graphify-windows

Run /graphify on a local path or GitHub URL. It detects files, extracts code structure via AST parsing and document semantics via LLM subagents, then clusters the graph and outputs graph.json, an interactive HTML view, and a GRAPH_REPORT.md summary.

Once graphify-out/graph.json exists, run graphify query "your question" for BFS traversal, or add --dfs to trace a specific dependency path. Use graphify path "A" "B" for shortest paths and graphify explain "Node" for single-concept explanations.

No API key is required. Code extraction is fully structural via AST with no LLM. Semantic extraction for docs, papers, and images uses Gemini only if GEMINI_API_KEY or GOOGLE_API_KEY is set; otherwise the host agent itself performs extraction.

Yes. Use --neo4j to generate a Cypher file or --neo4j-push bolt://localhost:7687 to push directly to Neo4j. FalkorDB is supported the same way via --falkordb and --falkordb-push, using MERGE so re-runs do not create duplicates.

Run /graphify <path> --update to re-extract only new or changed files and prune deleted ones. Code-only changes skip LLM extraction entirely, and a --watch mode or git post-commit hook can trigger rebuilds automatically.

It handles code files across many languages, Markdown and text documents, PDF papers, images via vision extraction, and video or audio files which are first transcribed to text with Whisper before being treated as documents.