKilled, and Pending pods, fixes NotReady nodes and DiskPressure, debugs service networking and PVC failures, investigates Helm releases, and runs post-incident reviews.

This skill is a kubectl-driven SRE playbook. It walks through structured triage for failing pods, node-level issues, networking and storage failures, and performance degradation, then guides remediation and a blameless post-incident review.

When to use

Use when a pod is crashing or stuck, a node goes NotReady, service connectivity breaks, a PVC won't bind, or you're responding to a Kubernetes production incident.

Examples

CrashLoopBackOff

Diagnose a crashing pod

A pod is in CrashLoopBackOff with exit code 137 — walk me through the kubectl steps to find the root cause and fix it

Service unreachable

Debug networking

My Service returns connection refused from other pods but the deployment is healthy — systematically debug the networking