My team and I have worked on an agentic benchmark to see how well agents can perform at reverse engineering threats. The results are pretty interesting, we would love the communities feedback.