EKS Model Context Protocol (MCP) Server
About
# EKS Model Context Protocol (MCP) Server A lightweight, efficient server that implements the Model Context Protocol for EKS operations. This server provides a standardized interface for GenAI agents to interact with EKS clusters. ## Features - Fast response times with proper timeout handling - Comprehensive EKS…
Details
- License
- MIT
Explore
- Fast response times with proper timeout handling
- Comprehensive EKS operations support
- Kubernetes resource management (pods, services, deployments, etc.)
- Multiple authentication methods for EKS clusters
- Robust error handling and logging
- Docker containerization for easy deployment
- ECS deployment support
Setting up with Highlight
This MCP is not yet compatible with Highlight’s one-click setup. However, you can still use it with Highlight by following these steps:
- Download and install Highlight from highlightai.com/download
- Navigate to the plugins tab and select "Add Custom Plugin"
-
Configure the plugin with the settings below
Plugin Name
EKS Model Context Protocol (MCP) ServerCommand (node, npx, python, etc.)Please refer to the README for specific instructions on how to obtain API keys or other required environment variables.
- Enable "Start Automatically" if you want the plugin to start when Highlight launches
From the repository
- Docker
- AWS credentials (for production use)
- Python 3.8+ (for local development)
The server supports multiple authentication methods for EKS clusters:
1. AWS SDK Authentication: Uses the AWS SDK to authenticate with EKS and get cluster information
2. Kubectl with Generated Kubeconfig: Creates a temporary kubeconfig file and uses kubectl
3. Direct Kubernetes API Calls: Makes direct API calls to the Kubernetes API server
The current implementation uses the Kubectl method with proper authentication, which provides the most reliable results.
pip install -r requirements.txt
See the ECS Deployment README for detailed instructions.
cd ecs-deployment
./deploy.sh
./update-service.sh
The project includes a client implementation that uses Amazon Bedrock for conversational interaction with the MCP server. See the Client README for details.
For the server to authenticate with EKS clusters, the ECS task role needs to be added to the EKS cluster's aws-auth ConfigMap. See the ECS Deployment README for detailed instructions.
The server is currently deployed to ECS and accessible at:
- Endpoint: http://3.90.45.69:8000
- Health check: http://3.90.45.69:8000/health
- Operations discovery: http://3.90.45.69:8000/mcp/v1/operations
Claude Desktop / Cursor
Paste into your MCP client config file to install this server.
{
"mcpServers": {
"eks model context protocol (mcp) server": {
"eks-mcp-server-new": {
"command": "python",
"args": [
"main.py"
]
}
}
}
}McpServers
{
"eks-mcp-server-new": {
"command": "python",
"args": [
"main.py"
]
}
}A lightweight, efficient server that implements the Model Context Protocol for EKS operations. This server provides a standardized interface for GenAI agents to interact with EKS clusters.
Features
- Fast response times with proper timeout handling
- Comprehensive EKS operations support
- Kubernetes resource management (pods, services, deployments, etc.)
- Multiple authentication methods for EKS clusters
- Robust error handling and logging
- Docker containerization for easy deployment
- ECS deployment support
Project Structure
- main.py: Main server implementation with FastAPI
- eks_operations.py: EKS API operations implementation
- k8s_operations.py: Kubernetes API operations implementation
- k8s_operations_sdk_v4.py: SDK-based Kubernetes operations implementation
- k8s_operations_kubectl.py: Kubectl-based Kubernetes operations with proper authentication
- k8s_auth_config.py: Authentication configuration for Kubernetes
- direct_k8s_client.py: Direct Kubernetes API client implementation
- test_*.py: Various test scripts for different implementations
- Dockerfile and Dockerfile.amd64: Docker configuration files
- docker-compose.yml: Docker Compose configuration
- requirements.txt: Python dependencies
- client/: Client implementation directory (Client README)
- ecs-deployment/: ECS deployment scripts and configuration (ECS Deployment README)
API Endpoints
- GET /health - Health check endpoint
- GET /mcp/v1/operations - List available operations
- POST /mcp/v1/query - Execute MCP operations
Supported Operations
EKS Cluster Operations
-list_clusters - List EKS clusters in a region
- describe_cluster - Get detailed information about a cluster
- list_nodegroups - List nodegroups for a cluster
- describe_nodegroup - Get detailed information about a nodegroup
Kubernetes Operations
-list_namespaces - List Kubernetes namespaces
- list_pods - List pods in a namespace
- describe_pod - Get detailed information about a pod
- get_deployments - List deployments in a namespace
- describe_deployment - Get detailed information about a deployment
- get_services - List services in a namespace
- describe_service - Get detailed information about a service
- get_pod_logs - Get logs from a pod
Authentication Methods
The server supports multiple authentication methods for EKS clusters:
1. AWS SDK Authentication: Uses the AWS SDK to authenticate with EKS and get cluster information
2. Kubectl with Generated Kubeconfig: Creates a temporary kubeconfig file and uses kubectl
3. Direct Kubernetes API Calls: Makes direct API calls to the Kubernetes API server
The current implementation uses the Kubectl method with proper authentication, which provides the most reliable results.
Getting Started
Prerequisites
- Docker
- AWS credentials (for production use)
- Python 3.8+ (for local development)
Running Locally
```bash
Sign in to leave a review
Use Google, GitHub, or an email account so ratings stay tied to real people.
No reviews posted yet.
