Project Overview
RedEye Guard is a Chrome-based phishing detection extension designed to identify, analyze, and classify potentially malicious websites in real time using a machine learning-based detection engine.
The system evaluates URLs through structured feature extraction, statistical scoring, and learned model weights to determine the likelihood of phishing or malicious activity. It operates fully within the browser environment, ensuring fast and lightweight execution without dependency on external cloud services.
The extension supports an advanced learning mode where detection accuracy can be improved using external model training. In this mode, a Python-based training pipeline can be executed in Windows or Linux environments to process datasets, update model parameters, and enhance classification performance over time.
This architecture enables the system to evolve into an adaptive, continuously improving machine learning security model.
Key Features
Real-time phishing and malicious URL detection Machine learning-based detection engine URL feature extraction and entropy analysis Brand impersonation and threat pattern recognition Lightweight browser-native execution (no backend required) Chrome toolbar integration for instant access Whitelist and blacklist domain management Optional external Python-based training pipeline Cross-platform support (Linux,Mac,Windows environments)
Installation Guide
- Download the Project
Clone the repository using Git:
git clone https://github.com/JANIW18/RedEyeGuard.git
Alternatively, download the project as a ZIP file and extract it to a local directory.
- Open Chrome Extensions Manager
Launch Google Chrome and navigate to:
chrome://extensions/ 3. Enable Developer Mode
Activate Developer Mode using the toggle switch located in the top-right corner of the page.
This is required to load unpacked extensions manually.
- Load the Extension
Click on Load unpacked and select the extracted project folder:
RedEyeGuard/
Ensure the folder contains the manifest.json file.
- Activation
Once loaded successfully, the extension icon will appear in the Chrome toolbar.
The system will automatically begin analyzing visited websites in real time.
Users can click the extension icon to view detection results and risk scores.
Advanced Training Mode
RedEye Guard supports external model training for improved detection accuracy.
In advanced mode:
Python scripts are used to process phishing datasets
Model parameters are updated based on labeled training data
Detection accuracy improves through iterative learning
Trained outputs can be reintegrated into the extension
Supported Environments
Windows (Python 3.8+) Linux (Ubuntu/Debian-based distributions recommended) Macos
System Architecture
The extension uses a machine learning pipeline based detection flow:
URL Feature Extraction Layer Machine Learning Prediction Engine Threat Pattern Analysis Layer Risk Scoring System
The final output is a normalized risk score between 0 and 1, representing the probability of malicious activity.
Notes:
This project is intended for cybersecurity research, educational purposes, and experimental machine learning development. It demonstrates how browser-based extensions can integrate lightweight machine learning models for real-time phishing detection.