I’d love to have more visibility into token usage for a given call. For a give call, I can see the number of tokens user for each LLM request, but I don’t know what’s going into that request. I have an agent that doesn’t have a knowledge base and the initial LLM request varies greatly in the token usage for the first request (around 3k in some cases). For most calls, the initial request is around 3300 tokens, but I have some examples where the initial request is over 6000 tokens. I’d love to be able to see what’s sent to the LLM for each request, or at least some visibility so I can optimize.