Verify Before You Trust.
Advanced real-time phishing detection using 512-bit vector analysis and Machine Learning. Protect your digital identity from sophisticated malicious attacks.
Triple-Tiered Protection
A modern full-stack implementation merging React, Node.js, and Python ML microservices.
Neural Frontend
High-performance dashboard built with React 19 and Vite for instant URL analysis and history tracking.
Secure API
Robust Express.js backend handling JWT authentication, encrypted profile data, and session integrity.
ML Microservice
Python Flask API running Scikit-learn models trained on 11,000+ phishing signatures for 98% accuracy.
Proactive URL Neutralization
SafeWebVerify doesn't just block; it analyzes. Every link is cross-referenced against 30 unique features to identify zero-day threats.
Instant Confidence Scoring
Receive a real-time probability score for every URL, categorizing it as Legitimate, Suspicious, or Phishing.
Dynamic History Audit
Track your entire history of URL checks with secure, per-user encrypted storage and one-tap cleanups.
Smart Whitelisting
High-speed bypass for trusted global domains ensures zero latency for verified institutional websites.
def extract_features(url):
features = [
using_ip, long_url, short_url,
symbol_at, redirect_path, ...
]
return model.predict(features)
// Prediction: Phishing (Confidence 98%)
Complete System Installation
Follow these steps to deploy the neural mesh on your local machine.
1. System Core
Clone the primary repository containing the React frontend and Node.js API.
2. Neural Assets
The ML model is hosted separately due to its high-dimensional weights.
⬇️ Download .PKL Modelcd ml-api
pip install -r requirements.txt
# Important: Place the downloaded 'phishing_model.pkl' here
python app.py
# B. Setup Secure Backend
cd ../server
npm install
# Configure your .env (MONGODB_URI, JWT_SECRET)
npm start
# C. Launch Neural Frontend
cd ../client
npm install
npm run dev
Missing the weights?
If you see a "Model not found" error, ensure you have downloaded the phishing_model.pkl from the link above and placed it inside the /ml-api directory.