whiteboard README

The project's README from github.com/devdotfast/whiteboard. As one markdown file: README.md.

Review logo

Review

Cursor for code review.

Download for macOS · Docs · Website · Contributing

Review is an open-source desktop app for understanding and reviewing agent-written code. Coding agents turn a branch or pull request into a guided, interactive Review connected to the exact code behind it.

Explore architecture and data flow, inspect diffs, ask questions, and explore coding traces from one interface.

Review Desktop showing a guided code review and interactive sequence diagram

[!NOTE] Review is in beta and currently ships for macOS on Apple silicon.

Quickstart

  1. Download Review and open the app.
  2. Connect Claude Code, Codex, and other coding agents from the welcome screen.
  3. Ask your agent to review your current branch against up-to-date main and open the result in Review.

Review works especially well for large changes where a file-by-file diff does not explain whether the system is right. See the quickstart for the complete first-review flow.

Review authors can add user and repository guidance for generated documents. See Review guidance.

Documentation

How Review works · Coding agents · CLI · Privacy · Troubleshooting

Known limitations

If there are any other features you need, feel free to ask on Discord or open an issue or discussion!

FAQs

Who is the Review app for?
  • If you still want the gains of generating code with AI but also want to understand how your codebase works and maintain quality, Review is for you.
  • If you don't use AI to write a lot of code and are content reading diffs line by line, Review probably isn't for you.
How is this different from Greptile, Bugbot, CodeRabbit, etc.?

These tools review your LLM-generated code with another LLM. Review is a complementary tool. Review doesn't review your code for you, but helps you understand it and spot regressions quickly.

How is this different from Plannotator?

Plannotator is meant as an alternative to using the terminal for reviewing plans and simple code diffs. Review takes some inspiration from Plannotator in its lifecycle but is engineered as an opinionated framework for reviewing AI-generated code diffs.

How are you guys gonna make money on this?

Eventually, we'll charge companies for a hosted Review product that manages review creation alongside features like trajectory storage, review sharing, etc., but everything will remain easily self-hostable.

Development

Review requires Node.js 24 and pnpm 11. See the desktop prerequisites, then run:

pnpm install
pnpm dev

Run pnpm run ci before submitting a pull request. See CONTRIBUTING.md for more.

A note on vendoring Code OSS

With everyone using dedicated agent TUIs and desktop apps, we only use our text editors for reviewing line-by-line diffs now, so we figured why not have a text editor meant for reviewing code. In that case, might as well start off with the most successful open source editor out there as a baseline.

We vendor Code OSS unlike other forks that maintain patches because coding agents have a hard time with patches and there's a lot of stuff from stock VS Code (i.e., ~45% of the codebase is Copilot these days 😬) that we don't need.

We regularly monitor upstream Code OSS and merge in security/feature patches as they come in.

Privacy

Review runs against local checkouts. Anonymous telemetry does not include your code, diffs, Review text, comments, questions, prompts, or model output. Read the privacy overview, inspect the complete telemetry reference, or turn telemetry off at any time.

License

Review is available under the MIT License. The vendored Code - OSS fork retains Microsoft's MIT license and third-party notices; see apps/review-desktop/LICENSE and apps/review-desktop/UPSTREAM.

Influences

whiteboard (devdotfast/whiteboard)

Searching Database…

3.2K ·GitHub
Whiteboard
这是什么？Review 是一款开源桌面应用，能把编程智能体生成的分支或拉取请求变成有引导的交互式审查，让你探索架构、检查 diff，并就生成的代码提问。
这是通用代码编辑器，还是专门的审查工具？它是专门构建的审查应用，不是通用编辑器。它基于一个内嵌（vendored）的 Code OSS（VS Code 的开源基础）分支构建，经过精简，只做审查 AI 生成 diff 这一件事。
怎么让它跑起来？下载 Review 应用并打开，然后在欢迎界面连接 Claude Code、Codex 或其他编程智能体，让智能体对照 main 分支审查你当前的分支。
它支持哪些编程智能体？自己构建需要什么？Review 可以从欢迎界面连接 Claude Code、Codex 和其他编程智能体；从源码构建需要 Node.js 24 和 pnpm 11。
目前还有哪些不好用的地方？Review 仍处于 beta 阶段：它只能梳理单个仓库内的架构变更，不能正确处理堆叠式 PR，在机器之间分享审查目前也很麻烦。
Review 到底适合谁用？它适合那些用 AI 生成代码、但仍想理解代码工作原理并捕捉回归问题的开发者；如果你乐于逐行阅读每一个 diff，它可能并不适合你。
Review 会把我的代码发送到别处吗？Review 针对本地检出运行，只发送匿名遥测数据，其中不包含你的代码、diff、评论、问题、提示词和模型输出。
Review 与 Greptile、Bugbot 或 CodeRabbit 这类工具有什么不同？那些工具是让另一个 LLM 替你审查 AI 生成的代码。Review 是互补的：它不评判你的代码，而是帮你理解背后的 diff、执行轨迹和架构。
Review 与 Plannotator 有什么不同？Plannotator 是审查计划和简单 diff 时替代终端的方案。Review 从 Plannotator 借鉴了生命周期设计，但构建为一个更有主见的框架，专门用于审查 AI 生成的代码变更。
为什么内嵌 Code OSS，而不是在 VS Code 上维护补丁？编程智能体难以应用补丁，而且原版 VS Code 的代码库里有很大一部分是 Review 用不到的 Copilot 相关代码，所以团队直接内嵌 Code OSS，并按需合并上游的安全和功能修复。
团队打算如何让一款免费、可自托管的应用盈利？他们计划最终向企业收取托管版 Review 产品的费用（附带轨迹存储、审查分享等额外功能），同时保持核心应用免费且可自托管。
What are similar projects?plannotator, pi-gui, CanvasTTY, monocode, zuse, bb, codex-host, wicked-studio
Which topics is it in?diff, code-editor, code-review, monorepo, apple-silicon, electron, agents, developers, coding-agent, codex, privacy, claude-code
Creator
devdotfast
Size
99.3 MB
Stars
3.2K
Updated
CommitOct 16Releasev0.3.2Oct 16
Age
1mo
Commits · 90 Days