🚀 AI Assistant for Local Docker & Kubernetes (MCP Agent)

This project connects an AI Assistant (like Claude Desktop) directly to your local Docker and Kubernetes environment.

Instead of manually running terminal commands, copying logs, and pasting them into a chat window, the AI agent can run commands, check container logs, and inspect cluster health for you in real time—all with your approval.

🌟 Why Use This?

• No More Copy-Pasting Logs: The AI checks live container and cluster status directly.

• Safe & Secure: The AI asks for your permission before running any command on your machine.

• Easy Debugging: Helps you troubleshoot local DevOps setups faster.

Related MCP server: Docker & Kubernetes MCP Server

📁 Project Files

AGENTIC/
├── mcp_server.py          # The backend server that runs Docker/K8s commands
├── agent_with_mcp.py      # Script to run and test the MCP agent
├── agent.py               # Main agent execution script
├── generative_ai.py       # Helper functions for the AI model
├── requirements.txt       # Python libraries needed for this project
└── README.md              # Project documentation

⚡ Quick Setup Guide

1. Clone the Repository & Set Up Python

# Clone this repository

# Create a virtual environment
python -m venv .venv

# Activate the virtual environment
# On Windows:
.venv\Scripts\activate
# On Mac/Linux:
source .venv/bin/activate

# Install required packages
pip install -r requirements.txt

⚙️ How to Connect Claude Desktop to Your Local MCP Server

Follow these steps to open the configuration file directly from Claude Desktop:

1. Open Claude Desktop on your computer.

2. Click on the Claude Menu (top left on macOS or via your Profile/Settings on Windows).

3. Select Settings... from the dropdown menu.

4. Go to the Developer tab in the left sidebar.

5. Click the Edit Config button. This will automatically open or locate the claude_desktop_config.json file in your system text editor.

6. Paste the following configuration block into the file:

{
  "mcpServers": {
    "docker-ops-server": {
      "command": "C:\\path\\to\\your\\AGENTIC\\.venv\\Scripts\\python.exe",
      "args": [
        "C:\\path\\to\\your\\AGENTIC\\mcp_server.py"
      ]
    }
  }
}

⚠️ Important: Replace C:\path\to\your\AGENTIC with the actual full path to your project folder. Make sure you point directly to the Python interpreter inside .venv to prevent connection errors.

7. Save the file and completely restart Claude Desktop for the changes to take effect.

🚀 How to Use

1. Make sure Docker Desktop is running on your machine.

2. Open Claude Desktop.

3. Look for the hammer / tool icon at the bottom of the chat box—this confirms your local MCP tools are active.

4. Ask Claude questions like:

   • "Show me all running Docker containers."

   • "Check the health of my Kubernetes nodes."

5. Click Allow when Claude requests permission to run a tool!

🛠️ Common Issue & Fix

• Error: Server disconnected in Claude Desktop

  • Fix: Open Settings → Developer in Claude Desktop and click View Logs to check the error stream. Double-check that your path to .venv\Scripts\python.exe in claude_desktop_config.json is completely accurate.