Skillsby SefarAI
Back to Learn
Video tutorial

Make a viral What's the Difference? Short

Use a consistent stickman host to turn two similar ideas into a clear, fast, and memorable vertical video.

Source checked 15 August 2026

Install the required skill

npx skills add SalimLouDev/whats-the-difference-shorts-skill --skill whats-the-difference-short
View What's the Difference Short

What you will produce

A complete 1080x1920 What's the Difference? Short with an original stickman host, researched contrasts, synchronized narration, captions, publishing copy, and a source record.

Before you begin

Node.js 18 or later, npm, and Git

Python 3.10-3.12 for Kokoro narration

One or two pairs of similar concepts to compare

An image-generation tool for creating the original stickman scenes

Choose the comparison and hook

Start with two ideas people often confuse, then establish the stickman host and the signature What's the Difference? opening.

  1. 1Install the linked skill using the catalog command above.
  2. 2Choose one focused pair that can be explained quickly. Add a second pair only when both distinctions can remain clear.
  3. 3Open immediately with: This is X. This is Y. What's the difference? Skip logos, greetings, and filler introductions.
  4. 4Ask for one original stickman host with the same head shape, body proportions, line weight, face, and wardrobe accent in every scene.
  5. 5State the language, voice, duration, caption, and visual preferences that should override the documented defaults.

Starter prompt

$whats-the-difference-short Create a viral-style Short about empathy vs. sympathy. Use one consistent original stickman host who points to each idea, reacts to the contrast, and presents every complete 9:16 scene. Open with: This is empathy. This is sympathy. What's the difference?

Prepare the topic package

Start from a compatible Remotion project, verify the environment, and scaffold one folder for the complete production.

  1. 1Inspect the current directory for an existing Remotion project before creating another one.
  2. 2If the directory is empty, bootstrap the bundled starter and install its dependencies.
  3. 3Run the project doctor and resolve required failures before producing assets.
  4. 4Scaffold the comparison so narration, sources, images, audio, configuration, publishing copy, and output stay under one video folder.

Initialize an empty project

python <skill-dir>/scripts/bootstrap_project.py --target .
npm install
python scripts/doctor.py --project .

Scaffold the comparison

python scripts/scaffold_video.py --project . --pair "Empathy|Sympathy"

Design the stickman scenes

Turn the hook and explanation into complete vertical frames led by one recognizable stickman host.

  1. 1Verify both definitions with reliable sources and write the contrast so the narration still works without the visuals.
  2. 2Storyboard the host introducing X, introducing Y, asking the question, and acting out the clearest contrast.
  3. 3Generate each scene as a finished 1080x1920 image. Do not construct the stickman, labels, or background from React layout fragments.
  4. 4Keep the stickman design consistent while changing poses, expressions, props, and side-by-side situations to support each narration beat.
  5. 5Keep essential text and the presenter's face away from Shorts interface zones.
  6. 6Record every downloaded media source and its usage notes in the topic source file.
  7. 7Write one narration chunk per scene, generate the Kokoro voiceover, and synchronize durations from its timing metadata.

Generate narration

.venv/bin/python scripts/generate_voiceover.py --input public/videos/<video-id>/narration.txt --output public/videos/<video-id>/audio/voiceover.wav --voice af_bella --lang a --speed 1.08 --pause-ms 90

Synchronize the timeline

python scripts/sync_timeline.py --project . --video-id <video-id>

Validate and prepare publishing

Inspect the hook, character continuity, pacing, and final publishing package before rendering the Short.

  1. 1Synchronize the registry, validate the selected video, run lint, and inspect the project in preview.
  2. 2Confirm the stickman remains recognizable, scene timing follows the narration, and no audio overlaps unexpectedly.
  3. 3Render the MP4 only when a final render is requested.
  4. 4Prepare a searchable title that names both ideas and ends with or implies What's the Difference? Include any caveat the topic requires.

Validate and preview

python scripts/sync_registry.py --project .
python scripts/validate_video.py --project . --video-id <video-id>
npm run lint
npm run dev

Optional final render

npx remotion render <video-id> public/videos/<video-id>/output/<video-id>.mp4

Verify the result

Every scene is a complete 1080x1920 image rather than a layout assembled in Remotion.
The same original stickman host is recognizable in every scene.
The Short opens directly with This is X. This is Y. What's the difference?
The narration remains understandable without the images.
Scene durations follow the generated voice timing.
Essential content avoids Shorts interface collision zones.
The topic folder contains content, sources, images, audio, configuration, publishing copy, and output locations.
The validated preview was inspected before handoff or rendering.

Keep in mind

Remotion sequences finished scene images. It does not reconstruct visual scenes from layout fragments.
A viral-style format can improve clarity and familiarity, but it cannot guarantee views or audience growth.
Third-party movie, anime, celebrity, and brand images are not bundled. You are responsible for media rights and platform policies.