MCP Server Kubernetes, developed by Flux159, functions as a critical bridge, enabling AI models and assistants to interact directly with Kubernetes and OpenShift clusters. Built with TypeScript and running on the Bun JavaScript runtime, this free tool extends the capabilities of Large Language Models that support the Model Context Protocol. It is distributed through GitHub repository cloning, npx execution, pre-built archives, or via Docker Desktop's MCP Toolkit Catalog. This server is designed to enhance cluster management through AI-driven interactions.

What does MCP Server Kubernetes connect to?

This MCP (Model Context Protocol) server connects to both Kubernetes and OpenShift clusters. Its core function is to act as an intermediary, allowing AI models and assistants to communicate with these container orchestration systems. By establishing this connection, MCP Server Kubernetes makes it possible for AI to understand and respond to cluster states, execute commands, and manage resources within these environments, extending AI capabilities into infrastructure management.

Core features and tools

MCP Server Kubernetes exposes several callable tools for cluster management. These include generic Kubernetes resource operations for CRUD actions on any resource, along with Pod-specific operations. It also supports listing Kubernetes Namespaces, viewing Kubernetes Events, and listing OpenShift Projects. Furthermore, it offers Helm v3 support for conversational chart management, and specific tools like kubectl_get for listing or getting resources, and kubectl_describe for resource details.

Who is it for?

This tool is for DevOps Engineers, developers, and operators who manage Kubernetes clusters and are looking to use AI assistants for improved management. It helps those who want to use natural language commands to interact with their cluster infrastructure. Compatible AI clients include GitHub Copilot (GPT-4.1), Claude Desktop, Cursor, and Visual Studio Code (with Copilot). This is not the right fit if you do not manage Kubernetes or OpenShift clusters.

How to get started

The server authenticates using permissions configured in the kubectl setup, inheriting context from the kubeconfig file. It supports loading kubeconfig from environment variables like KUBECONFIG_DATA or K8S_SERVER and K8S_TOKEN, and an mcpAuthToken can be used for HTTP/SSE transport when running without other authentication.

Use cases

• Diagnosing crash-looping pods by pulling logs and events through natural language commands.
• Automating routine Kubernetes management tasks such as listing namespaces, creating pods, scaling deployments, and fetching logs.
• Enabling AI-assisted GitOps for analyzing cluster state, troubleshooting deployment issues, and performing operations using conversational prompts.

MCP Server Kubernetes offers a valuable pathway for AI assistants to interact with Kubernetes and OpenShift environments. Its ability to perform various cluster operations and integrate with popular AI clients makes it a useful tool for enhancing operational efficiency. The reliance on existing kubectl configurations simplifies initial setup for many users.