Mobilewright CLI
A Playwright-style iOS and Android framework: the agent drives a phone through the accessibility tree and a skill writes tests.
- 471 ★ on the repository
- Last commit
- Apache-2.0
Extended access
Why this level
- The agent taps buttons and enters text on a simulator or a real phone, including working accounts.
- Cloud providers receive the app build and a device session.
Install
- Works in
Manual install
npm install mobilewright @mobilewright/test
npx mobilewright initThen `npx mobilewright doctor` checks the environment and `npx mobilewright test` runs tests.
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- The agent taps buttons and enters text on a simulator or a real phone, including working accounts.
- Cloud providers receive the app build and a device session.
This is third-party code. Review the repository files before installing.
What it does
Mobilewright automates iOS and Android devices, simulators and emulators through a Playwright-style API built on `mobilecli`. Elements are found by role and label (`screen.getByRole("button", { name: "Sign In" }).tap()`) and actions wait on their own until the element is visible, enabled and stable. Assertions such as `expect(...).toBeVisible()` poll the screen until the timeout. `@mobilewright/test` fixtures extend Playwright Test with `screen` and `device` objects. For agents the framework exposes the screen accessibility tree: deterministic, token-efficient, with no vision model and no coordinates. It is used together with mobile-mcp, Claude or Cursor. The `mobilewright-cli` skill teaches an agent the `mobilewright-cli devices`, `launch`, `snapshot`, `fill`, `tap` and `expect` commands and to write tests from the results. `npx mobilewright doctor` checks Xcode, the Android SDK, simulators and ADB. You can connect to a remote `mobilecli` and to cloud devices at Mobile Next Cloud, BrowserStack and Sauce Labs.
Who it is for. QA engineers and mobile developers who ask Claude Code or Cursor to walk through app scenarios and record them as tests.
Good fit when
- The agent must walk through sign-in in an app on a simulator and write a test itself.
- One test set is needed for iOS and Android with auto-waiting instead of manual pauses.
- Tests must run on remote devices in a lab or in the cloud.
- Xcode, ADB and simulators need a check with `doctor` before the first run.
Not a fit when
- A web app with no mobile shell, where regular Playwright is enough.
- There is no simulator, emulator or connected device.
- Node.js is older than 22.12.
Example request
Launch the app on the booted simulator, walk through sign-in with `mobilewright-cli` and turn the scenario into a test that checks "Welcome back".Limitations
Node.js 22.12 or newer and a booted iOS simulator, Android emulator or connected device are required. `mobilecli` handles the devices and Mobilewright starts it by itself for local addresses. iOS needs Xcode and Android needs the Android SDK and ADB, and `npx mobilewright doctor` shows what is missing. Cloud devices at BrowserStack, Sauce Labs and Mobile Next Cloud need separate accounts. The last commit in the repository is from November 2026.
How to disable. Run `npm uninstall mobilewright @mobilewright/test` in the project and delete the `mobilewright-cli` skill from the agent skills directory.
FAQ
How do I install Mobilewright?
With `npm install mobilewright @mobilewright/test` and `npx mobilewright init`. Tests run with `npx mobilewright test`.
How does it differ from Appium?
Mobilewright has a Playwright-style API with built-in auto-waiting and setup comes down to one npm command. Appium needs a server and drivers and waits are configured by hand.
How does the agent see the screen?
Through the accessibility tree: roles, labels and elements with refs. No vision model, XPath or coordinates are needed, and the output saves tokens.
Which commands does the mobilewright-cli skill know?
`devices`, `launch`, `snapshot`, `fill`, `tap` and `expect`. After each action the CLI prints a fresh screen snapshot with element refs.
Can it run in the cloud?
Yes, Mobile Next Cloud, BrowserStack and Sauce Labs are supported, along with a remote `mobilecli` for your own device lab.
Discussion
No questions or reviews yet.
Related
CLI toolsPlaywright CLI
playwright-cli
Microsoft's official Playwright CLI with an agent skill: drive a browser through short commands without heavy MCP schemas
CLI, MCP server and typed Node.js API so an agent can verify its changes in a live mobile app on iOS, Android and more
A community Playwright MCP server: browser automation, screenshots, device emulation and test code generation
A skill giving Claude Code a background browser: screenshots at three sizes, frontend audits, before and after diffs, login flows.
Runs Maestro flows on Android, iOS and in the browser as a single binary without a JVM and installs a skill for the agent.
tapflow MCP server: an agent controls iOS simulators and Android emulators on your Mac and replays flows for CI
README in short
Playwright for phones: one API for iOS and Android, auto-waiting and an agent mode over the accessibility tree. A skill writes tests for you.