pensarai/apex: Autonomous AI Penetration Testing Agent With Argus Benchmark of 60 Vulnerable Apps

GitHub·medium signal

Pensar Apex is an AI-powered penetration testing agent that runs autonomous black-box and white-box pentests directly in your terminal — no source code access required. The real contribution is Argus, an open-source benchmark of 60 self-contained Dockerized vulnerable web apps spanning Node.js/Express, Python/Flask/Django, and multi-service architectures. Argus covers WAF/IDS evasion, multi-step exploit chains, multi-tenant isolation failures, and modern auth bypasses.