Powered by Machine Learning + Google Gemini Multimodal AI
.jpeg){kind=link}
.jpeg){kind=link}
.jpeg){kind=link}
.jpeg){kind=link}
.jpeg){kind=link}
.jpeg){kind=link}
.jpeg){kind=link}
.jpeg){kind=link}
TrapEye is an AI-powered cybersecurity platform designed to protect users from modern digital threats such as:
- Phishing links
- Deepfake media
- Fake news & misinformation
- Job scams
- Payment fraud
- Credential breaches
The system combines machine learning, browser extensions, threat intelligence APIs, and community reporting to deliver real-time protection across web, messaging platforms, and browsers.
TrapEye is a full-stack AI cybersecurity ecosystem that uses a hybrid of traditional machine learning and Google Gemini multimodal AI to detect digital threats across multiple vectors — on the web, via browser extensions, WhatsApp, and a dedicated mobile app.
| Module | Purpose | Tech |
|---|---|---|
| TrapEyeX Web Platform | Full threat intelligence dashboard | Next.js 14, FastAPI, Gemini AI |
| Chrome Extension (LinkLens) | In-browser phishing link scanning | Manifest V3, TrapEye API |
| WhatsApp Bot (LinkLens) | Instant URL scanning via WhatsApp | Twilio / WA API, Python |
| Sentinelix Mobile App | Comprehensive mobile cybersecurity guardian | React Native / Flutter |
The flagship web dashboard featuring URL Scanner, Fake News Detector, Deepfake Analyzer, Threat Sandbox, and Live Intelligence Dashboard.
Full-stack Next.js + FastAPI application with Gemini AI integration.
An AI-powered mobile security guardian protecting users from phishing, scams, QR fraud, job scams, and credential breaches in real-time.
Splash & Onboarding
Main Dashboard — Device Protection Score
QR Shield — AI Threat Interception for QR Codes
.jpeg)
Credential Guard — Email Breach Checker
.jpeg)
Credential Guard — Password Leak Checker
Threat Radar — Real-Time India Feed
.jpeg)
Job Shield — AI Recruitment Scam Analyzer
Guardian Setup — Family Emergency Alerts
QR Scam Scanner — Live Camera Mode
.jpeg)
All Tools Overview
A browser extension that scans any link in real-time as you hover or click, flagging phishing attempts directly in your browser.
Key Features:
- One-click URL scanning via TrapEye API
- Visual risk indicator (green/yellow/red)
- Works on all websites in real-time
- Manifest V3 compliant
Send any suspicious URL to our WhatsApp bot and get an instant threat analysis report — no app download required.
Key Features:
- Powered by Twilio/WhatsApp Business API
- Returns risk level, phishing probability, and reasons
- Ideal for non-tech-savvy users and senior citizens
Extracts 20+ features including:
- URL length, dot count, subdomain depth
- IP address usage, HTTPS status
- Domain entropy (randomness)
- Suspicious TLD detection (
.xyz,.tk,.ml) - Brand impersonation check
- Phishing keyword detection
- ML Layer: TF-IDF vectorization + Logistic Regression
- Language Analysis: Sensationalism, clickbait, emotional language patterns
- Source Credibility: Pre-scored database of 30+ news sources
- Gemini Layer: Contextual reasoning and fact-check analysis
- Heuristic Analysis: JPEG quality, compression artifacts
- Gemini Vision: Facial feature analysis, lighting consistency
- Supports: JPEG, PNG, WebP, MP4, WebM (up to 50MB)
- Isolated virtual environment for threat payload analysis
- Behavioral heuristics logging
- Simulated C2 connection interception
- Real-time containment status tracking
- Fetches from NewsAPI across 6 categories
- Auto-scores every article for credibility
- Flags suspicious articles for alerts
- Auto-refresh every 5 minutes
- Live threat timeline and distribution charts ======= 🌐 Web Application
https://trapeye-demo.onrender.com/
- Next.js 14 (App Router)
