Expand a half-written note, shorten a rambling one, or make a description read like the rest of your guides.
Only the text you selected. No screenshot leaves your computer for these actions, which is the difference between them and AI step descriptions.
That also means a text-only model works fine here. If you are paying for a vision model to generate descriptions, the writing actions do not need one. Setup is the same provider and key: see Set Up AI in Folge.
Left alone, the AI writes in its own voice, which is wordier than most documentation should be. Custom instructions in AI settings is added to every one of these actions.
This matters more for Expand than for anything else, because Expand is the action most likely to invent padding you then have to delete.
Folge captures a screenshot on every click, so a guide writes itself as you work.
