This project provides a platform-agnostic MCP server that bridges AI agents with iOS and Android devices for mobile automation. It enables scalable app interaction using accessibility snapshots without requiring platform-specific test automation frameworks.
Project overview
It drives mobile apps from the native accessibility tree instead of relying on vision models, providing one API that works across iOS and Android simulators, emulators, and real devices.
Project type
MCP · AI Agent
Use cases
Automation
Deployment
Refer to project documentation
License
Apache-2.0
Best for
Developers and AI engineers seeking to automate mobile app interactions across iOS and Android using MCP-supported foundational models or agents.
Key capabilities
Drives apps from the native accessibility tree without needing a vision model or image tokens, falling back to screenshots and coordinates only when necessary.
Allows running the MCP server over Server-Sent Events instead of stdio to bind to a specific network interface.
Supports requiring Bearer token authorization for SSE server requests via environment variable.
Supports running on background emulators or simulators when a real device is not connected.
Limitations and risks
Collects anonymous usage telemetry via PostHog; telemetry is optional and external to the core automation tasks.
Agent-driven workflows can perform unintended real-world actions like commenting, booking, or messaging.
Getting started
Setup difficulty is medium because it requires Node.js v20+ and external mobile SDKs. Users must install mobile platform tools, configure the mcpServers object in their MCP client settings, start a simulator or connect a real device, and prompt the agent to list available devices.
Alternatives and comparisons
A fair-code platform to build and deploy AI agents and workflows with a visual canvas, custom code, and extensive integrations.
A visual authoring experience for AI agents and workflows with built-in API and MCP servers.
A collection of tested, open-source AI agents, agent skills, and RAG apps.
README: This is a [Model Context Protocol (MCP) server](https://github.com/modelcontextprotocol) that enables scalable mobile automation, development through a platform-agnostic interface…
README: 🤖 Enables agent-to-agent communication for mobile automation usecases, data extraction
README: 🚀 **Accessibility-first — fast and cheap**: drives apps from the native accessibility tree (no vision model, no image tokens), falling back to screenshots + coordinates only when…
README: **`mobile_list_available_devices`** - List all available devices (simulators, emulators, and real devices)
README: **`mobile_install_app`** - Install an app from file (.apk, .ipa, .app, .zip)