Interesting, have you considered adding benchmarks comparing AgentKit to other frameworks? Would help teams evaluating options