Transformation Studio · مهارة Open Skill عامة

Transformation Studio

Transformation Studio هي Open Skill عامة على Digen. افتحها لتعرف ماذا ينشئ هذا التدفق.

@fangthebeardie

Custom

19 4 5

https://d7qlwi7tl2u27.cloudfront.net/skill-covers/prod/1499793/2026-09-22/vx2kxmmd.png

---
name: Transformation Studio
description: >
  Creates animated 9:16 character transformation videos. Upload a character
  image and describe what they should become. Transformation Studio first
  normalizes non-vertical source images to a true 9:16 portrait composition,
  then animates the same character progressively transforming into the requested
  person, creature, fantasy character, age, sex/presentation, material, or style.
  It prioritizes correct target anatomy, recognizable identity when compatible,
  continuous movement, clothing continuity, no nudity, and true 9:16 output.
---

# TRANSFORMATION STUDIO

## PURPOSE

Transformation Studio creates ANIMATED character transformation videos.

The user uploads an image containing a character and describes what they want
that character to become.

Examples:

- older
- younger
- woman
- man
- Medusa
- vampire
- werewolf
- skeleton
- undead skeleton
- mermaid
- merman
- witch
- wizard
- elf
- fairy
- demon
- angel
- superhero
- fantasy warrior
- cyberpunk character
- medieval character
- living clay
- stone
- ice
- metal
- monster
- custom transformation

The SAME character must visibly transform during the video.

The transformation itself is the main event.

The final output MUST be a VIDEO.

---

# FIVE HARD LOCKS

These requirements override creative decisions.

1. TRUE 9:16 PORTRAIT
2. CORRECT TARGET FORM / ANATOMY
3. NO NUDITY
4. CONTINUOUS ANIMATED MORPH
5. IDENTITY PRESERVATION WHEN COMPATIBLE WITH THE TARGET

Do not sacrifice these requirements for cinematic effects.

---

# STEP ZERO — NORMALIZE EVERY SOURCE TO 9:16

This step occurs BEFORE video generation.

Inspect the uploaded source image's aspect ratio.

If the source image is already a genuine 9:16 portrait composition:

USE IT.

If the source image is:

- square
- landscape
- horizontal
- 1:1
- 4:3
- 3:2
- another non-9:16 ratio

DO NOT send the original image directly into image-to-video generation.

This is extremely important.

Some image-to-video generators inherit the source image's dimensions even when
the text prompt requests vertical video.

Therefore:

FIRST convert the source into a genuine 9:16 portrait source image.

Use an available image editing, image generation, outpainting, canvas
expansion, reframing, or equivalent capability.

Create:

VIDEO_SOURCE_IMAGE

with a TRUE 9:16 PORTRAIT composition.

Preferred dimensions:

1080 × 1920

Other acceptable true 9:16 dimensions include:

720 × 1280
540 × 960

or the closest supported true portrait equivalent.

---

# 9:16 SOURCE EXPANSION

When converting a square or landscape source:

PRESERVE:

- same character
- same face
- same identity
- same clothing
- same hairstyle
- same skin tone
- same pose when practical
- same environment
- same lighting
- same visual style

Naturally extend the scene vertically.

Generate believable additional background above and/or below the original
image.

When additional body area must be generated, maintain plausible anatomy and
matching clothing.

DO NOT:

- stretch the original image
- squash the image
- distort the person
- add black bars
- add blurred bars
- place the square picture in the center of a vertical blank canvas
- create a landscape image inside a vertical border

The new image must look like the original photograph or artwork had naturally
been composed as 9:16.

---

# 9:16 INPUT LOCK

Once VIDEO_SOURCE_IMAGE exists:

ONLY use the normalized 9:16 version as the visual source for video generation.

Do NOT accidentally send the original square or landscape image to the video
generator.

The workflow must be:

ORIGINAL UPLOAD
↓
9:16 NORMALIZATION
↓
VIDEO_SOURCE_IMAGE
↓
VIDEO GENERATION

not:

ORIGINAL SQUARE IMAGE
↓
VIDEO GENERATION
↓
TRY TO FORCE 9:16

---

# ABSOLUTE VIDEO OUTPUT REQUIREMENT

EVERY Transformation Studio video MUST be:

9:16 PORTRAIT
VERTICAL

Preferred resolution:

1080 × 1920

The agent MUST use the selected video tool's actual:

- aspect ratio control
- orientation control
- can