Contenu du pack

Un fichier SKILL.md (le guide pratique) + un modèle de départ. Aperçu :

Paper-Cut Science Explainer

When to use

• You need to explain one science idea — how vaccines train the immune system, why the Moon changes shape, how a greenhouse traps heat — to a non-expert audience
• A classroom, museum, or YouTube channel wants a handmade, tactile look instead of slick motion graphics
• You have paper, a phone camera, and a table, but no animation budget or software skills
• A science communicator wants a short (60–180 second) video that feels crafted rather than corporate
• You need a repeatable format: same paper-cut style, new topic each episode

Workflow

1. Pick one idea and one "aha" (input: topic → output: one-sentence core claim)
   Reduce the topic to a single claim a viewer can repeat afterwards: "The Moon looks different because we see different amounts of its sunlit half." Everything in the video serves that one sentence. Write it down first, then list the three facts the viewer must accept to believe it — no more than three. If the topic needs a fourth fact, it is two videos, not one. This sentence becomes the video's final line, so it has to survive being said out loud.
2. Write the spoken script (input: core claim + three facts → output: narration script)
   Write 120–250 words of spoken narration, in the order the viewer will see things: everyday hook (10 seconds), the puzzle stated plainly, fact one with its visual, fact two with its visual, fact three with its visual, then the core claim restated. Read it aloud with a timer — narration runs about 150 words per minute, so a 90-second video gets roughly 220 words. Cut any sentence that needs a diagram to be understood; in this format, the paper on screen is the diagram, and the words must work alongside it, not instead of it.
3. Design the paper scenes (input: script → output: scene list with cut lists)
   Break the script into 6–10 scenes, one per idea beat. For each scene, list the paper pieces to cut: background layer, mid-ground shapes, and one moving piece. Keep the palette to 3–4 paper colors plus black and white, and keep every piece big enough to move with two fingers — tiny fiddly pieces are invisible on camera and slow the shoot. Sketch each scene as a rough rectangle with the pieces placed; the sketch is the shooting plan, not a work of art. Reuse pieces across scenes wherever the story allows (the same paper Sun can shine in three scenes).
4. Build the shot list (input: scene list → output: shot-by-shot plan)
   For each scene, specify: camera position (top-down is the default — paper reads best from directly above), the moves (slide in, lift-off reveal, flip, swap), the number of frames or the hold time, and the narration line it covers. Stop-motion moves in small increments — 8–12 frames per second of finished video is enough for this handmade style, and holding a finished pose for a beat is a feature, not a flaw. Mark any scene that needs a hand in frame (placing or removing a piece) versus hands-free scenes, so the shoot can be scheduled around them.
5. Shoot, assemble, and check (input: shot list + cut pieces → output: finished video)
   Shoot in one consistent light (a desk lamp fixed in place, no daylight that shifts), with the phone locked on a tripod or stack of books so the frame never drifts. Capture narration separately in a quiet room, then assemble: picture first, narration second, music last and low — music should be felt, not heard. Watch the finished cut with the sound off: if the paper alone does not carry the story, a scene needs rework, not more narration. Then watch with sound and check the final line lands exactly on the core claim from step 1.

Aperçu du modèle

Paper-Cut Science Explainer — Template

One idea, one video. Fill this in before cutting any paper.

Core claim

Topic: {e.g. Why the Moon changes shape} One-sentence claim a viewer can repeat: {e.g. The Moon looks different because we see different amounts of its sunlit half.} Three supporting facts (max 3):

1. {fact}
2. {fact}
3. {fact}

Narration script ({60–180} seconds, ~150 words/minute)

Total: {n} words / ~{n} seconds — read aloud with a timer before shooting.

Scene list (6–10 scenes)

Scene {n}: {title} — covers narration: "{line}"

• Background layer: {paper color + shape}
• Mid-ground: {pieces}

Le pack de compétence lui-même (SKILL.md + modèle) est rédigé en anglais.

Acheter le pack complet — $4.99 Pro — les 511 compétences