How to install graphify so Claude Code queries a knowledge graph instead of grepping

graphify is an open-source tool that turns a project folder into a knowledge graph your agent can query instead of grepping through files. Code is parsed locally with tree-sitter, so no LLM is called; docs, PDFs and images get a second semantic pass from your assistant's model. Version 0.9.92 shipped on October 16, and the repo has passed 120,000 GitHub stars.

1. Install the CLI and register the skill. You need Python 3.10+ and uv (or pipx). The official PyPI package is graphifyy with a double y; other graphify* packages aren't affiliated.

uv tool install graphifyy
graphify install

For Codex add --platform codex, and to install into the current repository only, add --project.

graphify install --platform codex
graphify install --project

1. Build the graph from Claude Code. In Codex, call it as $graphify ..

/graphify .

You get graphify-out/ with graph.html (an interactive graph for the browser), GRAPH_REPORT.md (key concepts and suggested questions) and graph.json (the full graph).

1. Query the graph straight from the terminal.

graphify query "how requests are validated"
graphify path "FastAPI" "ModelField"
graphify explain "APIRouter"

1. To make your agent check the graph first, run this once in the project. In Claude Code, a hook fires before search-style tool calls and before reading source files one by one with Read or Glob, and nudges the agent toward the graph.

graphify claude install

Codex needs multi_agent = true under [features] in ~/.codex/config.toml for parallel extraction. The code graph is free and built locally; the pass over docs and media uses your assistant's model or an API key you configure.