TY  - RPRT
TI  - SkillSecurer: Detecting and Patching Prompt-Injection Vulnerabilities in AI Agent Skills
AU  - Donato Mecca
AU  - Alberto Verna
AU  - Youness Bouchari
AU  - Nikhil Jha
AU  - Marco Mellia
PY  - 2026
UR  - https://arxiv.org/abs/2609.14079
ID  - 2609.14079
ER  -