HARV - AI Browser Assistant

HARV is a powerful browser extension that brings AI-powered assistance directly into your browsing experience. With support for multiple AI models (Gemini, Perplexity, xAI), it offers intelligent page analysis, smart search, and context-aware conversations.

⚠️ Warning: Development Status

This project is currently under active development and in beta stage. Some features might:

• Be incomplete or change without notice
• Have bugs or unexpected behavior
• Require additional optimization
• Need better error handling

Feel free to report issues and contribute to the development!

Features

🔍 Smart Search & Context-Aware Conversations

• Floating search interface accessible anywhere on the page
• Context-aware conversations with memory of previous interactions
• Multiple search scopes: Page, Domain, or Global knowledge
• Support for image analysis and screenshots

🤖 Multiple AI Models

• Gemini: Google's latest AI model with vision capabilities
• Perplexity: Advanced language model for precise responses
• xAI: Cutting-edge AI for sophisticated analysis
• Seamless model switching during conversations

📑 Page Analysis & Interaction

• Automatic page summarization
• YouTube video content analysis
• Text selection tools (Define, Explain, Search)
• Dark mode support
• Text-to-speech capabilities

Prerequisites

• Node.js (v16 or higher)
• npm (v7 or higher)
• Chrome/Edge/Firefox browser

Installation

1. Clone the repository:

   git clone https://github.com/iamakchavan/harv.git
   cd HARV

2. Install dependencies:

   npm install

3. Configure API Keys:

   • Open src/utils/ai-config.ts
   • Add your API keys:

     export const AI_CONFIG: AIConfig = {
       xai: {
         apiKey: 'YOUR_XAI_API_KEY',
         endpoint: 'https://api.x.ai/v1'
       },
       gemini: {
         apiKey: 'YOUR_GEMINI_API_KEY',
         model: 'gemini-1.5-flash'
       },
       perplexity: {
         apiKey: 'YOUR_PERPLEXITY_API_KEY',
         model: 'sonar'
       }
     };

   ⚠️ Important Deprecation Notice for Perplexity API

   Please note that as of February 22, 2025, several models and model name aliases will no longer be accessible. The following model names will no longer be available via API:

   • llama-3.1-sonar-small-128k-online
   • llama-3.1-sonar-large-128k-online
   • llama-3.1-sonar-huge-128k-online

   We recommend updating your applications to use the recently released Sonar, Sonar Pro, Sonar Pro Reasoning and Reasoning Pro models.

   Model Configuration Note:

   Each AI provider offers different model options that you can configure:

   • Gemini: Supports models like gemini-1.5-pro, gemini-1.5-flash, and other Google Vertex AI models
   • Perplexity: Use sonar for lightweight queries or sonar-pro for complex tasks requiring deeper analysis and sonar-reasoning for reasoning tasks with chain of thought.
   • xAI: Currently supports the grok-beta model

   You can modify these in the ai-config.ts file based on your specific needs and API access.

4. Build the extension:

   npm run build

5. Load the extension in your browser:

   • Chrome/Edge:
     1. Go to chrome://extensions/
     2. Enable "Developer mode"
     3. Click "Load unpacked"
     4. Select the dist folder from the project directory, and you're good to go.

Usage

Quick Access

• Use Alt+X (Windows/Linux) or Command+X (Mac) to open the extension
• Click the floating search icon for quick access > (coming soon)

Search Modes

1. Page Mode: Search within current page content
2. Domain Mode: Search & scrape across the current website
3. All Mode: Use AI's comprehensive knowledge & entire internet to get answers

Text Selection Features

1. Select any text on a webpage
2. Use the popup menu to:
   • Summarize the selection
   • Get detailed explanations
   • Copy text
   • Search for more information

Image Analysis

1. Click the camera icon to capture screenshots
2. Upload images for AI analysis
3. Ask questions about the visual content

Development

Project Structure

project/
├── src/
│   ├── components/    # React components
│   ├── utils/         # Utility functions
│   ├── App.tsx        # Main application
│   └── background.ts  # Extension background script
├── public/           # Static assets
└── dist/            # Build output

Contributing

1. Fork the repository
2. Create your feature branch
3. Commit your changes
4. Push to the branch
5. Open a Pull Request

License

This project is licensed under the MIT License - see the LICENSE file for details.