The five-slot video prompt checklist

If the same prompt keeps producing a different video every time, the prompt is missing information. Each missing field becomes a field the model fills in for itself. Paste a prompt below and the checker reports which of the five slots it can find evidence for.

Check a prompt

Paste your prompt

What this check can and cannot tell you

It matches words, it does not read meaning. The checker looks for vocabulary that usually belongs to each slot and reports what it found, listing the matched words so you can judge them yourself. It is a prompt for your attention, not a measurement of your prompt.

Two consequences worth knowing before you trust a result.

A slot can be written correctly and still be reported missing. If you described the camera as "the frame creeps closer", no camera vocabulary appears, and the checker will call the slot empty. False negatives are expected, not exceptional.
Finding vocabulary is not the same as writing the slot well. "Close-up" matches the composition slot whether or not that framing suits your shot. Nothing here checks whether the words are the right words.

We are not putting a score on it. We have not run a controlled comparison of structured against unstructured prompts, so any percentage would be invented. The checklist further down covers the things a word list cannot judge at all.

Why prompts feel like a lottery

Language models omit parts of a shot systematically. Lin, Mitra et al. (CVPR 2026) audited eight video-captioning datasets published between 2016 and 2025 and found the same pattern across them: captions describe the subject and the scene, and drop what the subject isdoing, how the frame is composed, and how the camera moves. The authors also tried hiring crowdworkers to write more careful captions, and the vocabulary still did not appear. Workers continued to confuse a dolly-in with a zoom-in, and to describe a fisheye lens as a round building.

That study is about captioning and supervision data. Reading it across to prompt writing is our inference, not its result. On our own shots the inference holds: every dimension you leave unspecified is a dimension the model chooses for you, and that is the main reason the same prompt can come back different each time.

It is not the whole story. Generation is stochastic, so specifying every slot does not remove run-to-run variation, and even a fully specified prompt can be partly ignored.Structure reduces ambiguity; it does not make the output deterministic.

The five slots

Slot	What it controls	What to write
1. Subject	Who or what the frame is about	Age, build, clothing, skin tone for people; colour and material for products. With multiple subjects, state which is primary.
2. Subject motion	What the subject does, in time order	"First she does X, then Y, then Z." Interaction between subjects counts here.
3. Scene	Where it happens and under what light	Location, time of day, and lighting (natural window light, side light, neon). Add environmental motion such as passing crowds or traffic.
4. Spatial / composition	How things are arranged in frame	Shot size (close-up, medium, wide), subject position (centre, off to the right), foreground/midground/background, and how the layout changes over the shot.This is the slot most often missing, and the usual cause of strange framing.
5. Camera	How the camera moves	Height, angle, focus behaviour, stability, and the movement itself (push in, pan, tilt). "Slow push from 1m to 40cm, steady handheld feel, focus locked on the face."

Slots 4 and 5 are not the same thing. Slot 4 is how the subject sits in the frame; slot 5 is how the camera behaves. Writing "camera moves closer" in slot 4 and leaving slot 5 empty still leaves the camera unspecified.

Before and after

Weak: covers roughly two of the five slots.

A woman uses a facial mist, her skin becomes dewy, the background is an office.

Composition, camera and motion order are all missing, so the model chooses them. This is where wrong-looking faces, odd framing and inexplicable camera moves come from.

Strong: all five slots filled.

Subject: woman in her early 30s, black business attire, dry flaky skin, holding an unbranded white mist bottle.
Motion: first turns a dry cheek toward the camera, then closes her eyes and sprays toward her face, mist fills the air, she pats her cheek and smiles.
Scene: modern office desk by day, natural light through a window, laptop and papers on the desk, blurred colleagues moving behind.
Spatial: opens on a medium shot with the subject slightly right of centre; pushes in to a facial close-up as she sprays, mist particles in the foreground.
Camera: slow continuous push from about 1m to 40cm, steady handheld feel, focus locked on her face.

Self-test while writing

Cover the other four slots and read only one. Can you picture the shot from that slot alone? If not, that slot is not written yet.

Pre-submit checklist

These are the ones a word list cannot judge. Work through them by hand before you submit.

Subject: appearance, skin tone and clothing all stated? Primary subject clear?
Motion: does it have a time order (first X, then Y)? Any interaction described?
Scene: location and lighting both given? Environmental motion included?
Spatial: shot size, subject position, foreground/background, and how the layout changes? Check this one twice.
Camera: height, angle, movement and focus specified? Clearly separate from slot 4?
Consistency: does it carry over subject, scene and lighting from the previous shot?
On-screen text: is the frame kept free of text, UI and screen recordings? Generate text as an overlay afterwards instead, because models render lettering unreliably.
Length: is the prompt inside the comfortable range shown by the checker above?
Deliberate variation: when running a batch for variety, which single slot are you intentionally leaving open? Vary one, not all five.

Reducing variance further

A text prompt leaves all five slots to be interpreted. A start-frame image locks three of them, subject, scene and composition, immediately. That leaves only motion and camera for the prompt to control, which is a large drop in variance. If hit rate matters more than variety, lock with an image first and write a prompt for the remaining two slots.

The checker is a word-list heuristic, not a model. It is deliberately conservative about what it claims, and its limits are described above.