mcp-server-kubernetes

MCP.Pizza Chef: Flux159

The mcp-server-kubernetes is an MCP server designed to facilitate Kubernetes management commands within the Model Context Protocol ecosystem. It acts as a lightweight adapter that exposes Kubernetes cluster operations and resources in a structured, model-readable format. This server enables LLMs and AI agents to interact with Kubernetes environments securely and efficiently, supporting real-time context updates and multi-step reasoning for cluster management tasks. It integrates seamlessly with MCP clients and hosts, providing developers with a powerful tool to automate, monitor, and manage Kubernetes clusters through AI-enhanced workflows.

Use This MCP server To

Manage Kubernetes clusters via AI-driven commands Automate deployment and scaling operations Monitor cluster health and resource usage Retrieve Kubernetes resource configurations Execute kubectl commands programmatically Integrate Kubernetes management into AI copilots Facilitate multi-step Kubernetes troubleshooting Enable real-time Kubernetes environment context for LLMs

README

MCP Server Kubernetes

CI Language Bun Kubernetes Docker Stars Issues PRs Welcome Last Commit smithery badge

MCP Server that can connect to a Kubernetes cluster and manage it.

MCPKubernetesClaude.mov

Usage with Claude Desktop

{
  "mcpServers": {
    "kubernetes": {
      "command": "npx",
      "args": ["mcp-server-kubernetes"]
    }
  }
}

The server will automatically connect to your current kubectl context. Make sure you have:

  1. kubectl installed and in your PATH
  2. A valid kubeconfig file with contexts configured
  3. Access to a Kubernetes cluster configured for kubectl (e.g. minikube, Rancher Desktop, GKE, etc.)
  4. Helm v3 installed and in your PATH (no Tiller required). Optional if you don't plan to use Helm.

You can verify your connection by asking Claude to list your pods or create a test deployment.

If you have errors open up a standard terminal and run kubectl get pods to see if you can connect to your cluster without credentials issues.

Usage with mcp-chat

mcp-chat is a CLI chat client for MCP servers. You can use it to interact with the Kubernetes server.

npx mcp-chat --server "npx mcp-server-kubernetes"

Alternatively, pass it your existing Claude Desktop configuration file from above (Linux should pass the correct path to config):

Mac:

npx mcp-chat --config "~/Library/Application Support/Claude/claude_desktop_config.json"

Windows:

npx mcp-chat --config "%APPDATA%\Claude\claude_desktop_config.json"

Features

  • Connect to a Kubernetes cluster
  • List all pods, services, deployments
  • List, Describe nodes
  • Create, describe, delete a pod
  • List all namespaces, create a namespace
  • Create custom pod & deployment configs, update deployment replicas
  • Create, describe, delete, update a service
  • Create, get, update, delete a ConfigMap
  • Get logs from a pod for debugging (supports pods, deployments, jobs, and label selectors)
  • Support Helm v3 for installing charts
    • Install charts with custom values
    • Uninstall releases
    • Upgrade existing releases
    • Support for namespaces
    • Support for version specification
    • Support for custom repositories
  • kubectl explain and kubectl api-resources support
  • Get Kubernetes events from the cluster
  • Port forward to a pod or service
  • Create, list, and decribe cronjobs
  • Non-destructive mode for read and create/update-only access to clusters

Local Development

Make sure that you have bun installed. Clone the repo & install dependencies:

git clone https://github.com/Flux159/mcp-server-kubernetes.git
cd mcp-server-kubernetes
bun install

Development Workflow

  1. Start the server in development mode (watches for file changes):
bun run dev
  1. Run unit tests:
bun run test
  1. Build the project:
bun run build
  1. Local Testing with Inspector
npx @modelcontextprotocol/inspector node dist/index.js
# Follow further instructions on terminal for Inspector link
  1. Local testing with Claude Desktop
{
  "mcpServers": {
    "mcp-server-kubernetes": {
      "command": "node",
      "args": ["/path/to/your/mcp-server-kubernetes/dist/index.js"]
    }
  }
}
  1. Local testing with mcp-chat
bun run chat

Contributing

See the CONTRIBUTING.md file for details.

Advanced

Additional Advanced Features

For more advanced information like using SSE transport, Non-destructive mode with ALLOW_ONLY_NON_DESTRUCTIVE_TOOLS, see the ADVANCED_README.md.

Architecture

See this DeepWiki link for a more indepth architecture overview created by Devin.

This section describes the high-level architecture of the MCP Kubernetes server.

Request Flow

The sequence diagram below illustrates how requests flow through the system:

sequenceDiagram
    participant Client
    participant Transport as StdioTransport
    participant Server as MCP Server
    participant Handler as Request Handler
    participant K8sManager as KubernetesManager
    participant K8s as Kubernetes API

    Client->>Transport: Send Request via STDIO
    Transport->>Server: Forward Request

    alt Tools Request
        Server->>Handler: Route to tools handler
        Handler->>K8sManager: Execute tool operation
        K8sManager->>K8s: Make API call
        K8s-->>K8sManager: Return result
        K8sManager-->>Handler: Process response
        Handler-->>Server: Return tool result
    else Resource Request
        Server->>Handler: Route to resource handler
        Handler->>K8sManager: Get resource data
        K8sManager->>K8s: Query API
        K8s-->>K8sManager: Return data
        K8sManager-->>Handler: Format response
        Handler-->>Server: Return resource data
    end

    Server-->>Transport: Send Response
    Transport-->>Client: Return Final Response
Loading

Publishing new release

Go to the releases page, click on "Draft New Release", click "Choose a tag" and create a new tag by typing out a new version number using "v{major}.{minor}.{patch}" semver format. Then, write a release title "Release v{major}.{minor}.{patch}" and description / changelog if necessary and click "Publish Release".

This will create a new tag which will trigger a new release build via the cd.yml workflow. Once successful, the new release will be published to npm. Note that there is no need to update the package.json version manually, as the workflow will automatically update the version number in the package.json file & push a commit to main.

Not planned

Authentication / adding clusters to kubectx.

mcp-server-kubernetes FAQ

How do I install the mcp-server-kubernetes?
You can install it by cloning the GitHub repository and following the setup instructions, which typically involve configuring access to your Kubernetes cluster and running the server with Bun runtime.
What Kubernetes operations can this MCP server perform?
It supports a wide range of Kubernetes management commands including deployments, scaling, resource querying, and executing kubectl commands programmatically.
Is the mcp-server-kubernetes secure to use in production?
Yes, it follows MCP principles for secure and scoped model interaction, ensuring safe access to Kubernetes clusters with proper authentication and authorization.
Can this server handle real-time updates from Kubernetes clusters?
Yes, it provides real-time context integration, allowing LLMs to receive up-to-date cluster state and resource information.
Does it support integration with multiple MCP hosts and clients?
Yes, it is designed to work seamlessly with various MCP clients and hosts, enabling flexible AI-enhanced Kubernetes workflows.
What runtime environment does the server require?
The server runs on Bun, a fast JavaScript runtime, ensuring efficient performance and easy deployment.
How does this server improve Kubernetes management with AI?
By exposing Kubernetes commands and state in a structured format, it enables LLMs to automate complex tasks, perform multi-step reasoning, and provide intelligent recommendations.
Where can I find the source code and contribute?
The source code is available on GitHub at the mcp-server-kubernetes repository, and contributions are welcome via pull requests.