How the engine reads a section
Three questions. One layout. You never choose the template.
Every section resolves through three questions. What does the lead item carry: a photograph, an icon, or text alone? How many supporting items follow? And what do those items carry?
The combination of those three answers points to exactly one layout. The engine applies it. You never open a template picker, never adjust a grid, never wonder whether two columns or three looks better.
A photograph, an icon, or text. This determines how the section opens: as a visual statement, a symbolic marker, or a headline. The lead item sets the register for everything that follows.
One, two, three, or several. This determines the grid. A single supporting item gets featured treatment. Three get a row. Several collapse into a structured list. The count is the layout instruction.
Photographs, icons, or text, evaluated across the group. If any item in the group carries a photograph, the group is treated as visual. The most expressive element in the set determines how all of them render.