Dami Dina: do you think the LLM should be doing this type of logic or should there be other

@provisionalidea do you think the LLM should be doing this type of logic or should there be other abstraction layers or semantics to help the LLMs do a better + more reliable interpretation of these rules?