I am one of the co-authors of the original AgentDojo benchmark done at ETH. Agent security is indeed a very hard problem, but we have found it quite promising t

I am one of the co-authors of the original AgentDojo benchmark done at ETH. Agent security is indeed a very hard problem, but we have found it quite promising to apply formal methods like static analysis to agents and their runtime state[1], rather than just scanning for jailbreaks. [1] https://github.com/invariantlabs-ai/invariant?tab=readme-ov-file#analyzer https://github.com/invariantlabs-ai/invariant?tab=readme-ov-...