#backdoors
2 briefs tagged #backdoors.
Research
Researchers flag backdoor risks in LLM-agent skills
New papers probe how reusable agent skills can be poisoned, compressed and evolved.
arXiv cs.AI+1 outlet·Aug 11
Research
Researchers detail SkillJack backdoor attack on agents
The attack targets self-evolving agents that turn past interactions into reusable skills.
HF Daily Papers·Aug 4