MCP Kubernetes Server

mcp - kubernetes - server is an MCP protocol server that serves as a bridge between AI assistants and Kubernetes clusters, enabling Kubernetes operations through natural language interaction.

Developer tools Cloud platform #Kubernetes #AI assistant #Cluster management #Natural language .Python

rating : 2.5 points

downloads : 19

update time : 2025-04-23

What is MCP Kubernetes Server?

MCP Kubernetes Server is a protocol server that allows AI assistants (such as Claude, Cursor, etc.) to interact with Kubernetes clusters through natural language. It converts users' natural language requests into Kubernetes operations and returns the results in a format that AI tools can understand.

How to use MCP Kubernetes Server?

Simply configure the server information in the supported AI clients, and then you can query or manage Kubernetes clusters using natural language. It supports two installation methods: Docker and UVX.

Applicable scenarios

Suitable for users who need to quickly query the status of Kubernetes clusters, diagnose problems, or perform simple management operations through natural language, especially developers and运维人员 who are not Kubernetes experts but need to interact with the clusters.

Main features

Query Kubernetes resourcesSupports querying the status of various resources in the cluster, such as pods, services, deployments, etc.

Execute kubectl and helm commandsCan directly execute kubectl and helm commands and return the results.

Diagnose cluster issuesGet pod logs, events, and status information to help diagnose problems.

Cluster managementSupports basic cluster management operations, such as scaling, rolling updates, etc.

Advantages and limitations

Advantages

No need to memorize complex kubectl commands. You can operate Kubernetes using natural language.

Integrates with multiple AI tools, such as Claude, Cursor, etc.

Provides rich query and management functions, covering most daily Kubernetes operations.

Supports multiple installation methods, including Docker and UVX.

Limitations

Complex operations may still require direct use of kubectl.

Requires pre - configuration of the kubeconfig file.

Some advanced functions may require additional permissions.

How to use

Installation preparation

Ensure that Docker or uv is installed and the kubeconfig file is ready.

Install via Docker

Run the server using Docker and mount the kubeconfig file.

Install via UVX

Run the server through uvx and set the KUBECONFIG environment variable.

AI client configuration

Configure the MCP server information in AI clients (such as Claude, Cursor).

Usage examples

Check cluster statusQuickly understand the overall health status of the cluster

Diagnose pod issuesFind out the reason why the pod cannot run normally

Scale the applicationAdjust the number of application replicas

Frequently Asked Questions

What permissions do I need to use this server?

Which AI clients are supported?

How to limit the operation permissions of the server?

Related resources

GitHub repository

Project source code and latest documentation

Kubernetes official documentation

Kubernetes official documentation and tutorials

MCP protocol

Model Context Protocol specification

🚀 MCP Kubernetes Server Project Documentation

The MCP Kubernetes Server is a tool designed for interacting with Kubernetes clusters, supporting communication via the MCP protocol. This project aims to simplify Kubernetes resource management and offers a wide range of features for operating workloads and other resources within the cluster.

🚀 Quick Start

Install the Command - Line Tool

pip install mcp-kubernetes-server

Initialize the Configuration

Create a configuration file named k8s-config.yaml:

apiVersion: v1
kind: MCPConfig
metadata:
  name: default
spec:
  kubernetes:
    server: http://localhost:6060 # Replace with your Kubernetes API server address
    inCluster: false
    namespace: default

Start the MCP Service

uv run -m src.mcp_kubernetes_server.main

✨ Features

Resource Operation Tools

The project provides various tools for operating Kubernetes resources, including:

Read Tools: Used to retrieve resource information.
Write Tools: Used to create, update, and delete resources.
Query Tools: Used to search and filter resources.

Detailed Feature Description

Read Tools

Tool Name	Description	Parameters
k8s_get	Retrieve a single resource	resource_type, name, namespace
k8s_list	Retrieve multiple resources	resource_type, label_selector, namespace
k8s_watch	Monitor resource changes in real - time	resource_type, name, namespace

Write Tools

Tool Name	Description	Parameters
k8s_create	Create a new resource	manifest_file
k8s_apply	Apply a configuration file to the cluster	manifest_file, namespace
k8s_replace	Update an existing resource	patch_file, resource_type, name, namespace

Query Tools

Tool Name	Description	Parameters
k8s_search	Search for resources that meet specific criteria	keyword, resource_type
k8s_describe	Display detailed information about a resource	resource_type, name, namespace

Tool Classification

Read Tools

Tool Name	Description	Parameters
k8s_get	Retrieve a single resource	`resource_type`, `name`, `namespace`
k8s_list	Retrieve multiple resources	`resource_type`, `label_selector`, `namespace`
k8s_watch	Monitor resource changes in real - time	`resource_type`, `name`, `namespace`

Write Tools

Tool Name	Description	Parameters
k8s_create	Create a new resource	`manifest_file`
k8s_apply	Apply a configuration file to the cluster	`manifest_file`, `namespace`
k8s_replace	Update an existing resource	`patch_file`, `resource_type`, `name`, `namespace`

Delete Tools

Tool Name	Description	Parameters
k8s_delete	Delete a specified resource	`resource_type`, `name`, `namespace`

📚 Documentation

Run the Project

uv run -m src.mcp_kubernetes_server.main

Debug with the Inspector Tool

# Start the Inspector tool for debugging
uv run --inspector http://localhost:6060 -m src.mcp_kubernetes_server.main

🤝 Contribution Guidelines

Submit Code

Fork the project repository on GitHub.
Create a new feature branch: git checkout -b feature/your-feature-name
Commit your code and push it to the remote repository: git push origin feature/your-feature-name
Submit a Pull Request to the main repository.

Report Issues

If you encounter any issues, please submit them in the Issues section of the GitHub repository.
Provide detailed error information and reproduction steps to facilitate quick issue location.