A browser extension to detect misinformation across the web

Features • Installation • How It Works • Backend Setup • Development • License

Pinocchio servers going offline. Your own server, api key, and build of the extension is now required to use it.

Pinocchio is a powerful browser extension that helps users navigate the complex landscape of online information by:

• Analyzing web content for potential misinformation using AI
• Providing reliability scores from 0-10 (with Pinocchio's nose growing accordingly!)
• Generating comprehensive reports with context and verified sources
• Creating a database of all analyzed pages for historical reference
• Supporting sharing of fact-check results

1. Download the extension ZIP file from our website
2. Unzip/extract the downloaded file to a location on your computer
3. Open Chrome and navigate to chrome://extensions/
4. Enable "Developer Mode" in the top-right corner
5. Click "Load unpacked" and select the extracted extension folder
6. The Pinocchio icon should now appear in your browser toolbar

The extension requires the backend server to be running:

1. Set up the server as described in the Backend Setup section
2. The extension is configured to connect to http://172.105.18.148:8080 by default
3. To use a different server address, modify popup.js and manifest.json accordingly

1. Installation: Add the Pinocchio extension to your browser.
2. Browse & Analyze: When visiting a webpage, click the Pinocchio icon to analyze the content.
3. Review Results: Get instant feedback with:
   • A misinformation score (0-10)
   • Visual indication via Pinocchio's growing nose
   • A detailed report explaining potential issues
   • Links to verified sources
   • Additional context about the topic
4. Download Reports: Save detailed HTML reports for future reference.
5. Access Dashboard: Visit the report dashboard to view all previously analyzed pages.

The backend server uses Flask with Google's Gemini AI for content analysis:

• Python 3.8+
• Google Gemini API key
• SQLite (included with Python)

1. Create a .env file in the backend directory with:

GEMINI_API_KEY=your_gemini_api_key
DATABASE_PATH=reports.db
REPORTS_DIR=reports

2. Install dependencies:

cd backend
pip install flask flask-cors google-generativeai python-dotenv

3. Start the server:

python app.py

The server will run at http://127.0.0.1:8080 by default.

Pinocchio/
├── extension/             # Chrome extension files
│   ├── images/            # Extension icons and images
│   ├── libs/              # JavaScript libraries
│   ├── background.js      # Extension background script
│   ├── content.js         # Content script for web pages
│   ├── manifest.json      # Extension manifest
│   ├── popup.html         # Extension popup UI
│   ├── popup.js           # Extension popup logic
│   └── styles.css         # Extension styling
│
├── backend/               # Server-side code
│   ├── reports/           # Stored HTML reports
│   ├── static/            # Static assets for web UI
│   ├── templates/         # HTML templates
│   │   ├── index.html     # Homepage template
│   │   └── reports.html   # Reports dashboard template
│   ├── app.py             # Flask application
│   ├── db_service.py      # Database operations
│   └── gemini_service.py  # Gemini AI integration
│
└── README.md              # This file

1. Make changes to the extension files in the extension directory
2. Reload the extension in Chrome (chrome://extensions/ and click the refresh icon)

1. Make changes to the server files in the backend directory
2. Restart the Flask server to apply changes

• POST /api/analyze: Analyze webpage content for misinformation
• POST /api/reports: Save a new report to the database
• GET /api/reports: Retrieve list of all reports
• GET /api/reports/{id}: Get specific report data
• GET /api/reports/{id}/html: Get HTML content of a specific report

• Adrian Maier
• Dimeji Aiyesan
• Ashton Grant
• Hariz Shirazi

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

• Inspired by the classic tale of Pinocchio, whose nose grows when he tells lies
• Powered by Google's Gemini AI for content analysis
• Built with Flask, SQLite, Gemini, and Chrome Extension APIs