LaunchKit · 2026
Back to Skills

atl-mobile

Mobile browser and native app automation via ATL (iOS Simulator).

0
1.5k downloads
by @jordancoin

Setup & Installation

openclaw skills install @jordancoin/atl-mobile

Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:

npx clawhub install atl-mobile

What This Skill Does

ATL (Agent Touch Layer) is an HTTP-based automation layer for iOS Simulator. It runs two local servers: port 9222 for mobile Safari and port 9223 for native iOS app automation via XCTest. Agents interact with web and native UI through HTTP commands, using a numbered mark system to locate elements without vision API calls.

The mark system assigns numbered labels to interactive elements with exact pixel coordinates, so agents can tap precisely without paying for vision model API calls on every screenshot.

When to use it

  • Automating checkout flows on mobile e-commerce sites in Safari
  • Testing native iOS app navigation without a physical device
  • Scraping web content as rendered in mobile Safari
  • Capturing screenshots of mobile-specific UI layouts for QA
  • Running scripts that configure iOS system app settings

Example Workflow

Here's how your AI assistant might use this skill in practice.

INPUT

User asks: Open iOS Settings, navigate to Wi-Fi, and take a screenshot

AGENT
  1. 1Sends openApp command to port 9223 with bundleId com.apple.Preferences
  2. 2Calls snapshot with interactiveOnly to retrieve the accessibility tree
  3. 3Uses find command to locate and tap the Wi-Fi cell by label text
  4. 4Sends screenshot command and decodes the base64 PNG response to a file
  5. 5Calls closeApp to terminate Settings
OUTPUT

PNG screenshot of the Wi-Fi settings screen saved to /tmp/wifi-settings.png

Requirements

Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.

Xcode installed on macOS (provides xcrun, xcodebuild, and iOS Simulator)