Mkp
๐๏ธ โ๏ธ - MKP is a Model Context Protocol (MCP) server for Kubernetes that allows LLM-powered applications to interact with Kubernetes clusters. It provides tools for listing and applying Kubernetes resources through the MCP protocol.
What is Mkp?
Mkp is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to ๐๏ธ โ๏ธ - mkp is a model context protocol (mcp) server for kubernetes that allows llm-powered applications to interact with kubernetes clusters. it provides tools for listing and applying kubernetes re...
๐๏ธ โ๏ธ - MKP is a Model Context Protocol (MCP) server for Kubernetes that allows LLM-powered applications to interact with Kubernetes clusters. It provides tools for listing and applying Kubernetes resources through the MCP protocol.
This server falls under the Cloud Services category on MCPgee, the world's largest MCP server directory with 33,000+ servers.
Features
- ๐๏ธ โ๏ธ - MKP is a Model Context Protocol (MCP) server for Kub
Use Cases
Maintainer
Works with
Installation
Manual Installation
npx mkpConfiguration
Configuration Details
claude_desktop_config.json
Performance
Response Metrics
Resource Usage
How to Set Up and Use Mkp
MKP (Model Kubernetes Protocol) is a Go-based MCP server that lets AI assistants interact directly with Kubernetes clusters using natural language. It connects to your cluster via kubeconfig and exposes tools for listing, retrieving, applying, and deleting resources, as well as executing commands inside pods. Built-in rate limiting and opt-in write operations make it safe for use against production clusters, while support for label selector filtering and subresource access (logs, status, scale) makes it genuinely useful for day-to-day cluster operations.
Prerequisites
- Go 1.21 or higher and the task build tool installed
- A valid kubeconfig file with access to your Kubernetes cluster
- Git to clone the repository
- An MCP-compatible client such as Claude Desktop or Cursor
Clone the MKP repository
Clone the repository from GitHub and enter the project directory.
git clone https://github.com/StacklokLabs/mkp.git
cd mkpInstall and build the server
Use the task command to install dependencies and build the binary. If you don't have task installed, run 'go install github.com/go-task/task/v3/cmd/task@latest' first.
task install
task buildConfigure kubeconfig access
MKP uses your existing kubeconfig for authentication. Ensure KUBECONFIG is set or that ~/.kube/config exists and points to the cluster you want to manage.
export KUBECONFIG=/path/to/your/kubeconfigConfigure your MCP client
Add MKP to your MCP client configuration. By default, write operations are disabled for safety โ add --read-write=true only if you want to allow apply and delete operations.
{
"mcpServers": {
"mkp": {
"command": "/path/to/mkp/bin/mkp",
"args": ["--read-write=false"],
"env": {
"KUBECONFIG": "/path/to/your/kubeconfig"
}
}
}
}Adjust rate limits if needed
The default rate limits are 120 requests/minute for reads and 30 for writes. Override with environment variables if your use case requires higher throughput.
export MKP_RATE_LIMIT_READ=200
export MKP_RATE_LIMIT_WRITE=60Mkp Examples
Client configuration
Full configuration for claude_desktop_config.json with read-only access to a cluster
{
"mcpServers": {
"mkp": {
"command": "/Users/yourname/mkp/bin/mkp",
"args": ["--read-write=false", "--kubeconfig=/Users/yourname/.kube/config"],
"env": {
"MKP_RATE_LIMIT_READ": "120",
"MKP_RATE_LIMIT_DEFAULT": "60"
}
}
}
}Prompts to try
Example prompts mapped to the five MKP tools (get_resource, list_resources, apply_resource, post_resource, delete_resource)
- "List all pods in the production namespace and show me which ones are not Running"
- "Get the status and events for deployment nginx-deployment in the default namespace"
- "Show me the logs for pod api-server-xyz in the backend namespace"
- "List all services with the label app=frontend across all namespaces"
- "Execute 'df -h' inside the container 'app' in pod worker-001 in the default namespace"
- "Apply this ConfigMap manifest to the staging namespace"Troubleshooting Mkp
'task' command not found when building
Install the task build tool with: go install github.com/go-task/task/v3/cmd/task@latest. Alternatively, you can build directly with: go build -o bin/mkp ./cmd/mkp
Permission denied or 'Forbidden' errors when listing resources
MKP uses the permissions of the user or service account in your kubeconfig. Ensure that account has RBAC access to the resources you're querying. For cluster-wide listing, it needs ClusterRole permissions.
apply_resource or delete_resource tools are not available or return errors
Write operations are disabled by default. Add --read-write=true to the args in your MCP config to enable them. Use this flag only against clusters where you intend to make changes.
Frequently Asked Questions about Mkp
What is Mkp?
Mkp is a Model Context Protocol (MCP) server that ๐๏ธ โ๏ธ - mkp is a model context protocol (mcp) server for kubernetes that allows llm-powered applications to interact with kubernetes clusters. it provides tools for listing and applying kubernetes resources through the mcp protocol. It connects AI assistants to external tools and data sources through a standardized interface.
How do I install Mkp?
Follow the installation instructions on the Mkp GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.
Which AI clients work with Mkp?
Mkp works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.
Is Mkp free to use?
Yes, Mkp is open source and available under the Apache-2.0 license. You can use it freely in both personal and commercial projects.
Mkp Alternatives โ Similar Cloud Services Servers
Looking for alternatives to Mkp? Here are other popular cloud services servers you can use with Claude, Cursor, and VS Code.
Open WebUI
โ 138.2kUser-friendly AI Interface (Supports Ollama, OpenAI API, ...)
Anything LLM
โ 60.4kThe all-in-one AI productivity accelerator. On device and privacy first with no annoying setup or configuration.
LocalAI
โ 46.4kLocalAI is the open-source AI engine. Run any model - LLMs, vision, voice, image, video - on any hardware. No GPU required.
Nacos
โ 33.0kan easy-to-use dynamic service discovery, configuration and service management platform for building AI cloud native applications.
Xiaozhi ESP32
โ 26.7kๆฌ้กน็ฎไธบxiaozhi-esp32ๆไพๅ็ซฏๆๅก๏ผๅธฎๅฉๆจๅฟซ้ๆญๅปบESP32่ฎพๅคๆงๅถๆๅกๅจใBackend service for xiaozhi-esp32, helps you quickly build an ESP32 device control server.
Gateway
โ 11.8kA blazing fast AI Gateway with integrated guardrails. Route to 1,600+ LLMs, 50+ AI Guardrails with 1 fast & friendly API.
Browse More Cloud Services MCP Servers
Explore all cloud services servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.
Set Up Mkp in Your Editor
Choose your AI client for step-by-step setup instructions.
Quick Config Preview
Add this to your claude_desktop_config.json or .cursor/mcp.json
Ready to use Mkp?
Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.