Research

The Xtrinel research log.

Writeups, detections, and post-mortems from the Xtrinel team. Everything we publish is paired with a reproducible repro script and an entry in the VAAST vulnerability catalog.

April 14, 20268 min readVAAST
Vulnerability Assessment for AI Security Testing: Mapping the Attack Surface of LLM-Integrated Applications
Why the traditional DAST playbook misses most of the AI attack surface, and how VAAST closes the gap across prompt injection, tool-call abuse, RAG poisoning, and agentic pipelines.
Read post →

The Xtrinel research log.

Vulnerability Assessment for AI Security Testing: Mapping the Attack Surface of LLM-Integrated Applications