Guide · Workflow
A preset is a prompt that already works, packaged so anyone on the team can start from it. A good library is the difference between AI as a toy and AI as a production system. Here is how to build one that stays useful.
Do not sit down to write presets. Make work, and every time an output earns a save, turn it into a preset on the spot. A library built from proven outputs stays trusted. A library built from theory gets ignored within a week.
Product on marble, morning light is a preset. Nice product shots is not. Narrow presets compose: your team can stack a scene preset with a light preset and a format preset. Broad presets collapse into vague output nobody can predict.
Use a locked naming convention: subject, scene, light, format. When names follow the same grammar, search works, duplicates are obvious, and a new hire can navigate the library on day one without a walkthrough.
Once a month, sort by usage. Presets nobody touched get archived, presets everyone uses get promoted and refined. A library of forty presets people trust beats four hundred they scroll past. In Plenty Labs the whole team shares one library, so the pruning pays off for everyone at once.
