Use case: Wellness creator content, studio marketing, video-model motion tests, and calm lifestyle campaigns.
Create a calm handheld vertical video of a short yoga vinyasa flow.
Sequence:
1. seated breath moment on the mat
2. slow transition into tabletop
3. controlled downward dog
4. step forward into low lunge
5. gentle rise into warrior pose
6. final relaxed standing breath
Visual direction:
- warm natural light
- quiet studio or home environment
- stable handheld camera with subtle human movement
- realistic body alignment and transitions
- soft fabric texture and mat detail
- no extreme flexibility unless explicitly requested
Rules:
- keep limbs anatomically plausible
- preserve continuity of mat position and room layout
- make the pacing calm enough to follow
- avoid jump cuts that hide broken transitions
What to customize first
yoga level
studio setting
lighting
camera distance
sequence length
mood
How to use this template responsibly
This prompt is meant to be adapted into a brief for a real task, not copied into a model without
context. Start with the use case, then fill in the variables, run the quality checks, and keep the
source signal separate from your final prompt variant.
Decision
Use this page for
Do not skip
Task fit
Wellness creator content, studio marketing, video-model motion tests, and calm lifestyle campaigns.
Confirm the output will be reviewed by a person before reuse.
Variables
yoga level, studio setting, lighting
Replace placeholders with concrete details from your own brief.
Quality bar
Body alignment should remain plausible.
Compare the result against the checklist, not only against taste.
Failure prevention
Extra limbs or impossible poses appear during transitions.
Rewrite the prompt if the first run exposes this failure.
Why this prompt works
The sequence defines movement checkpoints. That makes it easier to evaluate continuity and prevents the output from collapsing into random wellness imagery.
Evaluation workflow
Use this page as a repeatable prompt test, not a one-off prompt dump. Save the exact prompt
version, model name, input references, and output settings before comparing results. Then judge
the output against the checks below so the decision is based on observable behavior instead of
whether the first image, video, page, or workflow looks impressive at a glance.
Run the unchanged template once to establish a baseline for the model and task.
Replace the variables with concrete details from your brief, audience, product, or review case.
Score the result against the first quality check before judging style or novelty.
If the first failure mode appears, rewrite the constraints before increasing generation volume.
Keep the best output and rejection notes together so future prompt changes can be compared fairly.
Rewrite record
Before saving this prompt as a team asset, write down what changed from the template and why. The
useful record is not only the final prompt text; it is the task, variables, model, source signal,
quality checks, failure notes, and rejected outputs that explain why this version is trusted.
Record which variables were changed from the public template.
Note whether the output is for exploration, internal review, or external publication.
Keep the first failed result if it reveals a useful constraint for the next version.
For client or brand work, keep rights, claims, likeness, and policy review separate from visual taste.
Quality checks before using the output
Body alignment should remain plausible.
The camera should feel handheld but not chaotic.
Transitions should be slow enough to inspect.
Common failure modes
Extra limbs or impossible poses appear during transitions.
The mat or room changes position between shots.
The model hides movement errors with too many cuts.
Originality and reuse boundary
The source signal explains why this pattern is worth watching, but the value of this page is the
rewritten structure, variables, quality checks, and failure analysis. Treat the final prompt as your
own working brief only after you have changed the subject, constraints, review criteria, and output
context for your own task.
Do not republish source creator text as if it were your own prompt.
Keep a record of the final prompt variant and the model used.
Use the failure modes to decide whether another model, reference image, or manual edit is needed.
For commercial work, review rights, brand claims, likenesses, and policy-sensitive content before publishing.