🕸️code-review-graph
tirth8205/code-review-graph · homepage ↗
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.
⭐ Very popular: 30k stars, gaining about 986 a week
View on GitHub ↗repo profile
momentum
durability
bus factor = how many people it takes to cover more than half the commits (6 months). 1 is a solo project; higher means the work is spread across a team. top-author share is the single busiest author's slice of those commits.
since we covered it
why it's a big deal
- It turns code review into a graph query: 8.2x average token reduction across six real repos in its own benchmarks, up to 49x on a Next.js-scale monorepo, with 100% recall on actually impacted files.
- It is agent-agnostic to an almost aggressive degree - a single install command auto-configures 14 platforms, from Codex and Claude Code to Cursor, Windsurf, Zed and Copilot. The graph is shared infrastructure, not a plugin for one vendor.
- It benchmarks honestly: the README publishes its own sub-1x failure case on small single-file changes and names its weak spots (search ranking, flow detection) right next to the wins. That culture deserves stars on its own.
under the hood
- Tree-sitter grammars for 24 languages plus Jupyter and Databricks notebooks; nodes and edges live in a local SQLite file, no external database, no cloud dependency.
- Incremental updates via git hooks and SHA-256 diffing re-index a 2,900-file project in under two seconds; a multi-repo daemon keeps graphs fresh for editors without hook support.
- 28 MCP tools (impact radius, budgeted graph traversal, hub and bridge detection, dead-code refactor previews) plus five workflow prompts, D3 visualization and GraphML/Neo4j/Obsidian exports. MIT-licensed.
our take from PR#39, 2026-07-30
star history
- PR#39 28k 2026-07-30
- now 30k + 3k since first covered
curve is sampled from GitHub's star history, plus our own daily readings since we covered it; the dashed stretch is before we first covered it, the solid line since. figures at coverage are the numbers we printed then (approx.), current count is live.
understory
Output and attention are roughly in balance.
- output, commits & releases
- clout, star velocity
output = commits & releases; clout = star velocity, both 0 to 100 monthly indices; the gap where output runs above clout is the understory. The understory →
covered in
-
blast-radius analysis for every AI code review
similar projects
compare these →- 🗜️ headroom
2.2× the stars
Compress tool outputs, logs, files, and RAG chunks before they reach the LLM. 20% fewer tokens for coding agents, 60-95% fewer tokens for JSON, same answers. Library, proxy, MCP server.
66k ACTIVE - 🤖 Serena
A powerful MCP toolkit for coding, providing semantic retrieval and editing capabilities - the IDE for your agent
28k ACTIVE - ⚡ RTK
Rust · 2.5× the stars
CLI proxy that reduces LLM token consumption by 60-90% on common dev commands. Single Rust binary, zero dependencies
76k ACTIVE
comments
Sign in with GitHub to add your blip on code-review-graph.