Super_Assistant

NanoBrowser

Open source multi-agent browser automation tool with built-in Chrome extension. NanoBrowser makes it easy to perform complex browser tasks using natural language commands.

Features

• 🤖 Multi-agent system for complex web automation tasks
• 🔌 Built-in Chrome extension with chat interface
• 🧠 Natural language commands for browser control
• 💬 Interactive chat sidebar for real-time interaction and status updates

Quick Start

Prerequisites

1. Chrome Browser:
2. Chrome browser installed

3. Chrome Extension Management page (chrome://extensions/) with "Developer mode" enabled

4. Python Environment:

5. Python 3.10 or higher

Installation

1. Install the Chrome Extension:
2. Open Chrome and go to chrome://extensions/
3. Enable "Developer mode" (top right)
4. Click "Load unpacked" (top left)

5. Select the nanobrowser/extension/src directory

6. Install the NanoBrowser Agent Server: ```bash

   Navigate to nanobrowser directory

   cd nanobrowser

Run installation script

./install.sh

> Note: The installation script has only been tested on MacOS. For other platforms, please refer to [uv's official documentation](https://github.com/astral-sh/uv) to install uv and dependencies manually.

3. **Configure API Keys**:

Create a `.env` file in the nanobrowser directory and add your API key. The default LLM provider is OpenAI:
```bash
OPENAI_API_KEY=your_openai_key_here

If you want to use Claude (Anthropic), you'll need to modify the provider in config.yaml and use:

ANTHROPIC_API_KEY=your_anthropic_key_here

1. Run NanoBrowser: bash uv run nanobrowser ## Example Tasks Here are some tasks you can accomplish with NanoBrowser using natural language: 1. News Extraction: > "Go to TechCrunch and extract top 10 headlines from the last 24 hours" 2. GitHub Research: > "Look for the trending Python repositories on GitHub with most stars" 3. Shopping Research: > "Find a portable Bluetooth speaker on Amazon with a water-resistant design, under $50. It should have a minimum battery life of 10 hours" 4. Recipe Analysis: > "Find a recipe with over 100 reviews for Fried Fish on Allrecipes, list the Full Nutrition Label and tell me the amount of Iron per Serving" ## Roadmap Here are the planned improvements and features we're working on: ### Standalone Extension - Developing a standalone Chrome extension without backend server dependency - Simplifying installation and deployment process - Improving overall system performance and reliability ### Enhanced Tab Management - Supporting multi-tab operations - Adding tab context awareness and management - Implementing better tab state synchronization ### Conversation Intelligence - Adding support for follow-up questions and continuous dialogue - Implementing context preservation between tasks - Implementing task and conversation history management ### Advanced DOM Analysis - Improving recognition and interaction with complex DOM elements - Enhancing canvas element handling for applications like Google Docs - Optimizing DOM tree traversal and manipulation performance - Implementing better detection of interactive elements and dynamic content --- ## Contributing Contributions are welcome! Feel free to: - Report bugs - Suggest new features - Submit pull requests ## Acknowledgments Special thanks to these inspiring projects: - python-cdp - Agent-E - fuji-web ## License This project is licensed under the Apache License 2.0 - see the LICENSE file for details. Made with ❤️ by the NanoBrowser Team