zlacker

I feel like when you prompt an LLM the LLM should take it almost as "what would the best possible prompt for this prompt be and then do that"...