- TailwindCSS (custom cybersecurity theme)
- Framer Motion (page transitions + animations)
- Recharts (threat charts)
- Axios (API client)
- FastAPI (Python)
- SQLAlchemy + PostgreSQL
- Scikit-learn (ML models — Random Forest, Logistic Regression, TF-IDF)
- Google Gemini API (multimodal AI for reasoning + vision)
- NewsAPI (real-time news feed)
- Chrome Extension: Manifest V3, Vanilla JS
- WhatsApp Bot: Python, Twilio API / WhatsApp Business API
trapeyeX/
├── backend/ # FastAPI backend
│ ├── main.py # App entry point
│ ├── config/ # Environment configuration
│ ├── routes/ # API endpoints
│ │ ├── phishing.py
│ │ ├── fakenews.py
│ │ ├── deepfake.py
│ │ ├── news_monitor.py
│ │ └── dashboard.py
│ ├── services/ # Business logic
│ │ ├── phishing_service.py
│ │ ├── fakenews_service.py
│ │ ├── deepfake_service.py
│ │ ├── news_service.py
│ │ └── gemini_service.py
│ ├── models/ # DB models & schemas
│ └── utils/ # Feature extractor, ML utils
├── frontend/ # Next.js 14 web app
│ ├── app/
│ │ ├── page.tsx # Landing page
│ │ ├── scanner/ # URL phishing scanner
│ │ ├── fakenews/ # News analyzer
│ │ ├── deepfake/ # Deepfake detector
│ │ ├── sandbox/ # Threat sandbox
│ │ └── dashboard/ # Threat intelligence dashboard
│ └── components/ # Shared UI components
├── chrome-extension/ # Browser extension (LinkLens)
├── linklens-whatsapp/ # WhatsApp scanning bot
├── ml_models/ # ML model training scripts
├── grme/ # App screenshots & media
├── start.bat # One-click launcher (Windows)
└── README.md
start.bat- Python 3.9+
- Node.js 18+
- PostgreSQL (optional — app works without DB using mock data)
cd ml_models
python generate_models.pycd backend
pip install -r requirements.txt
python -m uvicorn main:app --reload --port 8000cd frontend
npm install
npm run dev- Frontend: http://localhost:3000
- Backend API: http://localhost:8000
- API Docs: http://localhost:8000/api/docs =======
- Visit Google AI Studio
- Create an API key
- Add to
backend/.env
- Visit NewsAPI.org
- Sign up for a free key
- Add to
backend/.env=======
=======
Detect phishing links using hybrid threat intelligence models.
Features:
- Multi-source threat analysis
- Domain reputation scoring
- Malware/phishing detection
- Sandbox link verification
Security Engines Used:
- Safe Browsing
- OpenPhish
- IPQS
- Domain heuristics
Detect AI-generated manipulated media.
Capabilities:
- Image authenticity checks
- AI generated content detection
- Pattern recognition using ML models
Analyze suspicious articles using NLP based misinformation detection.
Features:
- Content credibility scoring
- Semantic analysis
- Misinformation pattern detection
Check whether emails or passwords appear in data breaches.
Capabilities:
- Dark web breach detection
- Password leak analysis
- k-Anonymity based privacy protection
Protect users from UPI and QR code payment scams.
Features:
- Fake UPI detection
- Collect request identification
- QR scam alerts
- Fraud VPA detection
Scan QR codes and analyze embedded links.
Protection Against:
- QR phishing attacks
- Malicious download links
- Payment redirection scams
Detect fake job offers and recruitment scams.
Capabilities:
- LinkedIn job analysis
- Scam pattern recognition
- Employer verification
- Risk scoring
TrapEye includes a browser extension that scans URLs automatically while browsing.
Features:
- Real time phishing detection
- Instant security alerts
- Browser level threat monitoring
- Safe browsing analysis
LinkLens analyzes suspicious links shared via messaging platforms.
Capabilities:
- WhatsApp link scanning
- Fraud detection in forwarded messages
- Reputation analysis
A community powered cybersecurity intelligence system.
Users can:
- Report suspicious links
- Share new scam patterns
- Build collective threat intelligence