Question
I would like to trace mcp request/response json parsing/generating on server (and eventually on client, but only on server for the moment). Is there any such tracing of pydantic-generated model dumps at the request/response level (so that not all model dumps are included)?
Additional Context
No response