RELATED 01
code-review-graph
Local code graph that gives a coding assistant review context, blast radius and change risk.
33kGitHub stars
302k/moPyPI downloads
103popularity score
graphify
Lets an assistant query a knowledge graph of a codebase and its docs instead of grepping through files
By Graphify Labs Python Developer toolsKnowledge and memory
72 popularity score ★ 144k
What it does
graphify, by Graphify Labs, maps a project (code, docs, SQL schemas, configs, PDFs, images and video) into a knowledge graph that a coding assistant can query. Code is parsed locally with tree-sitter, with no model involved; docs, PDFs and images get a semantic pass through your assistant's model or a configured API key. Every edge is tagged as extracted from the source or inferred, and there is no vector store: it is a graph you traverse.
The main way to use it is the /graphify skill, which writes graphify-out/graph.json, an interactive graph.html and a report. The optional MCP server then serves that graph.json to an assistant with tools to query the graph, look up a node and its neighbours, find the shortest path between two concepts, and list and triage pull requests by their impact on the graph.
The MCP server ships in the PyPI package graphifyy with the mcp extra. Its README launches it with uvx --from "graphifyy[mcp]" graphify-mcp (no API key), or with python -m graphify.serve graphify-out/graph.json. It runs over stdio by default; --transport http serves the same tools over Streamable HTTP so a team can share one process, with an optional API key.
Tools 7
query_graph- Query the knowledge graph with a question.
get_node- Get one node of the graph.
get_neighbors- Get the nodes connected to a node.
shortest_path- Find the shortest path between two nodes.
list_prs- List pull requests.
get_pr_impact- Show which parts of the graph a pull request affects.
triage_prs- Triage pull requests.