This module writes stack/orchestration.md, which 3 later modules read. It is part of the paid course.
You split one real job across two agents, put the handoff between them in a file you can open, then run the same job as a single agent and say which won and on what measure. You learn the three shapes multi-agent work takes, the three reasons that justify their cost, and the tax you pay in latency, money and debugging. You leave with a page recording which shape you use, why, and what happens when a worker is wrong.
Twelve modules, thirteen files. From typing into a box to a working setup you own — no technical background assumed.
Build AI that acts. Sixteen modules, fourteen files, two tracks, and the rules written down before anything runs.
Twenty-eight modules end to end, in the order they were meant to be read.
After you pay: a sign-in link lands in that inbox and the first paid module opens. Usually under a minute.