AI-Powered Web Vulnerability Scanner

Develop a machine learning-based web vulnerability scanner "Cult Secure" that analyzes traffic, identifies risks, suggests real-time fixes, and generates user-friendly reports to enhance penetration testing speed and accuracy.

Features

• Header Analysis: Analyzes HTTP headers for security misconfigurations.
• Vulnerability Detection: Detects CSRF issues, XSS vulnerabilities, and other common web vulnerabilities.
• Lighthouse Integration: Runs Google Lighthouse for a complete performance and security report.
• Scheduled Scans: Allows automated, scheduled vulnerability scans.
• Machine Learning-based Classification: Classifies cybersecurity rules using a pre-trained machine learning model.
• PDF Report Generation: Generates PDF reports summarizing scan results.
• Simulates User Interactions: Uses Selenium to automate web page interactions.

Implementation

Tech Stack

• Backend Framework: Flask
• Web Scraping: BeautifulSoup, Requests
• Automated Browser Interaction: Selenium (with ChromeDriver)
• Machine Learning: Scikit-learn, Hugging Face Datasets
• Scheduling: Python schedule library
• Logging: Python logging module
• Performance & Security Audits: Google Lighthouse
• Data Handling: Pandas
• Model Training: TfidfVectorizer, LogisticRegression

Dataset

The machine learning component uses the Cybersecurity Rules Dataset from Hugging Face, which contains labeled cybersecurity rules and questions to train the classification model.

Dataset Link: Cybersecurity Rules Dataset

Web App Interface