LLM App Exploration
A pattern for LLM agents to explore and control any app — mobile or desktop. Accessibility-first, not vision-based. Share with your agent and go.
What is LLM App Exploration?
LLM App Exploration is a Model Context Protocol (MCP) server that allows AI assistants like Claude, Cursor, and VS Code to pattern for llm agents to explore and control any app — mobile or desktop. accessibility-first, not vision-based. share with your agent and go.
A pattern for LLM agents to explore and control any app — mobile or desktop. Accessibility-first, not vision-based. Share with your agent and go.
This server falls under the Browser Automation category on MCPgee, the world's largest MCP server directory with 33,000+ servers.
Features
- A pattern for LLM agents to explore and control any app — mo
Use Cases
Maintainer
Works with
Installation
Manual Installation
npx llm-app-explorationConfiguration
Configuration Details
claude_desktop_config.json
Performance
Response Metrics
Resource Usage
How to Set Up and Use LLM App Exploration
LLM App Exploration is an accessibility-first pattern for giving AI agents the ability to explore and control mobile and desktop applications without relying on computer vision. It uses two companion MCP servers — mobile-mcp for Android and iOS devices, and desktop-mcp for macOS native and Electron apps — that expose the app's accessibility tree to the LLM, enabling it to read UI elements, tap buttons, type text, scroll, and navigate across screens. This approach is ideal for automated UX research, mobile testing, and building AI-driven QA workflows.
Prerequisites
- Node.js 18+ or a compatible TypeScript runtime
- For Android: Android Debug Bridge (ADB) installed and a device/emulator with USB debugging enabled
- For iOS: Xcode and a connected simulator or device with developer access
- For macOS desktop: Accessibility API permission granted in System Settings > Privacy & Security > Accessibility
- An MCP-compatible AI client such as Claude Desktop or Claude Code
Clone the repository
Clone the llm-app-exploration repository which contains both the mobile-mcp and desktop-mcp servers.
git clone https://github.com/ForrestKim42/llm-app-exploration.git
cd llm-app-explorationInstall dependencies
Install all Node.js dependencies for the project.
npm installGrant accessibility permissions (macOS desktop automation)
For macOS native app control, you must grant Accessibility API access. Go to System Settings > Privacy & Security > Accessibility and enable your terminal application.
Connect your Android device (mobile automation)
Enable USB debugging on your Android device and connect it. Verify ADB recognizes the device before configuring the MCP server.
adb devicesConfigure the appropriate MCP server in your client
Add mobile-mcp or desktop-mcp (or both) to your Claude Desktop or Claude Code MCP configuration, depending on your target platform.
Ask the agent to explore an app
Once connected, prompt the AI to explore a specific app. The agent will map screens to ROUTES.md and document transitions to TRANSITIONS.md before executing flows.
LLM App Exploration Examples
Client configuration
Configure mobile-mcp for Android app exploration in Claude Desktop.
{
"mcpServers": {
"mobile-mcp": {
"command": "npx",
"args": ["llm-app-exploration"],
"env": {}
}
}
}Prompts to try
Examples of app exploration and automation tasks you can perform with this server.
- "Explore the Settings app on the connected Android device and map all screens"
- "Navigate to the login screen of the app and fill in the credentials"
- "Tap the BUTTON:Submit element and take a screenshot of the result"
- "Find all text input fields on the current screen and list their labels"
- "Run a complete UX walkthrough of the checkout flow in the shopping app"Troubleshooting LLM App Exploration
ADB device not found or no devices listed
Ensure USB debugging is enabled in the Android device's Developer Options. Run `adb kill-server && adb start-server` to reset the ADB daemon, then reconnect the device and run `adb devices` again.
macOS desktop automation fails with permission errors
Open System Settings > Privacy & Security > Accessibility and ensure your terminal app (Terminal, iTerm2, or VS Code) is listed and checked. You may need to remove and re-add the app after granting permissions.
Electron app elements not visible in accessibility tree
Launch the Electron app with the `--remote-debugging-port=PORT` flag (e.g., `--remote-debugging-port=9222`) to enable Chrome DevTools Protocol access, which provides DOM-level element visibility beyond the OS accessibility layer.
Frequently Asked Questions about LLM App Exploration
What is LLM App Exploration?
LLM App Exploration is a Model Context Protocol (MCP) server that pattern for llm agents to explore and control any app — mobile or desktop. accessibility-first, not vision-based. share with your agent and go. It connects AI assistants to external tools and data sources through a standardized interface.
How do I install LLM App Exploration?
Follow the installation instructions on the LLM App Exploration GitHub repository. Clone the repo, install dependencies, and add the server config to your AI client.
Which AI clients work with LLM App Exploration?
LLM App Exploration works with all major MCP-compatible AI clients including Claude Desktop, Claude Code, Cursor, VS Code (GitHub Copilot), Windsurf, and Cline.
Is LLM App Exploration free to use?
Yes, LLM App Exploration is open source and available under the MIT license. You can use it freely in both personal and commercial projects.
LLM App Exploration Alternatives — Similar Browser Automation Servers
Looking for alternatives to LLM App Exploration? Here are other popular browser automation servers you can use with Claude, Cursor, and VS Code.
Chrome DevTools MCP
★ 40.6kAI-powered Chrome automation server with natural language element detection. Control Chrome browser through MCP protocol for testing, debugging, and performance analysis. Features 91% accuracy in element location, works with free AI models, and suppo
UI TARS Desktop
★ 34.9k📇 🏠 - Browser automation capabilities using Puppeteer, both support local and remote browser connection.
Playwright
★ 32.8kA production-ready browser automation server that enables AI assistants to interact with web pages using tools for navigation, element interaction, and data extraction. It features a built-in Inspector UI and robust crash recovery for reliable automa
Page Agent
★ 18.0kJavaScript in-page GUI agent. Control web interfaces with natural language.
Chrome
★ 11.7kAn extension-based MCP server that enables AI assistants to control your browser, leveraging existing sessions and login states for automation and content analysis. It provides over 20 tools for semantic tab search, interactive element manipulation,
LAMDA
★ 7.8kThe most powerful Android RPA agent framework, next generation mobile automation.
Browse More Browser Automation MCP Servers
Explore all browser automation servers available in the MCPgee directory. Each server includes setup guides for Claude, Cursor, and VS Code.
Set Up LLM App Exploration in Your Editor
Choose your AI client for step-by-step setup instructions.
Quick Config Preview
Add this to your claude_desktop_config.json or .cursor/mcp.json
Ready to use LLM App Exploration?
Browse our complete directory of 33,000+ MCP servers, read setup guides for your editor, and start building with the Model Context Protocol.