Diagnoses and resolves Kubernetes cluster issues using systematic workflows and MCP-powered kubectl tools.
The Kubernetes Debugger skill empowers Claude to systematically troubleshoot complex cluster environments by providing structured workflows for diagnosing failing pods, unreachable services, and unhealthy nodes. It leverages MCP tools to perform deep-dive inspections, including log analysis, event tracking, and container execution, helping developers quickly identify the root causes of common issues like CrashLoopBackOff, ImagePullBackOff, OOMKilled errors, and network misconfigurations. Whether managing a single deployment or an entire node pool, this skill provides the decision-making logic needed to restore cluster health efficiently.
Key Features
01Comprehensive pod diagnostics for CrashLoopBackOff, Pending, and ImagePullBackOff states
02Deployment rollout monitoring with automated rollback capabilities
032 GitHub stars
04Node-level health assessment and maintenance tools like cordon and drain
05Service and network connectivity verification with DNS and endpoint checks
06Detailed container log analysis including previous-run logs for crashed instances
Use Cases
01Identifying the root cause of application crashes and scheduling failures in production
02Debugging service-to-service communication issues and internal cluster DNS resolution
03Investigating resource bottlenecks, CPU throttling, and memory-related container terminations
