Smart Tracker

Smart Tracker is a browser extension designed to help you build a private digital library of the content you consume online. It automatically logs the articles you read and videos you watch, using a local AI model to summarize them. You can then share your intellectual journey with friends you connect with.

Overview

We all consume a vast amount of information online, but keeping a meaningful record of it is a challenge. Smart Tracker was created to solve this by providing an effortless way to curate your most insightful readings and viewings without any manual work.

The browser extension identifies significant content during your browsing sessions. A lightweight, privacy-preserving AI model running directly in your browser then summarizes this content and adds it to your personal library. This library is private by default but can be shared with friends. The sharing model is based on mutual consent: once two users accept each other's friend request, they can view each other's libraries.

Core Features

• Automated Content Logging: The extension seamlessly tracks relevant articles and videos in the background.
• AI-Powered Curation: A local, in-browser language model analyzes and selects meaningful content.
• Automatic Summaries: Key insights from your content are condensed into brief summaries.
• Private Friend-Based Sharing: Your curated library is private. You can send and receive friend requests, and only upon mutual acceptance can you and a friend view each other's libraries.
• Privacy-Focused: All AI analysis happens locally in your browser, and your library is only shared with friends you approve.

Technical Details

• Backend: Node.js, Express.js (API for user management, friendships, and data storage)
• Frontend (Extension): JavaScript, HTML, CSS
• AI: WebLLM running Llama-3.2-1B-Instruct-q4f16_1-MLC.
• Database: MongoDB

Getting Started: Setup and Installation

Follow these instructions to get the project running on your local machine. The backend is required to handle user accounts, friendships, and data persistence for the extension.

Prerequisites

Please ensure you have the following installed:

• Node.js (which includes npm)
• Git
• Google Chrome

1. Clone the Repository

Start by cloning the project repository to your computer.

git clone https://github.com/37nomad/ReadingAnd_Watching_Extension.git
cd ReadingAnd_Watching_Extension

2. Configure and Run the Backend

The backend server is responsible for managing user data and API requests for the extension.

1. Navigate to the backend directory:

   cd backend

2. Create an Environment File: Before installing dependencies, create a .env file in the backend directory. This file will store your environment variables.

   touch .env

   Open the .env file and add the following variables. Replace the placeholder values with your actual configuration details.

   PORT=6969
   MONGODB_URI=your_mongodb_connection_string
   JWT_SECRET=your_super_secret_jwt_key

3. Install Dependencies:

   npm install

4. Start the Server:

   npm run dev

   The backend server will now be running on the port you specified (e.g., http://localhost:6969).

3. Build the Chrome Extension

The smart-tracker folder contains the source code for the browser extension.

Note for Quick Testing:

To reduce the inference time from 3 minutes to 15 seconds for quick testing, you can do the following:

Comment out Line 492 in the content.js file.

(Refer to the guidelines provided in the file for more details.)

1. Navigate to the extension directory (from the project's root folder):

   cd ../smart-tracker

2. Install Dependencies:

   npm install

3. Build the Extension:

   npm run build

   This command creates a dist folder containing the production-ready files for the extension.

4. Load the Extension in Chrome

Finally, load the built extension into your browser.

1. Open Chrome and go to chrome://extensions.
2. Enable Developer mode in the top-right corner.
3. Click Load unpacked.
4. Select the smart-tracker/dist folder from your project directory.

The Smart Tracker extension will now be active in your browser and connected to your local backend.

5. Debugging of LLM's Output

1. We can check it in the devtools of the browser regarding the model downloading or model loading from browser storage cache or the output generation i.e Summary.

How to Contribute

We welcome contributions. If you'd like to improve Smart Tracker, please feel free to fork the repository, make your changes, and submit a pull request.

License

This project is licensed under the MIT License. See the LICENSE file for more details.