Control Android devices with an on-device AI agent using PokeClaw

Run an on-device Android control loop that inspects screens and performs app actions locally for mobile automation and testing workflows.

Prerequisites

PokeClaw; Android device or emulator; project-documented runtime dependencies

Installation

Requirements and caveats from upstream:

• DroidRun/Mobilerun local/open-source runs the agent loop in a Python CLI/container on a host machine and talks to a Portal APK through ADB/port forwarding. The Portal APK uses Android Accessibility for tree/gesture/sc...
• minitap/mobile-use, Mobile-Agent, AppAgent, and AppAgentX are external SDK/research harnesses that normally require Python plus ADB, UIAutomator, screenshots, Docker, cloud phones, or host-side model/API services.
• does not require a computer to be connected while tasks run

Basic usage or getting-started notes:

• It can run Gemma 4 on-device for local, private phone control, and it also supports optional cloud models when you want stronger reasoning for harder tasks.

• task state leaks into the next run

• model download, storage, or LiteRT startup fails before the model can run

• Source: https://github.com/agents-io/PokeClaw

• Extracted from upstream docs: https://raw.githubusercontent.com/agents-io/PokeClaw/HEAD/README.md

Documentation

• https://github.com/agents-io/PokeClaw

Source

• Agent Skill Exchange