MCP KQL Server

Example Use Cases

• 1

  Convert natural language requests into validated KQL and execute them against Azure Data Explorer clusters.

• 2

  Automated discovery and local caching of database/table schemas plus AI-generated table context to improve query correctness.

• 3

  Structured query results (JSON/CSV/table) and concise summaries or visualizations for downstream processing or presentation.

Description

This server converts natural-language prompts to KQL, runs raw KQL queries, and returns results in JSON/CSV/table formats while automatically discovering and caching table schemas. It provides AI-driven table context, schema analysis reports, and visualization-ready summaries to speed query development and improve accuracy. Authentication leverages the Azure CLI (no stored credentials) and schema memory is stored locally to reduce API calls and provide richer context for agents. The project is designed for fast integration with MCP-compatible clients like Claude or VS Code.