Enables interaction with CodeQL through structured commands for tools and AI agents.
This project hosts a Model Context Protocol (MCP) server, acting as a wrapper for the CodeQL query server. It facilitates communication between CodeQL and external tools like Cursor or AI agents, enabling them to execute queries and access documentation through a structured command interface.
Key Features
01Registers CodeQL databases
02Locates predicate/class symbol positions
03Decodes `.bqrs` files into JSON format
04Runs full queries or quickly evaluates symbols
0555 GitHub stars
Use Cases
01Enabling natural language interaction with CodeQL databases
02Automating security vulnerability checks using AI agents
03Integrating CodeQL analysis into AI-powered code editors