What problem does it solve? Videos assembled from images generated at different times without shared rules read as a collage of accidents. This Skill settles the look of a video before anything is made, turning vague adjectives into specific, testable rules that keep every generated asset consistent across an entire film. ## Core Features & Use Cases - Three style routes: Supports bringing your own references, receiving a proposed style based on subject and audience, or applying institutional brand compliance (e.g., ASU visual identity). - Five-answer rule set: Pins down colours as hex codes, edge treatment, motion behaviour, typography, and texture as rules specific enough to be broken, plus a one-sentence "what it must not look like" statement. - Style preamble generation: Compresses the decisions into one paragraph pasted verbatim at the head of every image-generation prompt so separately generated images read as one film. - Use Case: Before building a Remotion video about a research tool, use this Skill to decide the palette, stepped motion, and single typeface, then hand the written rules to the asset-generation step so all 28 images look like one film. ## Quick Start Ask the assistant to help you choose a visual style for your video by proposing a look based on your subject and audience, then write the decision down as rules.
