Adx MCP Server

🚀 Azure Data Explorer MCP Server

An Azure Data Explorer server supporting the Model Context Protocol (MCP).

This server provides standard MCP interface access to your Azure Data Explorer clusters and databases, enabling AI assistants to execute KQL queries and explore data.

✨ Features

• [x] Execute KQL queries against Azure Data Explorer.
• [x] Discover and explore database resources:
  • [x] List tables in the configured databases.
  • [x] View table schemas.
  • [x] Get sample table data (optionally specify the number of samples).
• [x] Support authentication:
  • [x] Token credential support (Azure CLI, MSI, etc.).
• [x] Support Docker containerization.
• [x] Provide interactive tools for AI assistants.

The list of tools is configurable, allowing you to choose which tools to expose to MCP clients. This is useful when you don't use certain features or don't want to occupy too much of the context window.

🚀 Quick Start

💻 Usage Examples

Basic Usage

1. Log in to your Azure account with ADX cluster permissions using the Azure CLI.

2. Configure environment variables via the .env file or system environment variables to specify your ADX cluster:

# Required: Azure Data Explorer configuration
ADX_CLUSTER_URL=https://yourcluster.region.kusto.windows.net
ADX_DATABASE=your_database

3. Add the server configuration to the client configuration file. For example, for Claude Desktop:

{
  "mcpServers": {
    "adx": {
      "command": "uv",
      "args": [
        "--directory",
        "<Full path to your adx-mcp-server directory>",
        "run"
      ],
      "environment": {
        "ADX_CLUSTER_URL": "https://yourcluster.region.kusto.windows.net",
        "ADX_DATABASE": "your_database"
      }
    }
  }
}

Advanced Usage

Docker Support

Build and run the container with the following commands:

# Build the Docker image
docker build -t adx-mcp-server .

# Run the container
docker run --name adx-mcp-server -p 8000:8000 adx-mcp-server

Configure the server URL in Claude Desktop as follows:

{
  "mcpServers": {
    "adx": {
      "url": "http://localhost:8000"
    }
  }
}

🔧 Technical Details

Tools

Project Structure

.
├── src/
│   ├── server.py          # MCP server implementation
│   └── main.py           # Main application logic
├── Dockerfile            # Docker configuration file
├── docker-compose.yml    # Docker Compose configuration
├── .dockerignore         # Docker ignore file
├── pyproject.toml        # Project configuration file
└── README.md             # This file

Testing

The project includes a comprehensive test suite to ensure functionality and prevent regressions. Run the tests:

# Install development dependencies
uv pip install -e ".[dev]"

# Run the tests
pytest

# Generate a coverage report
pytest --cov=src --cov-report=term-missing

📄 License

MIT