Instructions for your coding agent

Your agent checks its own work when its instructions tell it to. Put this in CLAUDE.md, AGENTS.md or .cursor/rules, with your scheme and app names:

CLAUDE.md

## Verifying UI changes
After you change a screen in this iOS app, verify it before you say it works.
1. Build for the simulator:
   xcodebuild -scheme <Scheme> -destination 'generic/platform=iOS Simulator' -derivedDataPath .mobster/build CODE_SIGNING_ALLOWED=NO build
2. Call mobster's verify_start with app_path set to the absolute path of
   .mobster/build/Build/Products/Debug-iphonesimulator/<App>.app, the steps in plain English, and expect:
   what must be true when the steps are done.
3. Drive with screen, tap, type_text and swipe, then call verify_finish.
4. If the verdict is failed, fix the code and verify again. Report the verdict and the report path.

Make your controls findable

Give the controls your checks need an accessibility identifier, such as .accessibilityIdentifier("streak_days") in SwiftUI. An id selector then survives copy changes and translations, where a label wouldn’t.

Run the whole suite

To run every saved check instead of one flow, the agent calls run_tests, or runs mobster test in a shell:

AGENTS.md

Before you open a pull request, run `mobster test`. If a check fails, read its line and the report it names, fix the code and run it again.

Run every check explains the suite, and MCP server the tool.

Set up

Use

Remote

Instructions for your coding agent

Make your controls findable

Run the whole suite

​Make your controls findable

​Run the whole suite

Make your controls findable

Run the whole suite