Code Review Graph

Code Review Graph helps developers review changes with AI coding assistants without repeatedly supplying an entire repository. It keeps a local map of code relationships and returns the context relevant to a change or question.

• Change impact: Trace affected callers, dependencies and tests so an assistant can inspect the files connected to a diff.
• Review risks: Identify risky functions, affected execution flows, test gaps and unexpected connections between parts of a project.
• Assistant integration: Supply context to Claude Code, Cursor, Codex, Copilot and other MCP clients, with CLI and VS Code access too.
• Visual exploration: Explore code relationships in an interactive graph and export them to tools such as Gephi, Neo4j or Obsidian.
• Pull request analysis: A GitHub Action runs on your CI runner and posts risk findings without sending source code to an external analysis service.

The open-source tool uses the MIT license. Its graph and core analysis run locally, with incremental updates as files change. Language support includes Python, TypeScript, Rust, Swift and Java, plus Jupyter and Databricks notebooks. It also follows supported Laravel and Spring relationships.

Optional semantic search can use local sentence-transformers embeddings or providers such as Gemini and Voyage AI. OpenAI-compatible endpoints include vLLM and LocalAI. Cloud embeddings send data to the selected provider; the connected coding assistant determines where supplied review context is processed.