Prompt Engineering in Practice cover
welcome to this free extract from
an online version of the Manning book.
to read more
or
welcome

welcome

 

Dear Reader,

A prompt that worked on Friday quietly stops working on Monday. The diff is one word. There is no obvious place to look. If you have shipped a feature on top of ChatGPT, Claude, Copilot, or any other Language Model, you have lived some version of this, and you have probably fixed it by tweaking the wording until the next run looked good enough. We have both spent enough time fixing prompts at three in the morning to be convinced that prompt engineering needs the same engineering rigour as the rest of the stack. That is the book you are holding.

Thank you for picking up the MEAP for Prompt Engineering in Practice. This book treats prompt engineering as a design discipline in its own right: a formal system built from the Structural Elements of a prompt and a curated library of Prompt Patterns, the prompting equivalent of software design patterns. The companion volume, AI Engineering in Practice, will do the same for the wider stack: Language Models, agents, harnesses, context engineering, and the rest of modern AI engineering.