k8s-debug

Diagnose Kubernetes pod and cluster issues with structured kubectl commands.

3|Updated Dec 30, 2025
One-click install
npx skills add https://github.com/martin-janci/claude-marketplace --skill k8s-debug-martin-janci
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: k8s-debug
Source: https://github.com/martin-janci/claude-marketplace/tree/main/plugins/k8s-toolkit/skills/k8s-debug
Command: npx skills add https://github.com/martin-janci/claude-marketplace --skill k8s-debug-martin-janci

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Provides a systematic, repeatable workflow to diagnose and troubleshoot Kubernetes pod and cluster issues, reducing mean time to resolution by guiding users through structured checks and commands.

Core Features & Use Cases

  • Diagnostic Commands Cheatsheet with practical kubectl commands for quick cluster health checks and pod diagnostics.
  • Pod Status & Lifecycle Troubleshooting covering CrashLoopBackOff, ImagePullBackOff, Pending, and OOMKilled scenarios.
  • Service, DNS, Ingress, and Networking debugging guidance, including log analysis, events, and quick fixes to common issues.
  • Standalone debugging utilities such as ephemeral debug pods, network reachability tests, and quick resource usage checks to reproduce problems.

Quick Start

Run the kubectl-based diagnostics outlined in this guide against the affected namespace or pod to begin rapid troubleshooting.

Frequently Asked Questions about k8s-debug

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I troubleshoot a Kubernetes pod stuck in CrashLoopBackOff?

Troubleshoot a Kubernetes pod in CrashLoopBackOff by running structured kubectl diagnostics to inspect pod events, analyze container logs, and identify failing startup commands or missing dependencies. This enables rapid root-cause analysis and resolution across affected namespaces.

What is the best way to diagnose Kubernetes OOMKilled pods and Pending states?

Diagnose Kubernetes OOMKilled and Pending pods by following a repeatable diagnostic workflow that checks resource limits, node capacity, and scheduling constraints using targeted kubectl commands. This ensures quick identification of memory exhaustion or resource starvation issues.

How do I debug Kubernetes DNS and networking problems across deployments?

Debug Kubernetes DNS and networking problems by executing guided diagnostic steps that include service connectivity checks, ephemeral debug pods, and network reachability tests. This structured approach isolates ingress routing failures and DNS resolution errors effectively.

Why does my Kubernetes deployment fail with ImagePullBackOff errors?

Kubernetes deployments fail with ImagePullBackOff errors due to incorrect image names, missing registry credentials, or inaccessible repositories. A systematic kubectl diagnostics workflow inspects pod events to verify image references and validate pull secrets for rapid troubleshooting.

Can I use kubectl commands to perform end-to-end Kubernetes cluster diagnostics?

You can use kubectl commands to perform end-to-end Kubernetes cluster diagnostics by following a structured workflow covering pod lifecycle, events, and resource usage checks. This provides safe, reusable guidance for comprehensive health assessments and rapid issue resolution.