MOMO 送餐猫短片 · публичный навык Open Skill
MOMO 送餐猫短片
MOMO 送餐猫短片 — публичный Open Skill на Digen. Откройте, чтобы увидеть, что создаёт этот процесс.
Anime
2 3 3
https://d7qlwi7tl2u27.cloudfront.net/skill-covers/prod/4161311/2026-09-23/uwwe76r6.png
# MOMO 送餐猫短片 (MOMO Delivery Cat)
Create a ~15s landscape (16:9) heartwarming video of MOMO, a cute fluffy orange British shorthair cat working as a McDonald's delivery driver. Pixar-style 3D look, realistic soft fur, cinematic warm lighting, funny and heartwarming tone, smooth motion, 4K-quality feel.
## When to use
- The user wants a video of MOMO the orange British shorthair cat as a McDonald's delivery worker (packing food, riding a yellow scooter through the city, greeting people, handing food to a customer).
- The user wants to revise one stage of an existing MOMO video: the character design, a single shot, the background music, or the final mix.
## Reference files
- `references/shots.md`: The fixed shot list (4 shots) with detailed English prompts. Read it before generating clips.
## Steps
1. First call `write_todos` to list the stages (character ref → video clips → BGM → final mix) and update it as you progress.
2. **Character reference image** — call `generate_image` with `model='t2i'`, `orientation='landscape'`, `resolution='1080P'`, `name='MOMO_character_ref'`, `entity_type='characters'`. Prompt: a cute fluffy orange British shorthair cat named MOMO wearing a red-and-yellow McDonald's delivery uniform, tiny employee hat, and a name tag reading "MOMO", standing in a bright modern city street, Pixar-style 3D, realistic soft fur, cinematic warm lighting, full body, centered. This image is the identity anchor for all clips.
3. **Video clips** — read `references/shots.md` for the 4 shot prompts. For each shot, call `generate_video` with `model='ref2v'`, `input_urls=[MOMO_character_ref]` (the character reference from step 2), `orientation='landscape'`, `resolution='1080P'`, `duration=4`, `name='MOMO_shotN'`, `entity_type='storyboard_videos'`. Keep MOMO's design consistent across all shots.
4. **Background music** — call `generate_music` with `name='MOMO_bgm'`, `entity_type='music'`. Prompt: upbeat, cheerful, heartwarming light pop with a playful bounce, warm and cozy, no vocals, ~15 seconds.
5. **Final mix** — call `mix_clips` to concatenate the 4 clips in order and lay the BGM underneath: `clips=[{videos:[shot1,shot2,shot3,shot4], audios:[{url:MOMO_bgm}]}]`, `orientation='landscape'`, `name='MOMO_delivery_final'`, `entity_type='final_videos'`.
6. Summarize the result in natural language (don't paste raw URLs). Call `set_guidance` to suggest next actions (e.g. tweak a shot, change the music, redo the character).
## Notes
- Keep the character consistent: always reuse the same character reference image in `input_urls` for every clip.
- If the user only wants to revise one stage, redo only that stage and re-run the final mix; don't regenerate everything.
- Report failures honestly; never fabricate URLs. Politely decline any request to make the content adult/NSFW (video generation does not support it).
- Intermediate products (character ref, clips, BGM) will be visible on the canvas alongside the final video.