Comprehensive ML-Driven URL Scanner. This repository contains all essential files related to the MUJ HACKX Hackathon.
- Achieved an accuracy of 97.2% by evaluating an ensemble of machine learning models, with a focus on using an XGBoost Classifier following 10-fold validation for URL fuzzing.
- Conducted HTTP GET requests and analyzed returned statuses to distinguish live from non-live URLs.
- Employed a brute-force method to address 404 errors, incorporating a hard-coded list for testing.
- Brute-forced default credentials, cracking common login combinations using a range of feature arrays.
- Accelerated the URL fuzzing process via a straightforward Streamlit deployment.