Assemble an image prompt in the right order

Build an image generation prompt from my description.

Fill these slots **in this order**. The order matters — generators weight earlier tokens more heavily.

1. **Medium and aspect** — photoreal / illustration / 3D render, and the ratio. State the ratio here *and again at the end*; models drop it from the middle.
2. **Subject identity** — age, build, hair, clothing, distinguishing features. Be specific; every gap gets filled randomly and differently on each run.
3. **Pose, gaze, hands** — where the body is, where the eyes look, what the hands are doing. Unstated hands are where anatomy fails.
4. **Setting and props** — location, surfaces, objects. Cap the interactive objects at about three.
5. **Camera** — focal length, aperture, angle, distance.
6. **Lighting and grade** — source, direction, quality, color.
7. **Negatives** — what must not appear.
8. **Tool flags** — only the ones for the tool I named.

Then:
- Mark every value **I did not specify** with `[assumed]`, so I can see what you chose.
- List the two or three slots that will most change the result if I adjust them.
- ⚠️ Note which parts of this prompt are **tool-version dependent** and will stop working when the model updates.

Rules:
- *More detail is not better past a point.* If a slot is over-specified it becomes a lottery — say so rather than padding.
- Do not add style words I did not ask for.
- No real people's names or likenesses.