Claude Code Adds iOS Simulator Testing

Anthropic's documentation says the desktop pane drives Apple's simulator directly. It can tap, type, launch an app, and take screenshots after the developer grants access to the simulated device. Because the pane does not rely on Claude's broader computer-use feature, it does not need the macOS Accessibility and Screen Recording permissions that screen control requires.

This is distinct from Claude Code's command-line computer-use workflow. The dedicated desktop pane keeps the simulator visible alongside the conversation and does not take over the rest of the screen. MacRumors independently reported the release and the same direct-control design.

The beta requires Claude Desktop 1.24012.0 or later, a Mac, and Xcode with the iOS platform installed. Anthropic lists the pane for Pro, Max, and Team plans, but not Enterprise. It works only in local Claude Code Desktop sessions; cloud and SSH sessions cannot reach simulator devices on the developer's Mac.

The integration controls simulated iPhones and iPads, not physical hardware. That makes it useful for build-and-run checks and interactive UI flows, but it does not replace validation on real devices for hardware-specific behavior, performance, sensors, push notifications, or other platform integrations.

Anthropic says screenshots taken from the simulated device are sent to Anthropic and kept under the user's normal conversation retention settings. Its documentation advises developers not to sign in to real accounts on a simulator Claude controls. MacRumors also highlighted that limitation.

For development teams, that makes synthetic fixtures, mock credentials, and isolated test backends important defaults. A simulator can still display tokens, customer-like records, or internal screens even though it is not a physical device. The feature shortens the feedback loop between code generation and visible runtime checking, but teams still need ordinary controls around test data, secrets, and final release validation.