About
I’m Aniket. I go by Awwfensive online. I break software, study how AI systems fail, and build tools for vulnerability research.
My research sits across three areas:
- AI for Security: autonomous vulnerability discovery, offensive agents, harness engineering, and evaluation through Hackbot.
- Security of AI: prompt injection, tool abuse, connector authorization, agent permissions, and agent isolation.
- Vulnerability Research: IDOR, BFLA, access control, tenant isolation, business logic, and API security.
I work independently. This site brings together published disclosures, technical analyses, and notes. Experiments and future research questions are described as work in progress.
Experience
Independent research · Now
Open-source vulnerability research, responsible disclosure, and security tooling. My Flyto disclosure is tracked as CVE-2026-77314.
Penetration Testing Intern · we45 · Jan–May 2026
Tested client APIs and web applications, from reconnaissance through written reports.
Security Intern · Dark Relay Security Labs · May–Dec 2025
Built deliberately vulnerable applications and hands-on security labs.