CyberSentinel

CyberSentinel is an AI-powered multi-modal cybersecurity system designed for real-time phishing detection across URLs, SMS, and QR codes. It combines a centralized dashboard with a proactive browser extension to provide end-to-end protection.

Phishing attacks have evolved beyond email into multiple channels. CyberSentinel addresses:

• Phishing: Malicious URLs targeting user credentials
• Smishing: SMS-based social engineering attacks
• Quishing: QR codes embedding fraudulent links

The system leverages machine learning and NLP to analyze threats and generate real-time risk assessments.

• Unified Detection Pipeline: Single /predict API for all input types
• XGBoost-based URL Detection: Uses lexical and structural feature analysis
• Smishing Detection: NLP + TF-IDF pipeline for SMS classification
• Frontend QR Decoding: Fast client-side extraction using jsQR
• Real-time Browser Protection: Chrome extension scans links and images
• Dynamic Risk Scoring: Confidence-driven risk meter (no hardcoded values)

git clone https://github.com/Sankethhhhhhh/CyberSentinel.git
cd CyberSentinel

cd backend
pip install -r requirements.txt
python app/main.py

Backend runs at: http://127.0.0.1:8000

cd frontend
npm install
npm run dev

Open: http://localhost:3000

1. Open chrome://extensions/
2. Enable Developer Mode
3. Click Load unpacked
4. Select CyberSentinel_Extension/

Pre-trained models are included, so the project runs out-of-the-box.

If you want to retrain the models:

Train URL Model

python backend/training/train_url_model.py

Train SMS Model

python backend/training/train_sms_model.py

After training, the models will be saved in: backend/models/

Ensure the following files are present:

• backend/models/url_models/best_url_model.pkl
• backend/models/sms_model/best_sms_pipeline.pkl

User Input (URL / SMS / QR)
 ↓
Frontend (React / Chrome Extension)
 ↓
FastAPI Backend (/predict)
 ↓
Inference Module
 ↓
ML Models (XGBoost + NLP)
 ↓
Prediction (label + confidence)
 ↓
Dashboard / Browser Highlighting

Endpoint POST /predict

Request

{
 "input_type": "url",
 "data": "http://suspicious-login.com"
}

Response

{
 "label": "phishing",
 "confidence": 0.942,
 "risk_level": "HIGH"
}

• Unified all detection workflows under /predict
• Fixed confidence calibration (removed constant 100% issue)
• Integrated QR scanning directly into frontend
• Improved SMS pipeline with optimized TF-IDF + Scikit-learn
• Enhanced extension to scan both links and images in real-time

CyberSentinel is currently optimized for local execution. Deployment to cloud infrastructure (AWS/Azure) is in progress and will be available in upcoming updates.

• Explainable AI (model reasoning output)
• Domain reputation integration (VirusTotal / Safe Browsing)
• Batch scanning optimization for large-scale webpages
• Cloud deployment and scalability improvements

This project is licensed under the MIT License and is intended for educational and research purposes.