Turn a rough clip idea into three generator-ready prompts

I have a rough idea for an AI-generated video clip. Rewrite it as three ready-to-paste prompts for a text-to-video model, each a different interpretation of the same idea.

My idea: [DESCRIBE THE CLIP IN YOUR OWN WORDS, HOWEVER VAGUE].
Intended use: [WHERE THE CLIP WILL APPEAR, E.G. AD, TITLE SEQUENCE, SOCIAL POST, MOOD PIECE]. Aspect ratio [16:9 / 9:16 / 1:1], length [SECONDS] seconds.

Each rewritten prompt must be a single paragraph of 60-120 words containing, in this order: shot size and framing; subject with two or three concrete visual identifiers; the single action taking place; setting with time of day; one camera movement with speed; lighting source and quality; color palette or film-stock reference; what moves in the frame and what stays still; audio (ambient bed, effects, or none).

Make the three versions differ in a meaningful way — for example one naturalistic handheld documentary take, one controlled take with slow deliberate camera movement, one stylized or abstract take — and label each with a two-word name. Do not use vague adjectives like "beautiful," "epic," "stunning," or "cinematic" on their own; every descriptor must be something a camera, a light, or a subject can physically do.

After the three prompts, list the three most significant details from my idea that you had to invent or assume, so I can correct them and ask for a revision.