Claude Code brings live iOS app testing into its Mac app

**Claude Code**’s Mac app now supports live iOS app testing through an interactive simulator pane. Users can exercise iOS apps from inside Claude Code itself, instead of switching out for a separate simulator workflow. The feature depends on having **Xcode** installed with the **iOS platform** available on the machine.

Under the hood, the Mac app surfaces the iOS simulator as a pane that stays in the same Claude Code workspace. That keeps the agent, code, and running app in one surface so you can prompt, change code, and observe simulator behavior without leaving the tool. The hard prerequisite is local Xcode and the iOS platform; without those, the simulator path is unavailable.

For engineers building iOS apps with an AI coding agent, the gain is tighter feedback. You can ask Claude Code to change UI, navigation, or app logic and check the result in the live simulator while the session continues. That shortens the loop between “edit” and “does it actually run on iOS,” which is usually where agent-assisted mobile work stalls.

In market terms, this pushes Claude Code further into full-stack Mac development for Apple platforms, not just code generation. Teams already on Xcode get a more integrated agent surface; those without the full iOS toolchain still cannot use this path. The practical bar remains the same as native iOS work: Mac, Xcode, and the iOS platform installed.

The takeaway is straightforward. If you ship iOS apps and already run Claude Code on a Mac with Xcode and the iOS platform, try routing a small UI or flow change through the simulator pane and see how much of the edit–run–inspect cycle stays inside the app. Watch whether Apple-target workflows (build failures, simulator state, multi-screen flows) stay reliable in that pane over longer sessions—that will decide how much of day-to-day iOS iteration can stay inside Claude Code.