Graphify
Turn any codebase into a queryable knowledge graph — ask questions, trace paths, and give your coding agent shared context, parsed locally.
Project intro
Graphify turns a codebase — along with its docs, SQL schemas, configs, and PDFs — into a knowledge graph you can query instead of reading file by file. It ships as a skill for Claude Code, Cursor, Codex, and Gemini CLI: ask a plain-language question, explain one concept, or trace the shortest path between any two parts of the code.
The graph is built with local, deterministic parsing across about 40 languages, and every edge carries a confidence tag — extracted means it was read straight from the source, inferred means it was worked out by resolution — so you can tell what the tool saw from what it guessed. There is no vector store, and the code never leaves your machine; only the optional semantic pass over docs and media calls a backend, and only if you configure one.
Nearly 150,000 stars on GitHub, with over 14,000 forks, 11.1 million PyPI downloads, and 15.2 million repositories indexed, according to the project's homepage. Version 0.9.112 shipped in November 2026, and the project reports its own benchmark numbers: 0.497 recall@10 on LOCOMO with zero LLM credits spent building the graph.
Two honest limits: the free open-source engine builds the graph on your machine, while the hosted Graphify Cloud — continuous multi-repo graphs, cross-repo links, and review tooling — is the paid side. And while code parsing is fully local, the docs-and-media pass needs a backend you set up yourself.
Actions
Tech stack
Project info
- Project category
- Developer Tools
- Project status
- Online
- Collaboration intent
- None