Discovery
We start with questions, not proposals. What problem are you solving? Who uses this? What already exists? What does success look like in six months?
OutputA written summary of scope, constraints, and what we understood — so misunderstandings surface now, not later.