# Create Cross Medium Reaction

> Generate a cross-medium comedy short — live-action night-market stir-fry stall footage and one perfectly flat 2D sticker character (the sticker Shiba Inu 旺财) sharing the frame and reacting to each other — as a 10s 16:9 24fps single continuous POV shot from the vendor's slightly-downward wok-tossing eye level. Use when the user wants a flat 2D sticker character interacting with real footage, a "sticker steals the chili / gets bonked / gets spicy-fired" gag, a cross-medium mixed-media clip, or a first-person wok-cooking comedy. T2VA (text-to-video) only, no reference image. Hard rules: only one live-action hand may appear at a time and never two hands in the same frame; the sticker dog stays perfectly flat 2D with no 3D volume, no realistic hair, shadow or clay texture; left hand never touches the spatula, right hand never holds the jar; no dialogue.

- Skill: `gordonlu/create-cross-medium-reaction` (Agent Skill, multi-file: 5 files)
- Install (CLI): `npx skillmds@latest add gordonlu/create-cross-medium-reaction`
- Raw SKILL.md: https://api.skillmd.com/api/skills/gordonlu/create-cross-medium-reaction/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: gordonlu (https://skillmd.com/u/gordonlu)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/gordonlu/create-cross-medium-reaction

---


# Create Cross-Medium Reaction — Sticker Dog vs. Real Night Market

Use this Skill when the user wants a **comedy clip where a completely flat 2D sticker character shares a real live-action frame and reacts to it**: a flat sticker Shiba Inu sneaks chili powder into a real stir-fry pan, gets bonked on the head by the real spatula, gets fed, and gets comically spicy-fired off its stool.

Core principle: **one real hand at a time, a perfectly flat sticker dog, and every reaction reads as a clean cross-medium gag — no 3D volume, no realism in the sticker, no continuity breaks.**

The prompt follows the H3 timecoded structure (core concept → five-segment screen description with positive and negative per beat → hard negative list). T2VA only — pure text, no reference image.

---

## 1. Trigger

Use this Skill when the user's request matches any of:

- A flat 2D sticker / paper-cut character interacting with real live-action footage
- A "sticker steals the chili / gets bonked on the head / gets spicy-fired" reaction gag
- A cross-medium mixed-media comedy short
- A first-person POV of a night-market wok-tossing cook with a mischievous 2D sidekick

Do **not** use this Skill when:

- The user wants a fully 3D or CGI character blended into the scene — this sticker must stay visibly flat
- The user wants a stable multi-camera setup — this is a single fixed wide POV with slight handheld drift
- The user wants dialogue or voice-over — the whole clip is silent
- A reference image must drive the look (this is pure T2VA)

## 2. Inputs

- A text description of the scene only: a real night-market stir-fry (炒粉) stall, black iron wok with glossy rice noodles and green veggies sizzling and steaming, a plastic stool to the right, and a flat 2D sticker Shiba Inu (旺财) sitting on it. The vendor's first-person slightly-downward wok-tossing eye level, fixed wide composition, only slight natural handheld drift.
- Duration 10s, aspect 16:9, 24fps, one continuous take with no cuts, no transitions, no zoom, no major re-composition.
- No reference image, no first frame — T2VA.

## 3. Core Concept

10 seconds, 16:9, comedy short. Live-action night-market stir-fry footage and one completely flat 2D sticker character (the sticker Shiba Inu 旺财) share the frame and interact. The whole film is the vendor's slightly-downward first-person wok-tossing POV: fixed wide composition, only slight natural handheld drift, no cuts, no transitions, no zoom, no major re-composition — one continuous take. The live-action right hand stir-fries, the live-action left hand carries the chili jar away; 旺财 steals chili powder and gets bonked lightly on the head by the spatula, then gets so spicy it spits a little flame and flips off the plastic stool like a sheet of paper.

