code-review-graph 1.8.2

Persistent incremental knowledge graph for token-efficient, context-aware code reviews with Claude Code

code-review-graph

Stop burning tokens. Start reviewing smarter.

Claude Code re-reads your entire codebase on every task. code-review-graph fixes that. It builds a structural map of your code with Tree-sitter, tracks changes incrementally, and gives Claude precise context so it reads only what matters.

---

Quick Start

Claude Code Plugin (recommended)

claude plugin marketplace add tirth8205/code-review-graph
claude plugin install code-review-graph@code-review-graph

pip

pip install code-review-graph
code-review-graph install

Restart Claude Code after either method. Requires Python 3.10+ and uv.

Then open your project and tell Claude:

Build the code review graph for this project

The initial build takes ~10 seconds for a 500-file project. After that, the graph updates automatically on every file edit and git commit.

---

How It Works

Your repository is parsed into an AST with Tree-sitter, stored as a graph of nodes (functions, classes, imports) and edges (calls, inheritance, test coverage), then queried at review time to compute the minimal set of files Claude needs to read.

Blast-radius analysis

When a file changes, the graph traces every caller, dependent, and test that could be affected. This is the "blast radius" of the change. Claude reads only these files instead of scanning the whole project.

Incremental updates in < 2 seconds

On every git commit or file save, a hook fires. The graph diffs changed files, finds their dependents via SHA-256 hash checks, and re-parses only what changed. A 2,900-file project re-indexes in under 2 seconds.

12 supported languages

Python, TypeScript, JavaScript, Go, Rust, Java, C#, Ruby, Kotlin, Swift, PHP, C/C++

Each language has full Tree-sitter grammar support for functions, classes, imports, call sites, inheritance, and test detection.

---

Benchmarks

All figures come from real tests on three production open-source repositories.

Code review benchmark details (6.8x average reduction)

Tested across 6 real git commits. The graph replaces reading entire source files with a compact structural summary (156-207 tokens) covering blast radius, test coverage gaps, and dependency chains.

Repo	Size	Standard Approach	With Graph	Reduction	Review Quality
httpx	125 files	12,507 tokens	458 tokens	26.2x	9.0 vs 7.0
FastAPI	2,915 files	5,495 tokens	871 tokens	8.1x	8.5 vs 7.5
Next.js	27,732 files	21,614 tokens	4,457 tokens	6.0x	9.0 vs 7.0
Average		13,205	1,928	6.8x	8.8 vs 7.2

Standard approach: reading all changed files plus the diff. Quality scored on accuracy, completeness, bug-catching potential, and actionable insight (1-10 scale).

Live coding task details (14.1x average, 49x peak)

An agent performed 6 real coding tasks (adding features, fixing bugs) across the same repositories. The graph directed it to the right files and away from everything else.

Task	Repo	With Graph	Without Graph	Reduction	Files Skipped
Add rate limiter	httpx	14,090	64,666	4.6x	58
Fix streaming bug	httpx	14,090	64,666	4.6x	59
Add rate limiter	FastAPI	37,217	138,585	3.7x	1,120
Fix streaming bug	FastAPI	36,986	138,585	3.7x	1,121
Add rate limiter	Next.js	15,049	739,352	49.1x	~16,000
Fix streaming bug	Next.js	16,135	739,352	45.8x	~16,000

The graph identified the correct files in every case. Savings scale with repository size.

Monorepo scale: the 49x case

Large repositories benefit most. In the Next.js monorepo (27,732 files, 739K tokens), the graph narrows the review context to ~15 files and 15K tokens, a 49x reduction with 27,700+ files excluded entirely.

---

Usage

Slash commands

Command	Description
/code-review-graph:build-graph	Build or rebuild the code graph
/code-review-graph:review-delta	Review changes since last commit
/code-review-graph:review-pr	Full PR review with blast-radius analysis

CLI reference

code-review-graph install     # Register MCP server with Claude Code
code-review-graph build       # Parse entire codebase
code-review-graph update      # Incremental update (changed files only)
code-review-graph status      # Graph statistics
code-review-graph watch       # Auto-update on file changes
code-review-graph visualize   # Generate interactive HTML graph
code-review-graph serve       # Start MCP server

MCP tools

Claude uses these automatically once the graph is built.

Tool	Description
build_or_update_graph_tool	Build or incrementally update the graph
get_impact_radius_tool	Blast radius of changed files
get_review_context_tool	Token-optimised review context with structural summary
query_graph_tool	Callers, callees, tests, imports, inheritance queries
semantic_search_nodes_tool	Search code entities by name or meaning
embed_graph_tool	Compute vector embeddings for semantic search
list_graph_stats_tool	Graph size and health
get_docs_section_tool	Retrieve documentation sections

---

Features

Feature	Details
Incremental updates	Re-parses only changed files. Subsequent updates complete in under 2 seconds.
12 languages	Python, TypeScript, JavaScript, Go, Rust, Java, C#, Ruby, Kotlin, Swift, PHP, C/C++
Blast-radius analysis	Shows exactly which functions, classes, and files are affected by any change
Auto-update hooks	Graph updates on every file edit and git commit without manual intervention
Semantic search	Optional vector embeddings via sentence-transformers
Interactive visualisation	D3.js force-directed graph with edge-type toggles and search
Local storage	SQLite file in .code-review-graph/. No external database, no cloud dependency.
Watch mode	Continuous graph updates as you work

Configuration

To exclude paths from indexing, create a .code-review-graphignore file in your repository root:

generated/**
*.generated.ts
vendor/**
node_modules/**

For semantic search, install the optional embeddings dependencies:

pip install code-review-graph[embeddings]

---

Contributing

git clone https://github.com/tirth8205/code-review-graph.git
cd code-review-graph
python3 -m venv .venv && source .venv/bin/activate
pip install -e ".[dev]"
pytest

Adding a new language

Edit code_review_graph/parser.py and add your extension to EXTENSION_TO_LANGUAGE along with node type mappings in _CLASS_TYPES, _FUNCTION_TYPES, _IMPORT_TYPES, and _CALL_TYPES. Include a test fixture and open a PR.

Licence

MIT. See LICENSE.

pip install code-review-graph && code-review-graph install