Quick Start Guide

This guide will help you get Kubernetes Claude MCP up and running quickly. We’ll use Docker Compose for a simple local deployment.

Prerequisites

• Docker and Docker Compose
• A Kubernetes cluster with kubectl configured
• ArgoCD instance (optional)
• GitLab access (optional)
• Claude API key from Anthropic

Step 1: Clone the Repository

git clone https://github.com/blankcut/kubernetes-mcp-server.git
cd kubernetes-mcp-server

Step 2: Configure Environment Variables

Create a .env file in the project directory:

# Required
CLAUDE_API_KEY=your_claude_api_key

# Optional ArgoCD credentials
ARGOCD_URL=http://argocd.your-domain.com
ARGOCD_USERNAME=admin
ARGOCD_PASSWORD=your_password
# Or use a token instead
# ARGOCD_AUTH_TOKEN=your_argocd_token

# Optional GitLab credentials
GITLAB_URL=https://gitlab.com
GITLAB_AUTH_TOKEN=your_gitlab_token

# Security
API_KEY=your_chosen_api_key_for_server_access

Step 3: Create Configuration File

Create a config.yaml file:

server:
  address: ":8080"
  readTimeout: 30
  writeTimeout: 60
  auth:
    apiKey: "${API_KEY}"

kubernetes:
  kubeconfig: ""
  inCluster: false
  defaultContext: ""
  defaultNamespace: "default"

argocd:
  url: "${ARGOCD_URL}"
  authToken: "${ARGOCD_AUTH_TOKEN}"
  username: "${ARGOCD_USERNAME}"
  password: "${ARGOCD_PASSWORD}"
  insecure: true

gitlab:
  url: "${GITLAB_URL}"
  authToken: "${GITLAB_AUTH_TOKEN}"
  apiVersion: "v4"
  projectPath: ""

claude:
  apiKey: "${CLAUDE_API_KEY}"
  baseURL: "https://api.anthropic.com"
  modelID: "claude-3-haiku-20240307"
  maxTokens: 4096
  temperature: 0.7

Step 4: Start the Server

Run the server using Docker Compose:

docker-compose up -d

Step 5: Verify the Installation

Check if the server is running:

curl http://localhost:8080/api/v1/health

You should see a JSON response with the server status and service availability.

Step 6: Make Your First API Call

Let’s query a pod in your Kubernetes cluster:

curl -X POST \
  -H "Content-Type: application/json" \
  -H "X-API-Key: your_chosen_api_key" \
  -d '{
    "action": "queryResource",
    "resource": "pod",
    "name": "your-pod-name",
    "namespace": "your-namespace",
    "query": "What is the status of this pod and are there any issues?"
  }' \
  http://localhost:8080/api/v1/mcp/resource

The server will analyze the pod and return Claude’s analysis, highlighting any issues and providing recommendations.

Next Steps

Now that you have Kubernetes Claude MCP running, you can:

1. Configure the server for your specific environment
2. Explore the API to learn about all available endpoints
3. Check out the Troubleshooting Resources guide for common use cases
4. Learn about GitOps Integration with ArgoCD and GitLab

For a more detailed setup and configuration, see the Installation Guide.