## 4. Screen Description (one continuous take)

| Time | Beat | What must happen |
|---|---|---|
| 0–3s | Chili theft | Fixed top-down POV: the black wok sits slightly left-of-center, glossy noodles and green veggies sizzling with natural steam, a plastic stool on the right. The live-action right hand (back of hand facing camera, thumb curled inward toward the left of frame) enters from the bottom-right edge holding the only spatula — a wooden-handled metal one — and keeps stirring. The sticker dog 旺财 sits on the stool with a mischievous grin, lifts a large enamel chili jar with its two clearly flat sticker paws, tilts it over the wok, and pours dry red chili powder in a stream that piles into a visible red mound on the noodles. The spatula never touches the jar; no live-action fingers near the jar. Keep the food sizzle and a dry papery pouring sound. No two live-action hands in any single frame; no second spatula; no mirrored, floating or extra hands. |
| 3–3.8s | Handover | The chili stream stops; the red mound stays visible. The right hand lays the same spatula flat into the wok and fully exits the bottom-right edge — a brief moment with no live-action hand and the spatula lying still in the wok. Then the empty live-action left hand enters from the top-left edge (back of hand facing camera, thumb curled inward toward the right of frame), takes the enamel jar directly from 旺财's two sticker paws, and exits the top-left edge carrying it away — the left hand never touches the spatula. Right and left hands never share a frame; no two left hands, no duplicate arms; the left hand touches no utensils. |
| 3.8–5s | Bonk | After the left hand is fully gone, the same right hand returns from the bottom-right (same orientation), picks up the same spatula from the wok, and lifts it to 旺财 along one continuous path: a light tap of the spatula's flat face on its head — a harmless cartoon bonk with a light metallic "DUANG" plus a cartoon spring sound. A flat red cartoon bump pops up on 旺财's head, its paper body bounces vertically once, eyes widen, both sticker paws clutch its head. The right hand then returns the spatula to the wok; the left hand stays absent. No forcefulness, no realistic pain like injury or burning. |
| 5–8s | Feeding | The same right hand and the same spatula stay visible. 旺财's eyes become flat spiral cartoon eyes and two bright blue 2D sticker tear-fountains spray from the sides of its eyes. The right hand scoops a small bite off the chili pile with the spatula, brings it to 旺财's open cartoon mouth, and the food pops gently into its mouth with a "boop". Its cheeks instantly balloon into two warm red-orange round sticker balloons; the crying is exaggerated and cartoonish, the spatula scraping softly through the chili powder. No realistic gagging, coughing or burning pain; no second hand. |
| 8–9.7s | Spicy flip | 旺财 swallows; its flat body flashes bright red-orange and freezes for half a beat. Its eyes become two flat spiral heat-rings; its open little mouth spits a small puff of flat 2D sticker flame (like a mini dragon breath); two wisps of flat white steam rise from its ears. Then it flips backward off the plastic stool like a light sheet of paper — short limbs up, one sticker paw still fanning its mouth — a ring of flat cartoon red chilies spinning around the bump on its head, a wisp of flat orange steam from its nose. A short "hiss", a light papery "puff" landing sound and a comical descending whistle. Throughout the flip: fur, collar, clothing, sticker white border and round belly stay consistent. No 3D volume in the flip, no realistic falling. |
| 9.7–10s | Finale | Hold clearly on the final visual punchline — 旺财 lying on its back, spiral eyes, a flicker of flame — to close the film. |

## 5. Hard Rules

- **One live-action hand at a time** — exactly one live-action hand visible, and it must never share a frame with another hand.
- **Hand duty split** — the left hand never stir-fries and the right hand never holds the jar; the left hand never touches the spatula; no live-action hand helps 旺财 pour chili.
- **Perfectly flat sticker** — 旺财 must stay a completely flat 2D sticker: no 3D volume, no realistic fur, no volumetric light, no plastic depth, no clay texture, no realistic shadow; only mild flat-cartoon squash-and-stretch. No text, numbers, logos, jewelry or extra accessories on the character.
- **One continuous take** — fixed wide composition, only slight handheld drift, no cuts, no transitions, no zoom, no major re-composition.
- **Silent clip** — no dialogue anywhere.

## 6. Negative Rules (Do Not)

3D volume or realistic fur/shadow/clay texture on the sticker; two live-action hands in any frame; two left hands or two right hands; duplicate arms; extra palms; a second spatula; a continuity break of one spatula in the wok and another in the air; the left hand stir-frying; the right hand holding the jar; the left hand touching the spatula; a live-action hand helping 旺财 pour chili; text/numbers/logos/jewelry/extra accessories on the sticker; realistic pain (gagging, coughing, burning); dialogue.

## 7. Generation

- Mode: T2VA (pure text-to-video) — no reference image, no first frame.
- 10s, 16:9, 24fps, native H3 audio (diegetic food sizzle, papery pours, cartoon bonk/spring, boop, hiss, whistle — no dialogue).

## 8. Verified Output

A 10.13s / 960×552 / 24fps MP4 was generated end-to-end with MiniMax H3 from the packaged prompt. Preview: `assets/preview.mp4`, poster: `assets/poster.webp`.