Expo 推出云端模拟器，让 AI 编码智能体能自己跑 App 测试
做 React Native 的人可以看看：Expo 让 Claude Code 这类智能体写完代码后能自己装进 iOS 模拟器跑一遍，还给你录屏和在线链接。写码工具很多，能自己测 App 的还少见。
Expo 正在推出 Expo Simulators 云端模拟器，解决 AI 编码智能体只能写代码、却无法运行和验证 App 的问题。开发者执行 eas simulator --platform ios 即可启动一个远程会话，智能体可以把构建产物装进去，并通过 agent-device、Argent 或 Appium 驱动应用进行自动化操作测试。
每个会话都会录制屏幕，智能体最终会返回截图、视频，以及一个评审者可以直接点进去操作的在线浏览器链接。这意味着代码审查者不再只看 diff，还能实际体验智能体改完后的运行效果。
对 iOS/React Native 开发者来说，这补齐了智能体开发闭环中「验证」这一环，类似 Cursor、Claude Code 等工具解决的是写码环节，Expo 解决的是跑码环节。
AI coding agents could write app code but never see the app run. Expo is rolling out cloud simulators that allow agents to test their own changes first.
In Expo Simulators, eas simulator --platform ios starts a remote session where the agent installs its build and drives it with agent-device, Argent or Appium.
Every session records the screen, so the agent returns screenshots, video and a live browser link reviewers can tap through. https://t.co/sth8KC1Zvm
This report is categorized under AITOP's "Products" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows. Key entities and technical topics under continuous tracking include Expo, Claude Code, iOS模拟器, 编程助手.
Source attribution: curated from rohanpaul_ai (@rohanpaul_ai). Cross-source corroboration: 11 independent sources have covered this development (including TestingCatalog, Zed, Varun Mohan), confirming industry-wide signal strength.
