graphify

graphify by Graphify-Labs turns the code, documents and media in a project into a queryable knowledge graph, and comes as a command-line tool, an MCP server and a /graphify skill. It works with Claude, Cursor, Codex, Gemini and others, on macOS, Linux and Windows. The README says code is parsed locally with tree-sitter, with no LLM and nothing leaving your machine; only the semantic pass over documents, PDFs, images and video calls a backend model. Code, documents and media all end up in the same graph, so you can query them together.

Who it's for:Developers who need to get their bearings in a large codebase and its documentation quickly. It is of no use if you don't write code.

Before you start: Install the CLI with uv, pipx or pip, then run graphify install to register the skill with your coding agent. The README also offers piping a curl-downloaded script into sh; we do not recommend that. Documents, PDFs, images and video are processed by your agent's model or a service you configure such as Gemini; having no API key set does not mean your material stays local. When a module is missing, the skill installs and upgrades it automatically with uv, and when it falls back to pip it may use --break-system-packages, changing your system Python packages; we do not recommend that kind of automatic system change. It can also install Git hooks and an always-on mode. Graphify Cloud is a separate hosted service with a 14-day trial; check the official site for pricing after that. This project may control your computer, phone or a large number of account permissions; we have not tested it and do not recommend it. The code is open source under the Apache-2.0 license.

Based on the project's public documentation. Not tested by AI Inspiration World.