Kubernetes MCP Server

What it does

The Kubernetes MCP Server connects your AI coding assistant to your Kubernetes clusters via your local kubeconfig, enabling it to inspect pods, services, and deployments, read container logs, and apply manifests. This brings cluster management directly into your development environment.

This is a community-maintained MCP server that provides essential Kubernetes operations. It reads your existing kubeconfig for cluster authentication, so it works with any Kubernetes cluster you already have access to -- whether local (minikube, kind) or remote (EKS, GKE, AKS).

Use cases

• Inspect pod statuses and logs to debug application issues in your cluster
• View deployment rollout status to verify your latest changes are running
• List services and their endpoints to understand cluster networking
• Apply Kubernetes manifests to deploy updates directly from your editor
• Check resource health across namespaces during development and testing

Getting started

1. Ensure you have a valid kubeconfig at ~/.kube/config with access to your target cluster. Verify with kubectl cluster-info.
2. The MCP server uses your default kubeconfig context. Switch contexts with kubectl config use-context <context-name> before starting the server if needed.
3. No API keys or tokens are needed beyond your existing kubeconfig setup.
4. For production clusters, ensure your kubeconfig role bindings limit the scope of operations. Consider using a read-only ClusterRole for safety when exploring production environments.