Graphify is a skill you install into your coding agent. Point it at a folder, run /graphify ., and it turns the whole thing into a queryable knowledge graph instead of a pile of files your agent greps through blindly. 85.9k stars, 8.5k forks, +1,858 today — top of GitHub Trending.
The trick is that parsing is dumb on purpose. Tree-sitter AST across ~40 languages, deterministic, zero LLM calls, nothing leaves your machine. Only the semantic pass over docs, PDFs, images and video hits a model, and it’s your own backend. Founder Safi Shamsi built it, YC S26, MIT licensed.
What you actually get
Three artifacts: an interactive HTML graph, a markdown report of key concepts, and a JSON graph you can query programmatically. Plus god nodes (the most-connected concepts), Leiden community detection, cross-file relationship resolution, and confidence tags that separate extracted edges from inferred ones — so you know what the tool saw versus what it guessed.
The agent hook
It installs into 20+ assistants — Claude Code, Cursor, Codex, Copilot, Gemini CLI. Then you ask in English: graphify query "what connects auth to the database?" The agent traverses a graph instead of burning 50k tokens re-reading files. That’s the real pitch. Coding agents don’t fail because the model is weak; they fail because the codebase doesn’t fit in context. Graphify is a map, not a bigger window.
You Might Also Like
- Claude Replay Turns Your Anthropic Claude Code Sessions Into Shareable Video Like Replays
- Openviking Treats ai Agent Memory Like a File System and 9k Github Stars say its Working
- 27k Github Stars in Weeks Learn Claude Code by Shareai lab Breaks Down ai Coding Agents Into 12 Lessons
- From 739k to 15k Tokens how Code Review Graph Slashes Claude Code Costs With a Local Knowledge Graph
- Oh my Claudecode Turns Claude Code Into a 32 Agent dev Team and 11 8k Stars Agree
Top AI Product 
Leave a comment