PacktPublishing/30-Agents-Every-AI-Engineer-Must-Build — reverse-engineered prompt

Build me a hands on Python notebook project that teaches the main AI agent patterns from beginner foundations through more advanced real world systems. I want it organized chapter by chapter, with clear explanations, working examples, and notebooks I can open and run locally without a lot of setup. Please make it friendly for someone learning agent engineering, but still practical enough to show production minded concerns like memory, tool use, multi step decision making, reliability, cost, safety, and evaluation.

A big requirement is that every chapter should work in a no API simulation mode with built in mock responses, and also have an easy way to switch to live model providers if I add a key later. Support common cloud models plus a local option, and include simple setup docs, env examples, and pre run notebook outputs so I can browse results even before running anything.

Keep the structure clean and consistent across chapters, and add README guidance for choosing between simulation, cloud, and local use. Look up current docs online if you need to.