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.
code-review-graph is a local-first code intelligence graph that builds a persistent structural map of a codebase and supplies relevant context to AI coding tools through MCP and a CLI. It is used for code review and large-repository workflows, with catalogue add-ons providing integrations and instructions for operating it.
Latest release v2.3.9 · 18 Sept 2026
4 files for GitHub Copilot, Codex, OpenCode, Claude Code and Gemini CLI: code-review-graph copilot-instructions.md, code-review-graph AGENTS.md, code-review-graph CLAUDE.md, code-review-graph GEMINI.md — 5,290 tokens loaded in every session.
.github/copilot-instructions.md A 538 tok AGENTS.md C 1,120 tok CLAUDE.md A 3,522 tok GEMINI.md A 110 tok These files are tirth8205/code-review-graph's own configuration — they tell GitHub Copilot, Codex, OpenCode, Claude Code and Gemini CLI how to work on this repository, so they are not mods to install elsewhere. Copy one as a starting point and replace the parts that are about this project.