Ardan Ultimate AI #03 — Context injection into a prompt
Before RAG and tools, the original way to give an LLM extra information was to inject it into the prompt. The example shows the right way to format injected context and what the LLM does (and doesn't) pay attention to.