smithery/krzemienski

claude-mobile-ios-testing

Use when testing iOS apps on simulator, capturing screenshots for validation gates, automating UI testing with expo-mcp and xc-mcp, or verifying visual correctness - combines expo-mcp autonomous testing (React Native level) with xc-mcp simulator management (iOS level)

Installation

$ npx skills add smithery/krzemienski --skill claude-mobile-ios-testing

Similar popular skills

Related neighbors and high-traction skills in the same topics — useful to compare before installing.

Also in this package

Other skills from smithery/krzemienski.

npx skills add smithery/krzemienski

Browse all from smithery/krzemienski

More details

Agent compatibility

Declared targets from SKILL.md / docs. Unmarked agents are not listed — the skill may still install via the CLI.

Claude Code Declared
Cursor Not declared
Codex Not declared
GitHub Copilot Not declared
Windsurf Not declared
Gemini CLI Not declared
Cline Not declared
OpenCode Not declared

Also listed on

Alternate registries and mirrors of this skill.

Skill metadata

Parsed from SKILL.md frontmatter.

Declared agents claude-code

Package contents

Files included with this skill beyond the listing page.

  • skill md SKILL.md 5,741 B
  • docs SUMMARY.md 301 B

History

  1. First recorded snapshot · 0 installs

SKILL.md

iOS Testing Automation - expo-mcp + xc-mcp Hybrid

Overview

Autonomous iOS testing combining expo-mcp (React Native component testing) with xc-mcp (simulator management).

Core principle: expo-mcp for UI testing (testID-based). xc-mcp for simulator lifecycle. AI validates visually.

Announce at start: "I'm using the claude-mobile-ios-testing skill for autonomous iOS testing."

When to Use

  • Testing iOS app with visual verification (Gates 4A, 6A-E)
  • Capturing screenshots with AI analysis
  • Automating UI interactions by testID
  • Multi-device testing (iPhone SE, 14, Pro Max)
  • Verifying React Native components work correctly

Tool Selection Matrix

Operation Tool Why
Install packages expo-mcp Correct Expo SDK versioning + usage docs
Search docs expo-mcp Natural language Expo documentation search
Screenshot expo-mcp automationtakescreenshot (React Native level)
Find element expo-mcp automationfindviewbytestid (by testID prop)
Tap element expo-mcp automationtapby_testid (by testID, React Native level)
Boot simulator xc-mcp simctl-boot (iOS simulator management)
Install .app xc-mcp simctl-install (after xcodebuild)
Launch app xc-mcp simctl-launch (iOS app lifecycle)
Low-level tap xc-mcp idb-ui-tap (when testID not available, use coordinates)
Accessibility tree xc-mcp idb-ui-describe (iOS accessibility API)

Hybrid Workflow

1. Simulator Setup (xc-mcp)

// Boot iPhone simulator
mcp__xc-mcp__simctl-boot({
  deviceId: "iPhone 14",
  waitForBoot: true,
  openGui: true
});

2. Install and Launch (xc-mcp)

// Install .app bundle (after xcodebuild or expo run:ios)
mcp__xc-mcp__simctl-install({
  udid: "booted",
  appPath: "/Users/nick/Desktop/claude-mobile-expo/claude-code-mobile/ios/build/Build/Products/Debug-iphonesimulator/claudecodemobile.app"
});

// Launch app
mcp__xc-mcp__simctl-launch({
  udid: "booted",
  bundleId: "com.yourcompany.claudecodemobile"
});

3. Autonomous Testing (expo-mcp)

Requires: Metro running with EXPOUNSTABLEMCP_SERVER=1

Visual Verification:

"Take screenshot of Chat screen empty state"

AI analyzes screenshot:

  • ✅ Purple gradient background (#0f0c29→#302b63→#24243e)?
  • ✅ Input field visible?
  • ✅ Send button visible?
  • ✅ Colors match spec?

Element Verification:

"Find view with testID 'message-input' and verify it's accessible"

expo-mcp returns: {found: true, accessible: true, enabled: true}

Interaction Testing:

"Tap button with testID 'send-button'"

expo-mcp: Taps element successfully

Outcome Verification:

"Take screenshot and verify message appeared in chat"

AI analyzes: Message bubble visible ✅, correct styling ✅

4. Complete Test Cycle

// Test workflow for Chat screen
const tests = [
  "Take screenshot of Chat screen empty state, verify gradient background",
  "Find view with testID 'message-input'",
  "Tap view with testID 'message-input'",
  "Take screenshot and verify keyboard appeared",
  "Find view with testID 'settings-button'",
  "Tap button with testID 'settings-button'",
  "Take screenshot and verify Settings screen loaded",
];

for (const test of tests) {
  // AI executes test via expo-mcp prompt
  // AI analyzes result
  // AI determines pass/fail
}

testID Requirements

ALL interactive elements MUST have testID:

// ✅ CORRECT
<TouchableOpacity testID="send-button" onPress={handleSend}>
<TextInput testID="message-input" />
<Pressable testID="settings-button" onPress={openSettings}>
<FlatList testID="message-list" />

Why: expo-mcp automationtapbytestid and automationfindviewby_testid require testID props

Multi-Device Testing

const devices = ["iPhone SE (3rd generation)", "iPhone 14", "iPhone 14 Pro Max"];

for (const device of devices) {
  // 1. Boot (xc-mcp)
  mcp__xc-mcp__simctl-boot({deviceId: device});
  
  // 2. Install and launch (xc-mcp)
  // ... (same as above)
  
  // 3. Test with expo-mcp
  "Take screenshot of Chat screen on ${device} and verify layout correct"
  
  // 4. AI verifies responsive layout
  
  // 5. Shutdown
  mcp__xc-mcp__simctl-shutdown({deviceId: "booted"});
}

Common Mistakes

Mistake Reality
"Use npm install" WRONG. Use expo-mcp "Add package-name".
"Use xc-mcp screenshot" WRONG for RN testing. Use expo-mcp automationtakescreenshot.
"Tap by coordinates always" WRONG. Use expo-mcp automationtapby_testid with testID.
"Skip testID props" WRONG. Required for expo-mcp automation.
"Manual visual verification" WRONG. AI verifies screenshots autonomously.

Red Flags

  • "npm install is faster" → WRONG. expo-mcp provides correct versioning.
  • "xc-mcp screenshot is fine" → WRONG. expo-mcp works at React Native level.
  • "testIDs are optional" → WRONG. Required for autonomous testing.
  • "I'll verify manually" → WRONG. AI verifies via screenshot analysis.

All mean: Use expo-mcp for RN testing. Add testIDs. Let AI verify.

Integration

  • Use WITH: @claude-mobile-metro-manager (start Metro with MCP flag)
  • Use WITH: @claude-mobile-validation-gate (complete gate execution)
  • Provides to: Gate 4A, Gates 6A-E (visual validation)