September 2026
Intermediate
308 pages
7h 33m
English
In the previous chapter, we used EventStorming and Domain Storytelling to externalize domain knowledge and to stabilize the vocabulary the team will use for the rest of the book. We ended on a working principle: the sequence Discover, Refine, Name, Reinforce produces language stable enough to be trusted across conversations with an AI assistant. Stable vocabulary, however, is not the same as disciplined reasoning. This chapter closes that distance by introducing a rulebook.
A rulebook is a small, explicit set of constraints that govern how the model reasons inside one bounded context. The team writes it once, versions it alongside the code, and reuses it in every session that follows. In practice, ...
Read now
Unlock full access