BugShield: AI-Based Web Attack Surface Mapper For Automated Security Assessment

Abstract

Web application security is an important concern in modern software development because insecure coding practices and exposed credentials can introduce risks to application confidentiality, integrity, and availability. This paper presents BugShield, a source-code security scanning and findings-management platform designed to support preliminary security assessment of web applications. The system combines project management, automated source-code analysis using implemented detection rules and integrated security tools, vulnerability classification, structured findings, dashboard visualization, AI-assisted explanations, and PDF report generation. The platform aims to help developers and security researchers identify potential weaknesses, review supporting code evidence, prioritize findings, and understand possible remediation measures. The evaluation methodology considers detection outcomes against known test cases, severity classification, scan execution time, and report-generation functionality. Experimental findings are to be reported using verified measurements collected during testing. The system's capabilities depend on supported programming languages, implemented detection rules, and integrated tools. Potential false positives and false negatives require further verification, and source-code scanning alone cannot identify every runtime or business-logic vulnerability. BugShield is intended to complement established security testing practices by providing an integrated workflow for preliminary vulnerability analysis and findings management.

Keywords

• Web Application Security
• Vulnerability Detection
• Static Analysis
• Attack Surface Analysis
• Secure Coding
• AI-Assisted Security