Browser Use (used By Deploya.dev)

🚀 Browser-Use MCP Server

An AI-driven browser automation server that implements the Model Context Protocol (MCP) for natural language browser control. This MCP server is built on top of browser-use/web-ui. The core browser automation logic and configuration mode are inherited from the original project.

✨ Features

• 🧠 MCP Integration - Complete protocol implementation supporting communication with AI agents.
• 🌐 Browser Automation - Page navigation, form filling, and element interaction.
• 👁️ Visual Understanding - Screenshot analysis and vision-based interactions.
• 🛠️ Development Support - Configure multiple LLM providers via environment variables.
• 🔧 Troubleshooting - Quickly identify API errors and configuration issues.

🚀 Quick Start

Project Introduction

This is an MCP server based on browser automation technology, supporting multiple Large Language Model (LLM) providers. API configuration can be done through environment variables, making it suitable for browser automation scenarios that require natural language interaction.

Functional Modules

• Core Features

  • Browser automation control.
  • Support for multiple LLM providers.
  • Visual screenshot analysis.
  • Natural language interaction interface.

• Configuration Management

  • Flexible configuration via environment variables.
  • Customizable model names.
  • Temperature parameter adjustment.
  • Context length control.

📦 Installation

Installation and Running

# Clone the repository
git clone https://github.com/your-repository.git
cd browser-use-server

# Install dependencies
npm install

# Start the service
node index.js

Configuration Example

Create a .env file in the project root directory:

MCP_MODEL_PROVIDER=anthropic  # Supported: anthropic, openai, azure_openai, etc.
ANTHROPIC_API_KEY=your-key    # Required: Anthropic API key
OPENAI_ENDPOINT=https://api.example.com  # Optional: Custom OpenAI domain

💻 Usage Examples

Basic Usage

The basic usage of this MCP server is to start the server and configure it through environment variables, then it can be used for browser automation tasks.

Advanced Usage

For more complex scenarios, you can adjust the configuration parameters according to your needs, such as customizing the model name, adjusting the temperature parameter, and controlling the context length.

🔧 Technical Details

Troubleshooting

Common Issues

• API Errors: Check the environment variables to ensure they are correctly configured and consistent with the API interfaces of the LLM providers.
• Visual Support: Ensure MCP_USE_VISION=true to enable the screenshot analysis function.

Provider Configuration

The server supports calling large language models through the following providers:

⚠️ Important Note

• For the endpoints marked as optional, the default values will be used if not specified.
• The temperature can be configured through MCP_TEMPERATURE, with a default value of 0.3.
• The model name can be customized using MCP_MODEL_NAME.
• Ollama models support additional context settings such as num_ctx and num_predict.

📄 License

This project is licensed under the MIT License. For more details, please refer to LICENSE.

Acknowledgments

Thanks to the original authors of the browser-use/web-ui project for their contributions to the field of browser automation. This project is an extension based on their work, and the copyright belongs to the original authors.