Working with LLMs, prompt engineering, agents, and the strange new world of generative AI.
How dailyprog generates a fresh coding puzzle every day: a legally safe corpus, an LLM-built ontology, and a repair loop that runs real code.