device-mobile-automation

Operate iOS and Android devices via accessibility-tree-driven UI automation.

3|Updated Mar 14, 2026
One-click install
npx skills add https://github.com/Muvon/octomind-tap --skill device-mobile-automation
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: device-mobile-automation
Source: https://github.com/Muvon/octomind-tap/tree/main/skills/device-mobile-automation
Command: npx skills add https://github.com/Muvon/octomind-tap --skill device-mobile-automation

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It enables reliable UI automation on real and simulated mobile devices, eliminating slow manual tapping to complete device-driven workflows.

Core Features & Use Cases

  • Mobile device driving: Automates discovery and selection of iOS/Android devices and simulators/emulators, then performs app install/launch/terminate and UI interactions (tap, swipe, typing, and device button presses).
  • Accessibility-tree targeting: Finds UI elements using accessibility-id first, then label/name/role, with coordinate fallback after fresh snapshots.
  • Safety boundaries for mobile risks: Enforces stop-and-ask handling for OS/app-store/biometric/security-critical overlays and disallows jailbreak/root/DRM/biometric bypass and protected payment flows.
  • Use case: You want to run the same end-to-end UI flow on an iPhone simulator and an Android emulator to regression-test onboarding steps and capture consistent run artifacts.

Quick Start

Ask to automate your phone or tablet’s UI flow, for example: "Run the device-mobile-automation skill to launch the app, navigate to the target screen, and complete the steps while stopping for any permission or biometric prompts."

Frequently Asked Questions about device-mobile-automation

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automate UI testing on iOS and Android devices using natural language?

Mobile UI automation translates natural-language intent into actionable device and app operations, handling app lifecycle control, tapping, swiping, and typing across iOS and Android platforms. It targets UI elements via the accessibility tree to execute end-to-end flows.

How does accessibility tree targeting work for mobile automation?

Accessibility tree targeting finds UI elements using accessibility-id first, then label, name, or role. It falls back to coordinate-based interactions only after capturing fresh snapshots, ensuring accurate element identification during automated mobile testing.

Can I run the same end-to-end UI flow on both an iPhone simulator and an Android emulator?

You can automate device discovery and selection across iOS simulators and Android emulators to run identical end-to-end UI flows. This allows regression testing of onboarding steps and captures consistent run artifacts across both platforms.

What safety boundaries apply when automating mobile app flows with biometric or permission prompts?

Mobile automation enforces stop-and-ask handling for OS, app-store, biometric, and security-critical overlays. It disallows jailbreak, DRM/attestation bypass, biometric bypass, credential persistence, and protected purchase flows to maintain mobile-specific safety boundaries.

Do I need a specific mobile MCP capability to drive iOS and Android apps?

Driving iOS and Android apps requires the device:operator agent and the device-mobile capability, known as mobile-mcp. These components enforce safety boundaries and translate intent into device orchestration for UI automation.