adx-mcp-server

A Model Context Protocol (MCP) server for querying and analyzing Azure Data Explorer databases, designed to support time-series and analytics MCP use cases and enable AI assistants to interact with Azure Data Explorer through standardized interfaces.

Source: https://github.com/pab1it0/adx-mcp-server

Features

• Execute KQL queries: Run Kusto Query Language (KQL) queries against Azure Data Explorer databases.
• Database discovery and exploration:
  • List tables in the configured database
  • View table schemas
  • Sample data from tables (with optional sample size)
  • Get table statistics/details
• Authentication support:
  • Token credential support (Azure CLI, Managed Service Identity, etc.)
  • Workload Identity credential support for Azure Kubernetes Service (AKS)
  • Automatic fallback to DefaultAzureCredential if workload identity is not present
• Docker containerization: Includes Dockerfile and docker-compose setup for easy deployment and isolation.
• Configurable interactive tools for AI assistants: Select which tools (query execution, listing tables, schema viewing, etc.) are available to MCP clients.
• Development container support: Can be used as a dev container or in GitHub Codespaces for seamless development experience.
• Comprehensive test suite: Includes configuration validation, server functionality, error handling, and main application tests.

Provided Tools (APIs)

• execute_query: Execute a KQL query against Azure Data Explorer
• list_tables: List all tables in the configured database
• get_table_schema: Get the schema for a specific table
• sample_table_data: Get sample data from a table

Usage

• Configure environment variables for Azure Data Explorer cluster and database.
• Supports running directly, via Docker, or as a development container.
• Integrates with MCP clients (e.g., Claude Desktop) via configuration.

License

MIT

Pricing

Open-source software (no pricing; freely available under MIT license).