K8s MCP Server G9i

🚀 K8s MCP Server Configuration Guide

K8s MCP Server is a tool designed to simplify Kubernetes operations. It supports various functions such as executing kubectl commands, deploying Helm Charts, managing Istio service meshes, and integrating with ArgoCD for application management.

🚀 Quick Start

Launch the K8s MCP Server using the following command:

docker run -i --rm ghcr.io/alexei-led/k8s-mcp-server:latest

Interact with the MCP Server using natural language. For example:

flowchart TB
    User->MCP_Server: "Show all pods in the default namespace"
    MCP_Server->K8s_API: kubectl get pods -n default
    K8s_API->MCP_Server: Return the pod list
    MCP_Server->User: Display the results and provide explanations

✨ Features

K8s MCP Server is a tool for simplifying Kubernetes operations, supporting the following features:

• kubectl Command Execution: Parse user intentions through natural language understanding (NLU) and execute corresponding kubectl commands.
• Helm Chart Deployment: Automatically parse Helm charts and generate deployment commands.
• Istio Service Mesh Management: Support the configuration and status check of Istio service meshes.
• ArgoCD Application Management: Integrate ArgoCD for application delivery and version control.

💻 Usage Examples

Basic Usage

Querying Cluster Information

flowchart TB
    subgraph "Basic Commands"
        A1["Show all pods in the default namespace"]
        A2["Get all services in all namespaces"]
        A3["Show the logs of the nginx pod"]
    end

Troubleshooting

flowchart TB
    subgraph "Troubleshooting"
        B1["Why isn't my deployment starting?"]
        B2["Describe the failed pod and explain the error"]
        B3["Check if my service is correctly connected to the pod"]
    end

Advanced Usage

Specifying the Target Cluster and Namespace

Add the following configuration to config.json:

{
  "mcpServers": {
    "kubernetes": {
      "command": "docker",
      "args": [
        "run", "-i", "--rm",
        "-v", "/Users/YOUR_USER_NAME/.kube:/home/appuser/.kube:ro",
        "-e", "K8S_CONTEXT=production-cluster",
        "-e", "K8S_NAMESPACE=my-application",
        "-e", "K8S_MCP_TIMEOUT=600",
        "ghcr.io/alexei-led/k8s-mcp-server:latest"
      ]
    }
  }
}

Connecting to an AWS EKS Cluster

{
  "mcpServers": {
    "kubernetes": {
      "command": "docker",
      "args": [
        "run", "-i", "--rm",
        "-v", "/Users/YOUR_USER_NAME/.kube:/home/appuser/.kube:ro",
        "-v", "/Users/YOUR_USER_NAME/.aws:/home/appuser/.aws:ro",
        "-e", "AWS_PROFILE=production",
        "-e", "AWS_REGION=us-west-2",
        "ghcr.io/alexei-led/k8s-mcp-server:latest"
      ]
    }
  }
}

Connecting to an Azure Kubernetes Service (AKS)

{
  "mcpServers": {
    "kubernetes": {
      "command": "docker",
      "args": [
        "run", "-i", "--rm",
        "-v", "/Users/YOUR_USER_NAME/.kube:/home/appuser/.kube:ro",
        "-e", "K8S_CONTEXT=aks-cluster",
        "ghcr.io/alexei-led/k8s-mcp-server:latest"
      ]
    }
  }
}

Enabling Logging

Add the following content to config.json to enable debug mode:

{
  "logging": {
    "level": "debug",
    "file": "mcp_server.log"
  }
}

⚠️ Important Note

• Authentication: Authenticate through Kubernetes service accounts or IAM roles.
• Authorization: Use RBAC policies to restrict the access permissions of the MCP Server.
• Network Isolation: Deploy the MCP Server in a trusted network environment.

📄 License

This project is licensed under the MIT License. The specific terms are as follows:

Copyright (c) 2023, Alexei Ledovskoi
Permission is granted to use, modify and distribute this software under the MIT License.

📞 Contact Information

• GitHub: alexei-ledovskoi/k8s-mcp-server
• Email: support@k8s-mcp-server.com
• Community Forum: Discussions