# Best open-source AI repos for AI code review (2026) Source: https://dedupedai.com/best-ai-repos-for/ai-code-review/ (Deduped, updated 2026-10-29). Markdown twin of the page; JSON: https://dedupedai.com/best-ai-repos-for/ai-code-review/index.json ## What is the best open-source AI code review tool? The best open-source AI repos for AI code review right now are Juror, OpenQodex and code-review-graph. Use Juror to put several models on every pull request through GitHub Actions, OpenQodex to review a change locally before you push with your own Claude Code or Codex login, and code-review-graph to give any AI reviewer only the code a change touches. ## Comparison | Repo | What it does | Best for | Stars | Language | |---|---|---|---|---| | [Juror](https://dedupedai.com/use-cases/juror-ai-juror/) (Juror-AI/juror) | Juror is a GitHub Action and CLI that has several frontier models review each pull request in parallel, merges duplicate findings, and posts one review with a merge score and cost receipt. | Teams who want multi-model AI code review running on their own GitHub Actions with their own provider keys. | n/a | TypeScript | | [OpenQodex](https://dedupedai.com/use-cases/openqodex-openqodex/) (openqodex/openqodex) | OpenQodex is an open source tool that reviews your code change before you push, combining scanners with a separate AI reviewer that uses your own Claude Code or Codex login. | Developers and the AI coding agents they use, who want a local review of unpushed and uncommitted changes before a push or pull request. | 789 | TypeScript | | [code-review-graph](https://dedupedai.com/use-cases/tirth8205-code-review-graph/) (tirth8205/code-review-graph) | code-review-graph builds a local Tree-sitter knowledge graph of your codebase and serves precise review context to AI coding tools over MCP and CLI, so they read only what matters. | Developers using AI coding tools on large repositories who want cheaper, more precise code reviews. | n/a | Python | | [CodeAF](https://dedupedai.com/use-cases/agent-field-codeaf/) (Agent-Field/CodeAF) | CodeAF is an open-source Go coding harness for open models that runs as one local binary, with specialist subharnesses for coding, security audits and pull request reviews. | Software developers and teams who want to run coding, review and audit work with open models from one local window or a headless command. | 549 | Go | | [diff-tour](https://dedupedai.com/use-cases/init-co-jp-diff-tour/) (init-co-jp/diff-tour) | diff-tour is a read-only terminal app that turns a git change into a guided tour explaining what the change is for and what to check. | Developers and team leads who need to understand or review a git change, including changes an AI agent wrote, and who already have one of the supported LLM CLIs signed in. | 197 | Go | | [first-pass](https://dedupedai.com/use-cases/joetawil7-first-pass/) (joetawil7/first-pass) | first-pass is a set of rules, skills and checks for coding agents that makes them review the code around a change, prove their work, and fix bugs as a class. | Developers who use Claude Code, or Cursor and Codex, on repos they work on and want agents to check the surrounding code before calling work done. | 121 | JavaScript | ## When to pick which ### Juror (Juror-AI/juror) Juror is a GitHub Action and CLI that has several frontier models review each pull request in parallel, merges duplicate findings, and posts one review with a merge score and cost receipt. Best for: Teams who want multi-model AI code review running on their own GitHub Actions with their own provider keys. Needs: A GitHub repository with Actions; Provider API keys such as OpenAI and Fireworks; GitHub CLI for setting secrets. Start here: [Add multi-model AI review to every pull request](https://dedupedai.com/use-cases/juror-ai-juror/multi-model-pr-review-in-github-actions/) (Intermediate, 30 minutes) 1. Export your key with export JUROR_OPENAI_API_KEY=… 2. Run npx juror-ai init --set-secrets in the repository 3. Or preview changes first with npx juror-ai init --dry-run 4. Open a pull request and read the posted review All 3 use cases: https://dedupedai.com/use-cases/juror-ai-juror/ | GitHub: https://github.com/Juror-AI/juror ### OpenQodex (openqodex/openqodex) OpenQodex is an open source tool that reviews your code change before you push, combining scanners with a separate AI reviewer that uses your own Claude Code or Codex login. Best for: Developers and the AI coding agents they use, who want a local review of unpushed and uncommitted changes before a push or pull request. Needs: Node 22 or newer, and git; macOS or Linux (no test runs on Windows or WSL); Claude Code or Codex installed and logged in for a full review; A repository you can run commands in. Start here: [Try OpenQodex on a demo repo with planted bugs](https://dedupedai.com/use-cases/openqodex-openqodex/try-openqodex-on-demo-repo/) (Beginner, 15 minutes) 1. Make sure Node 22 or newer and git are installed. 2. Run npx openqodex demo. 3. Read the printed receipt and the findings it lists. 