A student developer built HAsh_Scanner, a web app that performs real-time phishing detection using machine learning and security checks. It analyzes URLs for suspicious patterns, evaluates domain intelligence, and inspects content for phishing keywords. The system generates a nuanced risk score from 0-100 instead of a simple safe/unsafe label.

Trained on 156,000 real-world phishing URLs, the tool identified brand impersonation and subdomain abuse. Testing showed 0 false positives on 51 legitimate sites and 70% detection on 20 known phishing patterns. Built with Python, Flask, and Gunicorn, it deploys on Render with rate limiting and HTTPS-only communications.

HAsh_Scanner prioritizes explainable results and privacy, storing no URLs or tracking users. The developer, a student, seeks feedback from cybersecurity professionals to improve accuracy and expand capabilities. This reflects growing demand for practical, transparent security tools that balance detection with minimal false positives in real-world scenarios.