Structured Output and Tool Schemas (the contract between model and code)

Building with LLMs. The board after “Context Engineering” (Context Engineering (what goes in the window, and who decides)) decided what goes into the string the model reads. This one decides what happens to the string the model writes back.

Before you start. A model writes one token at a time, a token being a word or a piece of one, the unit it writes in. Everything it has read so far - instructions, conversation, and now a description of what it may ask for - sits in front of it as one long stretch of text, the prompt. This lesson follows what happens to the text it hands back.