# tirth8205/code-review-graph

Source: [GitHub](https://github.com/tirth8205/code-review-graph)  
Feed7 permalink: https://feed7.dev/p/code-review-graph-0ag5r2t  
Published: Unknown  
Trust: Needs Review (needs_review)

## Why Included

code-review-graph gives coding agents a local, incrementally updated dependency map over MCP, narrowing review context while exposing affected callers, dependents, and tests.

## Source Summary

code-review-graph parses repositories with Tree-sitter into a persistent local graph, then serves targeted context through MCP. Its six-repository evaluation reports a **~65x median token reduction**, with graph queries typically returning **2,000–3,500 tokens**.

## Practical Implication

Use it when reviews or impact analysis repeatedly make agents rediscover a large codebase. Hooks and watch mode update only changed files and their dependents; a two-file edit in a roughly **3,000-file repository took about 2.5 seconds** on the measured hook path.

## Agent-Ready Context

code-review-graph parses repositories with Tree-sitter into a persistent local graph, then serves targeted context through MCP. Its six-repository evaluation reports a **~65x median token reduction**, with graph queries typically returning **2,000–3,500 tokens**.

Use it when reviews or impact analysis repeatedly make agents rediscover a large codebase. Hooks and watch mode update only changed files and their dependents; a two-file edit in a roughly **3,000-file repository took about 2.5 seconds** on the measured hook path.

The headline comparison uses whole-corpus reading as an upper bound, not a realistic agent workflow. Reported recall is circular because its ground truth comes from the same graph, while search MRR is **0.35** and flow-detection recall is 33%.

## Connected Context

Feed7 judgment across 386 accumulated Signals:

This turns graph-shaped code context from a general retrieval pattern into a persistent, incrementally updated MCP implementation for review and impact analysis. It reinforces bounded evidence retrieval over whole-repository loading, while sharply narrowing the performance claim: the token baseline is unrealistic, graph-derived recall is circular, and weak search and flow results leave practical retrieval quality unresolved.

- [Graphify-Labs/graphify](https://feed7.dev/p/graphify-1e0bs1f) — Both expose repository relationships as queryable agent context; code-review-graph adds measured incremental-update behavior but provides weaker evidence that the retrieved relationships cover real review needs.
- [AI on Your Lakehouse: Context Comes in Shapes, Not Queries — Zach Blumenfeld, Neo4j](https://feed7.dev/p/ai-on-your-lakehouse-context-comes-in-shapes-not-queries-zach-blumenfeld-0r4s9u5) — It implements the proposed graph-shaped retrieval surface specifically for source code, while its mixed evaluation results reinforce the candidate’s requirement to benchmark graphs on local data.
- [RLM: Recursive Language Models for Large Codebases - Shashi, Superagentic AI](https://feed7.dev/p/rlm-recursive-language-models-for-large-codebases-shashi-superagentic-ai-1w21nex) — Both keep large repositories outside the main context and return bounded evidence, but code-review-graph relies on a precomputed persistent graph rather than agent-written inspection at query time.

## Context Map

- Layer: context
- Domains: coding
- Topics: mcp, retrieval, context-engineering

## Uncertainty

- The headline comparison uses whole-corpus reading as an upper bound, not a realistic agent workflow. Reported recall is circular because its ground truth comes from the same graph, while search MRR is **0.35** and flow-detection recall is 33%.

## Agent Instruction

Use this item as source-backed context. Do not invent claims beyond the linked source. If this item conflicts with another source, call out the conflict.