3D Family Animation Creator · публичный навык Open Skill

3D Family Animation Creator

3D Family Animation Creator — публичный Open Skill на Digen. Откройте, чтобы увидеть, что создаёт этот процесс.

@jeeai

ShortDrama

5 2 6

https://d7qlwi7tl2u27.cloudfront.net/skill-covers/prod/1821775/2026-09-20/8gx7ahf8.jpg

# 3D Family Animation Creator

## When to use
- The user wants an original, colourful, cinematic, family-friendly 3D animated video (children's stories, nursery rhymes, nasheeds with animated visuals, educational adventures, moral stories, family-friendly music videos, fantasy and animal stories).
- The user wants the full creative package (concept → character bible → scene plan → style rules → thumbnail → publishing metadata) and/or the actual generated assets (character art, scene stills, voiceover, music, final animated video).
- The user wants to revise or continue an existing 3D family animation package.

## Style rules (apply to every prompt)
- High-quality theatrical 3D animation look: rounded character designs, large expressive eyes, expressive facial animation, warm cinematic lighting, detailed but uncluttered backgrounds, bright colours, smooth character movement, emotional close-ups, whimsical environments, gentle depth of field.
- Playful comedy + heartfelt storytelling.
- Do NOT copy specific copyrighted characters, films, logos, or exact studio branding. Always create original characters and worlds.

## Steps
1. First use `write_todos` to list the steps, and update the whole list as you progress.
2. Clarify the idea in one or two rounds if needed: story/theme, target age, characters, whether they want just the package or also the generated assets. Once clear, lock onto this one package.
3. Produce the full creative package as a Markdown document via `create_document(content=..., name="<video title> — 3D Family Animation Package", entity_type="scripts")`. Fill in every field below:
   - VIDEO TITLE / CONTENT TYPE / TARGET AGE / MOOD / VISUAL CONCEPT / STORY SUMMARY
   - CHARACTER BIBLE (per character): Name, Species, Age, Appearance, Clothing, Personality, Voice, Movement style
   - SCENE PLAN (per scene): Duration, Location, Character action, Facial expression, Dialogue or narration, Camera, Lighting, Sound, Image or video-generation prompt
   - STYLE CONSISTENCY RULES (same character proportions, same clothing & colours, consistent facial features, same lighting language across scenes, no unrequested characters)
   - THUMBNAIL PROMPT
   - YOUTUBE TITLE / DESCRIPTION / KEYWORDS / HASHTAGS
4. If the user wants generated assets, proceed (each asset-producing call must pass the `name`/`entity_type` below):
   - Character art: `generate_image(model='t2i', prompt=<character description in the style rules>, orientation='landscape', resolution='720P', name="<Character Name> — character art", entity_type="characters")`. For refinements/consistency edits, use `model='i2i'` with the previous image in `input_urls`.
   - Scene stills: `generate_image(model='t2i', prompt=<scene prompt from the scene plan>, orientation='landscape', resolution='720P', name="Scene <n> — <location>", entity_type="storyboard_images")`. Keep character proportions/colours consistent with the character art.
   - Narration voiceover: `generate_tts(model='tts.design', text=<dialogue/narration>, voice_instruction=<English voice description matching the character's Voice>, name="<Character> — VO", entity_type="voiceovers")`.
   - Music: `generate_music(model='music', prompt=<mood/genre description>, name="<Video Title> — soundtrack", entity_type="music")`.
   - Scene videos: `generate_video(model='ref2v', input_urls=[<character art / scene still asset_id>], prompt=<motion description>, orientation='landscape', resolution='720P', name="Scene <n> — video", entity_type="storyboard_videos")` to keep the character/subject consistent. For a single still with no consistency need, `model='i2v'` is fine.
   - Final cut: `mix_clips(clips=[...], name="<Video Title> — final", entity_type="final_videos", orientation='landscape')` to stitch scene videos and mix in the voiceover/music tracks.
5. When a stage completes, call `set_guidance` to suggest next actions (e.g. "generate the character art", "make the scene stills", "produce the voiceover & music", "render the fi