The components you invoke via python all “live” in a separate document. This document is never loaded into the canvas and is never included in previews. The node-to-code stuff only deals with inputs and outputs, any ancillary functionality is meaningless.
Ah ok, good to know. Ok next question is there a way to preview text within a script component or is my best bet just to make a cluster with a text tag 3D component and my python bake text script component?
In C# you can add preview override methods, perhaps python allows the same? There’s also the Rhino.Display.CustomDisplay class. But you’ll run into issues when a different document is loaded.