BigHugger
GH Repository · tirth8205

code-review-graph

Local-first code intelligence graph for MCP and CLI. Builds a persistent map of your codebase so AI coding tools read only what matters, with benchmarked context reductions on reviews and large-repo workflows.

stars
31,449
30-day movement
+10334/day
Related entries
67
Connections
5
python/uvmcp-servermcpclaude-codeai-codingclaudeknowledge-graphcode-reviewtree-sittergraphragincrementalllmpythonstatic-analysisPython

A local-first tool that builds a persistent knowledge graph of your codebase using static analysis (tree-sitter) and exposes it via MCP and a CLI. It aims to reduce the context AI coding tools need by letting them query only relevant parts of the repository.

When your codebase is too large for AI tools to read whole, this gives them a structured map to pull only what matters.

Use it to

  • Serve codebase context to Claude via MCP
  • Run code reviews with reduced context usage
  • Incrementally index a large repository locally
  • Query code structure from the CLI

For Developers using AI coding assistants on large repositories

Role
mcp-server
Language
Python
Licence
MIT
Forks
2,855
Open issues
96
Last push
2026-09-13
Latest release
v1.1.0 · 2026-02-26
Skills shipped
7
topicsmcpcode-reviewknowledge-graphstatic-analysistree-sitterllm