Kubernetes MCP Server
A native Go MCP server that lets AI assistants operate Kubernetes and OpenShift clusters directly through the API server: generic resource CRUD, pod operations, Helm releases, and Tekton pipeline management.
Kubernetes MCP Server is a Model Context Protocol server that gives an AI assistant direct access to a Kubernetes or OpenShift cluster. It automatically detects the current kubeconfig or in-cluster configuration, supports create, update, get, list and delete operations on any Kubernetes or OpenShift resource type, and provides pod-specific operations such as listing, viewing logs, getting resource usage metrics, executing a command inside a pod, and running a new container image.
It also lists namespaces and cluster events, lists OpenShift projects, installs, lists and uninstalls Helm releases, and starts, restarts, cancels and retrieves logs for Tekton pipelines and tasks. Optional OpenTelemetry tracing and metrics, including a statistics endpoint, provide observability into server usage.
Unlike some other Kubernetes MCP servers, it is a native Go implementation that talks to the Kubernetes API server directly rather than wrapping the kubectl or helm command-line tools, so it needs no external tool dependencies installed. It is released under the Apache-2.0 license.