Algonius Browser MCP

### 🌐 Browser Automation via Model Context Protocol **Algonius Browser** is an open-source MCP (Model Context Protocol) server that provides comprehensive browser automation capabilities to external AI systems. It serves as a bridge between AI assistants and web browsers, enabling programmatic control of web navigation, DOM interaction, and content extraction through a standardized protocol. ### ✨ Why Choose Algonius Browser? #### 🔗 **MCP Protocol Integration** - Standard interface for AI systems to control browser automation - Compatible with any MCP-compliant AI assistant or tool - Structured tool definitions with comprehensive parameter validation - Real-time resource notifications for browser state changes #### 🆓 **Open Source & Free** - No subscription fees or usage limits - Complete transparency with open-source codebase - Community-driven development and improvements - Apache 2.0 license for commercial and personal use #### 🔒 **Secure Architecture** - Chrome Native Messaging for secure communication - Local processing with no external data transmission - Sandboxed browser environment for safe automation - Minimal permissions required for operation #### 🛠️ **Enterprise-Grade Reliability** - Comprehensive test coverage with 50+ integration tests - 95%+ success rates across all automation tools - Robust error handling and graceful degradation - Cross-platform support (Windows, macOS, Linux) ### 🚀 Core MCP Tools & Resources #### **6 Browser Automation Tools** - **`navigate_to`**: Navigate to URLs with intelligent timeout handling - **`manage_tabs`**: Create, close, and switch between browser tabs - **`get_dom_extra_elements`**: Advanced DOM element extraction with pagination and filtering - **`click_element`**: Click DOM elements using CSS selectors or text matching - **`set_value`**: Set values in input fields, textareas, and form elements - **`scroll_page`**: Scroll pages in multiple directions with customizable parameters #### **2 Browser State Resources** - **`browser://current/state`**: Complete browser state in AI-friendly Markdown format - **`browser://dom/state`**: Current DOM overview with interactive elements and metadata #### **Advanced Features** - **Pagination Support**: Handle large DOM structures efficiently - **Element Filtering**: Filter elements by type (button, input, link, etc.) - **Progressive Typing**: Optimized text input for large forms - **Multi-Tab Coordination**: Manage complex workflows across multiple tabs - **Real-Time Updates**: Browser state resources update automatically ### 🎯 Perfect For #### **AI Assistant Integration** - Enable Claude, ChatGPT, or other AI systems with browser capabilities - Build custom AI tools that need web automation - Create specialized research and data collection assistants - Develop automated testing and monitoring systems #### **Developer Workflows** - Web scraping and data extraction automation - Automated testing and quality assurance - Research automation and competitive analysis - Form filling and repetitive task automation #### **Business Applications** - Lead generation and market research - E-commerce monitoring and price tracking - Content aggregation from multiple sources - Customer research and analysis automation ### 🏗️ Technical Architecture #### **Clean MCP Implementation** ``` External AI System → MCP Protocol → Go Host → Native Messaging → Chrome Extension → Browser APIs ``` #### **Component Overview** - **MCP Host**: Go-based native messaging host implementing MCP protocol - **Chrome Extension**: Background service worker with tool handlers - **Content Scripts**: DOM interaction and data extraction utilities - **Native Messaging**: Secure bidirectional communication channel #### **Technology Stack** - **Go**: High-performance MCP host with native messaging - **TypeScript**: Type-safe Chrome extension implementation - **Chrome APIs**: Advanced browser automation capabilities - **MCP Protocol**: Standardized AI-browser communication ### 📈 Common Use Cases & Workflows #### **Web Scraping Automation** 1. `navigate_to` → Go to target website 2. `browser://dom/state` → Get page overview 3. `get_dom_extra_elements` → Extract specific data with pagination 4. `scroll_page` → Load additional content 5. `browser://current/state` → Get complete extracted data #### **Form Automation** 1. `navigate_to` → Navigate to form page 2. `browser://dom/state` → Identify form fields 3. `set_value` → Fill multiple form fields 4. `click_element` → Submit form 5. `manage_tabs` → Handle form submission results #### **Multi-Site Research** 1. `browser://current/state` → Check available tabs 2. `manage_tabs` → Open multiple research targets 3. `navigate_to` → Load content in each tab 4. `get_dom_extra_elements` → Extract data from each site 5. `browser://current/state` → Compile results from all tabs ### 🛡️ Reliability & Performance #### **Proven Metrics** - 95%+ element location success rate - 98%+ navigation success rate - 97%+ form interaction success rate - Comprehensive error handling with detailed feedback #### **Advanced Capabilities** - Multi-strategy element location with intelligent fallbacks - Automatic timeout detection and optimization - Memory-efficient operation with proper resource cleanup - Graceful handling of complex websites and edge cases ### 🔧 Integration & Setup #### **Quick Installation** 1. Install Chrome extension from Web Store 2. Download and install MCP host from GitHub releases 3. Configure your AI system to connect to MCP server 4. Start automating with simple tool calls #### **System Requirements** - Chrome/Chromium 88+ or Microsoft Edge 88+ - 4GB RAM recommended for complex automations - 50MB disk space for installation - Stable internet connection for AI system communication #### **Supported Platforms** - Windows x86_64 - macOS (Intel and Apple Silicon) - Linux x86_64 ### 🌟 Developer Experience #### **Comprehensive Documentation** - Complete MCP tool schemas and examples - …