The Prompt Structure Converter

Convert this video concept into a properly structured prompt.

My concept, as I would describe it to a person:
[DESCRIBE IN PLAIN LANGUAGE]

Platform: [SORA / VEO / RUNWAY / KLING / LUMA / OTHER]
Duration: [SECONDS]
Aspect ratio: [SPECIFY]

Produce:
1. The structured prompt using: subject and action / camera / lens
   and light / tone / exclusions
2. What I left unspecified that the model will decide for me, and
   what it will probably decide
3. The single element most likely to be misinterpreted
4. A tighter version at half the length — tell me what you cut and
   what that costs
5. If my platform uses tags rather than prose, the tag version

Rules:
- One clear action. If my concept has several, tell me which to
  keep and why the others should be separate shots.
- Camera must be specified even if static
- Exclusions must cover the things video models add uninvited:
  people, hands entering frame, text, unmotivated camera drift