I built AI Guide because I noticed a gap in how AI coding tools work with large codebases. Coding agents are getting very good at writing and modifying code, but before making a change, they still need to figure out where the relevant code lives and what is connected to it. AI Guide takes a different approach: it analyzes a repository and builds a deterministic Unified Graph of files, symbols, and relationships. Developers and coding agents can then query that graph to find code, locate definitions, understand dependencies, and trace paths across the codebase. V1 is currently open source and supports Python, TypeScript/TSX, JavaScript/JSX, C++, and Java. Current validation: • 41 tests passing • 185 entities • 282 relationships • 0 failed files • Tested on a real project I'd love to get feedback from developers building with AI coding tools. What would you want an AI coding agent to understand about your codebase before it starts making changes?

Ai-Guide

Give AI coding agents a map of your codebase.

Give AI coding agents a map of your codebase.