$ loading_
Control iOS and Android devices from CLI for automation and agent testing.
The materials indicate an open-source local CLI for controlling iOS/Android devices, with no declared secrets or remote endpoints. Overall supply-chain trust is relatively strong, but its local execution and device-control capabilities still warrant caution for code execution and data access.
The material explicitly states there are no required keys or environment variables, and no API tokens, cloud accounts, or other sensitive credentials are mentioned; based on the available facts, credential exposure appears low.
No remote endpoints or external service hosts are declared, and the provided material does not show user data being sent to third-party network services; based on known facts, there is no explicit data-egress path.
The system checks mark it as executes-code, and its stated function is to control iOS/Android devices via CLI, implying local process execution and access to device-related capabilities. This is a standard high-privilege capability for this type of MCP tool and should be used in a controlled environment.
As a device-control CLI, it may typically interact with UI state, status, or related debugging data on connected devices; while the material does not specify exact read/write scope and shows no explicit overbroad permission request, device interaction itself implies some data exposure surface, so least-privilege deployment is advisable.
The source is an open-source GitHub repository under the MIT license with about 2.5k stars, making the code auditable and showing meaningful community adoption; although maintenance status is unknown, the available positive signals are sufficient to keep supply-chain risk low.
Copy the install command and let the AI configure it · recommended for beginners
No copy-paste install info for "agent-device" yet — see the docs or source repo.
Use agent-device to connect to an Android test device, launch our app, perform login, open the home screen, navigate to settings, and take screenshots; if any step fails, report the failed step and a summary of device logs.
A test result report with step screenshots, failure details, and a log summary.
Use agent-device to control an iPhone, open the latest build, go to the profile page, edit the avatar three times, record the screen, and when layout misalignment appears, save a screenshot and the current view hierarchy information.
A reproduction video, issue screenshots, and view hierarchy data for debugging.
Use agent-device to list all connected iOS and Android devices, then collect each device's OS version, resolution, battery level, network status, and installed target app version, and organize the results into a table.
A device inventory table for test preparation, compatibility checks, and environment verification.
Control Android devices with natural language for UI and system automation.
Let AI agents observe and act on Android and iOS devices.
Control Android devices with natural language for native app testing automation.
Give agents real phone numbers for calls, messages, and voice workflows.
Keep a private, local-first diary through any AI agent.
Control Android devices via ADB for UI interaction and inspection.