AI Attack: HalluSquatting Exploits Language Models
HalluSquatting is an attack where researchers pre-compute fake repository, package, or skill names that AI coding agents predictably invent, register those names first, and load them with malicious instructions.