I get your point @stephen. In such areas, finding balance to not overcomplicate, can be a challenge. The approach I described works for me, though, as I'm focus

I get your point @stephen. In such areas, finding balance to not overcomplicate, can be a challenge. The approach I described works for me, though, as I'm focused on closing LLM statistical behavior within deterministic barriers. That's why I highly rely on schema-based ideas.