FocusShield AI

About

🔒 Safe Browsing with AI  
FocusShield AI uses on‑device and server‑hosted machine learning to automatically block or obscure unsafe content—both text and images—so children can explore the web with confidence.

📚 Designed for Families  
• Password‑protected barrier page: prevent access to mature sites until a guardian code is entered  
• Popup controls: instantly pause filtering or rescan the current tab  
• Options page: pick which content classes to block (Profanity, Drugs, Violence, Gambling, Social Media, Explicit)

✅ Key Features  
1. TEXT FILTERING: replaces flagged words with “█”  
    - Example: “I like playing █ in the casino.”  
2. IMAGE FILTERING: swaps NSFW images for a child‑friendly placeholder  
3. CUSTOM THRESHOLDS: set confidence levels for text/image flagging  
4. LOGGING & REPORTS: view daily/30‑day logs of blocked categories  
5. LIGHTWEIGHT: runs as a Chrome MV3 service worker—minimal CPU & memory usage

🚀 How It Works  
1. You browse as usual.  
2. On each page load or on‑demand via the popup, URLs and page content are sent (securely, over HTTPS) to your trusted server.  
3. Our AI model scores each sentence or image.  
4. Anything above your threshold is automatically hidden or replaced.

⚙️ Configuration  
Open the “Options” page to:  
- Enable/disable specific categories  
- Adjust text vs. image thresholds  
- Clear your logs  

🔐 Privacy & Security  
• LOCAL DATA: only your filtering logs (timestamps + category codes) are stored in Chrome storage, and purged after 30 days.  
• SERVER: only page fragments or images are sent for scoring—nothing is retained on the inference server beyond the brief processing time.  
• NO PERSONAL INFO: we never collect usernames, passwords, browsing history, or cookies.