Graphify turns a folder of code or documentation into a queryable knowledge graph using deterministic tree-sitter AST parsing across approximately 40 languages without requiring LLM calls. It generates graph files for local analysis and exposes the data through a Model Context Protocol (MCP) server or browser-based visualizations.
Project overview
It builds a deterministic code knowledge graph locally using AST parsing, distinguishing direct extractions from inferred connections without requiring LLM API calls.
Project type
MCP · AI Coding · RAG
Use cases
Knowledge Q&A · Coding & Development
Deployment
Refer to project documentation
License
Apache-2.0
Best for
Developers and AI engineers seeking a deterministic AST-based knowledge graph of local codebases without relying on LLM calls for structural mapping.
Key capabilities
Parses code locally using tree-sitter AST across ~40 languages to deterministically build a knowledge graph without LLM calls.
Provides query, path, and explain commands to ask questions, trace connections between two things, or explain concepts against the generated graph.json.
Exposes the graph as an MCP stdio or HTTP server with structured access to query, node, neighbors, and path tools.
Generates an interactive graph.html file to explore nodes, filter communities, and visualize project structure in any browser.
Installs and registers the graphify skill into 20+ AI coding assistants (Claude Code, Cursor, etc.) via hooks or instruction files.
Extracts concepts from docs, PDFs, images, and videos by routing to configured LLM backends like Gemini, Claude, OpenAI, or Ollama.
Limitations and risks
Fully local and free operation is limited to code files; semantic extraction over docs or media requires a configured LLM backend.
Semantic extraction of documents, PDFs, and images sends data to a configured external AI provider.
Getting started
Install the package using a uv tool or pipx. The setup requires two commands: install the package and register the skill. Run /graphify . in the AI assistant to start the process.
Evidence and sources
GitHub project description: Turn any folder of code, SQL schemas, R scripts, shell scripts, docs, papers, images, or videos into a queryable knowledge graph. App code + database schema + infrastructure in on…
README: uv tool install graphifyy # install the CLI (or: pipx install graphifyy) graphify install # register the skill with your AI assistant
README: graphify-out/ ├── graph.html open in any browser — click nodes, filter, search ├── GRAPH_REPORT.md the highlights: key concepts, surprising connections, suggested questions └── gr…
README: Type `/graphify` in your AI coding assistant and it maps your entire project (code, docs, PDFs, images, videos) into a **knowledge graph** you can **query instead of grepping** th…
README: Not a vector index. No embeddings, no vector store: a real graph you traverse. Ask a question, trace the path between two things, or explain one concept.