Baby & Kitten Clips · öffentlicher Open Skill

Baby & Kitten Clips

Baby & Kitten Clips ist ein öffentlicher Open Skill auf Digen. Öffne ihn, um zu sehen, was dieser Workflow erstellt.

@condor.3000000

Anime

6 0 5

https://d7qlwi7tl2u27.cloudfront.net/skill-covers/prod/4178668/2026-09-29/dpxhmqmj.jpg

# Baby & Kitten Clips

Create adorable short vertical (9:16) clips and images of a baby playing with a small kitten, keeping the characters and style consistent across generations.

## When to use
- The user wants a cute video/image of a baby playing with a kitten (e.g. "طفل رضيع يلعب مع قط صغير").
- The user wants to continue, extend, or restyle a previous baby+kitten scene with the same characters.
- The user wants the baby/kitten look, art style, or color palette remembered for future requests.

## Steps
1. **Remember the user's preferences first.** Call `memory_query` (and `memory_whoami` if useful) to check for saved details about the baby's appearance, the kitten's appearance, the art style, and the color palette. If nothing is found, don't invent fixed characters — ask the user briefly for the baby/kitten look and style, or default to a cute cartoon/claymation style.
2. **Save new recurring details.** If the user states a recurring preference (specific baby look, kitten breed/color, art style, palette, brand), call `memory_store` with `kind` = `preference` (or `person` for a named baby/kitten) and a `namespace` of `baby-kitten-clips` so it persists for next time. Only store durable preferences, not one-off scene details.
3. **Generate the scene image.** Call `generate_image` with `model='t2i'`, `orientation='portrait'`, `resolution='720P'`, `name='baby-kitten-scene'`, `entity_type='baby_kitten_scenes'`. Prompt in English, describing a cute baby playing with a small kitten in a charming cartoon/claymation style (unless the user prefers another style), warm soft lighting, and the remembered palette.
4. **Animate it into a short clip.** Call `generate_video` with `model='i2v'`, passing the generated image's asset_id in `input_urls`, `orientation='portrait'`, `duration` ~5s, `name='baby-kitten-clip'`, `entity_type='baby_kitten_clips'`. Describe gentle, heartwarming motion (baby giggling, kitten pawing/playing).
5. **Optionally stitch multiple clips.** If the user wants a longer sequence, generate a few clips and combine them with `mix_clips` (`model='clip-mixer'`), `orientation='portrait'`, `name='baby-kitten-mix'`, `entity_type='final_videos'`.
6. **Reply naturally** summarizing what was made; don't paste raw URLs.

## Notes
- Always keep the same baby and kitten appearance across a session by reusing the first generated image as a reference (`input_urls`) for later edits.
- To restyle away from the default (e.g. to realistic or anime), use `generate_image` with `model='i2i'` on the existing scene image.
- To forget a saved preference, use `memory_query` to find its `id`, then `memory_forget`.
- Report failures honestly; never fabricate URLs. Video generation does not support NSFW/adult content — politely decline such requests.
- Intermediate step-by-step products (scene images, individual clips) will be visible on the canvas alongside the final result.