RedEvoAgent: Automatic Red-Teaming Agent with Experience-Driven Skill Evolution
LLM-based agents are increasingly deployed in product-level execution harnesses, where jailbreaks can trigger harmful tool use and persistent state changes, creating greater risks than unsafe text generation alone. Existing automatic red-teaming methods often rely on fixed attacks, while recent agentic attackers coordinate multiple jailbreak tools and show stronger potential through…
We haven't written up this one. arXiv cs.AI has the full story — the link below goes straight to it.
