Alfi - AI Web Navigation Assistant

Alfi is an AI-powered voice assistant that helps users navigate through websites efficiently. It provides voice-based interaction to understand website content and navigate to relevant sections using natural language processing.

Features

• 🎙️ Voice command recognition for web navigation
• 🤖 AI-powered content understanding and summarization
• 🔍 Smart navigation to specific sections
• 🌐 Works with any website
• 🔊 Text-to-speech responses
• 🛡️ Secure API key handling
• 📱 Responsive floating microphone interface

Prerequisites

• Node.js 16 or higher
• Chrome browser installed
• OpenAI API key
• A working microphone

Installation

# Install globally (recommended for CLI usage)
npm install -g @collin/alfi

# Or install in your project
npm install @collin/alfi
npm install -g alfi

Configuration

Before using Alfi, you need to set up your OpenAI API key:

export OPENAI_API_KEY='your-api-key-here'

Usage

As a CLI Tool

# Start Alfi with browser interface
alfi

This will open a Chrome browser window with a microphone button in the bottom right corner. Click the button to:

1. Start voice recording (button turns red)
2. Speak your command
3. Wait for Alfi to process and execute your command

As a Module

import { BrowserController } from '@collin/alfi';

// Initialize with your API key
BrowserController.initialize('your-api-key-here');

// Start the browser controller
const controller = await BrowserController.create();
await controller.start();

Example Commands

• "Go to example.com"
• "Summarize this page"
• "Navigate to the products section"
• "Tell me what this website is about"
• "Scroll down"
• "Go back"
• "Find the pricing section"

Security

• API keys are stored securely in memory
• Voice data is processed locally before being sent to OpenAI
• Website content is analyzed locally when possible
• All communication with external services is encrypted

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

MIT License - see the LICENSE file for details

Support

If you encounter any issues or have questions, please file an issue on the GitHub repository.