4. Open the report file the demo names to see the full output. All 4 use cases: https://dedupedai.com/use-cases/openqodex-openqodex/ | GitHub: https://github.com/openqodex/openqodex ### code-review-graph (tirth8205/code-review-graph) code-review-graph builds a local Tree-sitter knowledge graph of your codebase and serves precise review context to AI coding tools over MCP and CLI, so they read only what matters. Best for: Developers using AI coding tools on large repositories who want cheaper, more precise code reviews. Needs: Python with pip or pipx; A supported AI coding tool. Start here: [Give your AI reviewer only the code that matters](https://dedupedai.com/use-cases/tirth8205-code-review-graph/cut-tokens-on-ai-code-reviews/) (Beginner, 15 minutes) 1. Install: pip install code-review-graph (or pipx) 2. Run code-review-graph install to detect your AI tools and write an MCP server entry, hooks and rules for each 3. Run code-review-graph build to parse the codebase (about 40 seconds for a 3,000 file repo) 4. Hooks and watch mode keep the graph updated as files change All 3 use cases: https://dedupedai.com/use-cases/tirth8205-code-review-graph/ | GitHub: https://github.com/tirth8205/code-review-graph ### CodeAF (Agent-Field/CodeAF) CodeAF is an open-source Go coding harness for open models that runs as one local binary, with specialist subharnesses for coding, security audits and pull request reviews. Best for: Software developers and teams who want to run coding, review and audit work with open models from one local window or a headless command. Needs: macOS or Linux machine with a terminal; A provider key (OpenRouter, DeepSeek, GLM, Kimi, MiniMax, Qwen), a ChatGPT plan through Codex, or a running Ollama install; A git repository for the project you want to work on; Install CodeAF with the curl script, or build it from source with make build. Start here: [Audit local code changes for security problems](https://dedupedai.com/use-cases/agent-field-codeaf/security-audit-local-changes/) (Intermediate, 30 minutes) 1. Run curl -fsSL https://agentfield.ai/get/codeaf | bash and then codeaf. 2. Connect a provider on first start, in the browser through OpenRouter or with a key. 3. Open the project's folder in CodeAF and type /sec changes since main. 4. Open the sec task to follow the work, then read the findings and the coverage notes in the report. All 5 use cases: https://dedupedai.com/use-cases/agent-field-codeaf/ | GitHub: https://github.com/Agent-Field/CodeAF ### diff-tour (init-co-jp/diff-tour) diff-tour is a read-only terminal app that turns a git change into a guided tour explaining what the change is for and what to check. Best for: Developers and team leads who need to understand or review a git change, including changes an AI agent wrote, and who already have one of the supported LLM CLIs signed in. Needs: git installed; One of Claude Code, Codex CLI, Gemini CLI, OpenCode or Copilot CLI, installed and signed in; A git repository with a change to tour. Start here: [Learn why a branch change is written that way](https://dedupedai.com/use-cases/init-co-jp-diff-tour/learn-why-a-branch-change-is-written-that-way/) (Intermediate, 15 minutes) 1. Run curl -fsSL https://raw.githubusercontent.com/init-co-jp/diff-tour/main/install.sh | sh 2. Run cd your-repo 3. Run tour, press Enter twice in Setup the first time, then pick the current branch and press Enter 4. Wait 2 to 5 minutes while the tour is written All 5 use cases: https://dedupedai.com/use-cases/init-co-jp-diff-tour/ | GitHub: https://github.com/init-co-jp/diff-tour ### first-pass (joetawil7/first-pass) first-pass is a set of rules, skills and checks for coding agents that makes them review the code around a change, prove their work, and fix bugs as a class. Best for: Developers who use Claude Code, or Cursor and Codex, on repos they work on and want agents to check the surrounding code before calling work done. Needs: Claude Code to install the plugin (Cursor and Codex get the rules and skills, but hooks are Claude Code only); A folder that holds your repos, where you start your sessions; Git repos with a branch to review, for the review use case (GitHub access is needed to read pull requests, inferred from the README). Start here: [Check your own branch before asking for review](https://dedupedai.com/use-cases/joetawil7-first-pass/review-own-branch-before-pr/) (Intermediate, 30 minutes) 1. Run /plugin marketplace add joetawil7/first-pass in Claude Code. 2. Run /plugin install first-pass@first-pass. 3. In the folder where you start your sessions, run /first-pass:setup-first-pass. 4. Check out your branch in the repo, then run /first-pass:review with nothing after it. All 4 use cases: https://dedupedai.com/use-cases/joetawil7-first-pass/ | GitHub: https://github.com/joetawil7/first-pass ## Questions ### What is the best open-source AI code review tool? The best open-source AI repos for AI code review right now are Juror, OpenQodex and code-review-graph. Use Juror to put several models on every pull request through GitHub Actions, OpenQodex to review a change locally before you push with your own Claude Code or Codex login, and code-review-graph to give any AI reviewer only the code a change touches. ### Which repo for AI code review is easiest to start with? OpenQodex: its starter use case "Try OpenQodex on a demo repo with planted bugs" is rated Beginner and takes about 15 minutes. ### How do I get started with Juror? Export your key with export JUROR_OPENAI_API_KEY=…; Run npx juror-ai init --set-secrets in the repository; Or preview changes first with npx juror-ai init --dry-run. Full steps and a copy-paste agent prompt: https://dedupedai.com/use-cases/juror-ai-juror/multi-model-pr-review-in-github-actions/ ### Are these AI code review repos free? They are open-source projects on GitHub (check each license; some, like fair-code projects, limit commercial use). Some also sell a hosted version or need a paid model API key; the requirements for each repo are listed on this page. ### How is this list of AI code review repos chosen? Deduped tracks the hottest AI repos on GitHub and writes step-by-step use cases from each README. For this page we picked the repos that do this job, ordered by fit and then GitHub stars. Repos with a risk flag or a retired page are left out. The page is rebuilt with the site; stars are as of the last build in 2026. Third-party repos we don't own or vouch for. AI-written use cases can be wrong: review everything and run at your own risk.