WebDriverIO

WebDriverIO is an MCP server that connects Claude Desktop to Chrome browsers and mobile environments via Appium. Built for developers, QA engineers, and automated testing practitioners, it enables language models to manage automated sessions across web and native mobile interfaces. Users can launch headed or headless Chrome instances, inspect live document structures, click components, input form text, manage cookies, and capture visual page screenshots. Beyond traditional desktop browsers, the server extends control to native and hybrid iOS (.app/.ipa) and Android (.apk) applications. Through Appium, it exposes commands for touch gestures like swiping, tapping, and long-pressing, as well as device operations including screen orientation adjustments, app lifecycle events, and hybrid context switching between native views and webviews. While this repository has transitioned into maintenance following the launch of the official @wdio/mcp package, it provides complete Model Context Protocol tools for end-to-end web testing, viewport inspection, and automated mobile device validation.

Category: Browser & Web Automation

Tags: browser automation, screenshots, webdriverio, web scraping

Visit WebDriverIO

How to install and configure WebDriverIO

Follow these steps to configure the WebDriverIO MCP server with Claude Desktop: 1. Ensure you have Node.js (version 18 or higher) and Chrome installed on your machine. 2. Open your Claude Desktop MCP configuration file. 3. Add the server entry under the mcpServers object: json { "mcpServers": { "webdriverio-mcp": { "command": "npx", "args": ["-y", "webdriverio-mcp"] } } } 4. Save the configuration and restart Claude Desktop completely. 5. If automating mobile apps, start an Appium server running locally (appium) before invoking mobile tools.

What you can do with WebDriverIO

• Running automated web navigation tests to inspect page elements, fill out inputs, and capture screenshots for UI verification. - Validating native Android APK files on an emulator by automating taps, swipes, and permissions handling. - Testing iOS applications across simulator and physical device setups using UDID targeting and native XCUITest gestures. - Managing hybrid application workflows by switching execution contexts seamlessly between native app layouts and embedded webviews. - Auditing browser page structures using accessibility tree queries and viewport-specific element discovery commands.

Key facts

• Open Source
• https://github.com/Winify/webdriverio-mcp
• Browser & Web Automation, Design, Media & Creative
• browser automation, screenshots, webdriverio, web scraping

Part of MCP Servers

What can WebDriverIO MCP do?

WebDriverIO MCP allows LLMs to control Chrome browsers and mobile devices. It can launch browser sessions, interact with buttons and forms, scroll, extract accessibility trees, take screenshots, manage cookies, and automate native or hybrid iOS and Android applications via Appium gestures and context switching.

How do I install WebDriverIO MCP?

You can install it directly inside Claude Desktop by adding an npx entry for webdriverio-mcp to your MCP configuration file. You will need Node.js version 18 or higher and Google Chrome installed. For mobile automation, you must also run an Appium server locally.

Which MCP clients work with WebDriverIO MCP?

WebDriverIO MCP is designed primarily for Claude Desktop, though it follows standard Model Context Protocol specifications and can run in compatible MCP client applications that support local stdio server commands.

Is WebDriverIO MCP open source?

Yes, WebDriverIO MCP is an open-source project hosted on GitHub. Note that the original repository is archived, and the community is encouraged to migrate to the official @wdio/mcp package maintained by the WebDriverIO project.

Does WebDriverIO MCP support mobile testing?

Yes, it supports mobile testing for both Android and iOS applications by connecting through a local Appium server. It provides tools for device gestures, orientation, keyboard control, and context switching between native app views and webviews.