SafePost AI is a browser extension designed to analyze images and text on social media platforms, providing privacy-preserving AI-powered insights and content moderation. All processing is performed locally in your browser, ensuring that your data and images never leave your device.
- Local Image Analysis: Detects and analyzes images in social media posts using advanced AI models.
- OCR (Optical Character Recognition): Extracts text from images using Tesseract.js, all within your browser.
- Privacy-First Design: No images or data are ever uploaded to any server. All analysis is performed locally.
- Content Moderation: Identifies potentially sensitive or private information in posts before you share them.
- User-Friendly Interface: Simple popups and notifications to guide users and inform them about privacy.
- Image Detection: The extension scans your social media feed for images.
- Local OCR: When an image is detected, Tesseract.js is loaded (if not already present) directly in your browser. The image is processed locally to extract any text content.
- AI Analysis: The extracted text and image data are analyzed using AI models (also running locally) to detect sensitive information or privacy risks.
- Privacy Notice: A modal popup informs users that all processing is local and no data is sent to any server.
- User Action: Users are notified of any detected risks and can choose to proceed or edit their content.
Below is an example of the SafePost AI Privacy Risk Detection modal in action:
This example showcases SafePost AI detecting multiple sensitive items at a medium risk level and providing clear, actionable options to help users safeguard their privacy.
- No Data Upload: All image and text analysis is performed in the browser. No images, text, or metadata are sent to any external server at any point.
- Tesseract.js Integration: The extension dynamically loads Tesseract.js from a trusted CDN only if needed, and uses it exclusively in the browser context.
- No Tracking: The extension does not track user activity, store data externally, or use analytics scripts.
- Transparency: Users are always informed via a clear privacy notice modal before any analysis is performed.
- Open Source: The code is open for inspection, so users and developers can verify privacy claims.
background.js– Handles background tasks and extension events.content.js– Injects scripts into web pages and manages DOM interactions.manifest.json– Extension manifest and permissions.api/ai-privacy-analyzer.js– AI models for privacy analysis.api/huggingface-service.js– (If used) Local integration with HuggingFace models.api/image-analysis-service.js– Handles image analysis and OCR using Tesseract.js.options.html,options.js,options-styles.css– Extension options UI.popup.html,popup.js,popup-styles.css– Popup UI for user notifications.styles.css– Shared styles.icons/– Extension icons.
- Install the Extension: Load the extension in your browser (see INSTALL.md for instructions).
- Browse Social Media: As you use platforms like LinkedIn, the extension will automatically analyze images and text in your feed.
- Review Privacy Notices: When analysis is performed, a privacy notice will appear, confirming that all processing is local.
- Take Action: If sensitive content is detected, follow the extension's guidance to protect your privacy.
- The extension only requests permissions necessary for its operation.
- All third-party libraries (like Tesseract.js) are loaded from trusted sources.
- No persistent storage of analyzed data; all processing is ephemeral and local.
Contributions are welcome! Please see the code and submit issues or pull requests for improvements.
This project is licensed under the MIT License.
For questions or support, please open an issue on the repository.