Deep into prompt engineering for LLM fine-tuning. Building custom agents and sharing the code that makes them tick. x.com/Taffnykiki