Overview

code-review-graph revolutionizes AI coding workflows by building a persistent, local-first intelligence graph of your codebase. It empowers AI assistants to understand your project's structure and dependencies deeply, enabling them to read only the most relevant code for tasks like reviews and large-repo operations. This dramatically reduces token consumption and improves the efficiency and accuracy of AI-driven code analysis.

Key Features

• Intelligent Context Reduction: Utilizes a structural code map built with Tree-sitter and Model Context Protocol (MCP) to provide AI with precisely scoped context, drastically reducing the number of tokens read.
• Local-First Operation: The entire knowledge graph is built and queried locally, ensuring code privacy and seamless integration into existing development environments.
• Incremental Updates: Re-parses only changed files, with updates completing in under 2 seconds, keeping your AI's understanding of the codebase current.
• Broad Language and Notebook Support: Parses a wide array of languages including Python, JavaScript/TypeScript/TSX, Go, Rust, Java, C/C++, C#, Ruby, Kotlin, Swift, PHP, Scala, Solidity, Dart, R, Perl, Lua/Luau, Objective-C, shell scripts, Elixir, Zig, PowerShell, Julia, ReScript, GDScript, Nix, Verilog/SystemVerilog, SQL, Vue/Svelte SFCs, Astro files, and Jupyter/Databricks notebooks.
• Blast-Radius Analysis: Identifies all potentially affected functions, classes, and files when a change is made, providing a clear scope of impact.
• Risk-Scored CI Reviews: Integrates with GitHub Actions to provide risk-scored PR review comments and optional merge gating based on code changes.
• Extensible Language Support: Easily add support for new languages by defining grammars and node types in a languages.toml file.
• Performance Benchmarks: Demonstrates significant token efficiency gains (up to 82x median reduction) and accurate impact analysis.

Typical Use Cases

• Smarter Code Reviews: AI assistants can perform reviews with a fraction of the token cost, focusing only on code directly impacted by changes.
• Efficient Large Repository Navigation: Navigate and understand massive monorepos by providing AI with targeted context, overcoming the limitations of full-corpus scans.
• Enhanced AI-Powered Refactoring: Tools can leverage the graph for more accurate dead code detection, renaming previews, and community-driven suggestions.
• Automated CI/CD Code Quality Checks: Implement intelligent, context-aware code quality gates directly within your CI/CD pipeline.
• Developer Onboarding: Quickly bring new developers up to speed by using AI to explain complex code structures and dependencies.
• Deep Codebase Understanding: Gain insights into architectural patterns, identify structural weaknesses, and discover unexpected coupling through advanced graph analysis.