Implementasi Artificial Intelligence untuk Analisis Attack Surface dan Rekomendasi Remediasi Kerentanan Keamanan Berbasis Vulnerability Assessment pada Website UMKM
TL;DR
VulnScope, a web-based vulnerability assessment (VA) system that integrates a Large Language Model (LLM) to analyze the attack surface and generate contextual remediation recommendations, shows that integrating VA with artificial intelligence effectively helps MSME owners independently identify and remediate website security vulnerabilities.
Abstract
The rapid digitalization of Micro, Small, and Medium Enterprises (MSMEs) in Indonesia has raised the need for adequate website security, yet limited technical resources leave many MSME owners unaware of their platform's security posture. This research designs and implements VulnScope, a web-based vulnerability assessment (VA) system that integrates a Large Language Model (LLM) to analyze the attack surface and generate contextual remediation recommendations. The system is built on the FastAPI framework with eight core scanning modules, an aggressive scanner, and a VAPT Engine that classifies findings based on OWASP Top 10:2021 and CVSS v3.1. Artificial intelligence is integrated through the Claude and Nous Hermes models using an agentic tool-use pattern. Black box functional testing confirmed that all features work as specified. Applied to the MSME website ibadahterpanjang.com, the system detected four vulnerabilities (one Critical, one Medium, two Low) with an overall CRITICAL risk level (score 38/100). The results show that integrating VA with artificial intelligence effectively helps MSME owners independently identify and remediate website security vulnerabilities.
