Cinematic Golden Hour Romance · skill publik Open Skill

Cinematic Golden Hour Romance

Cinematic Golden Hour Romance adalah Open Skill publik di Digen. Buka untuk melihat apa yang dibuat alur ini.

@saharasaine

Beauty

8 0 6

https://d7qlwi7tl2u27.cloudfront.net/skill-covers/prod/4164563/2026-10-05/j32s4vkx.jpg

# Cinematic Golden Hour Romance

## When to use
- The user wants a warm, cinematic golden-hour romance short: a close-up of an attractive curvy Japanese model by a window at golden hour, with a playful/intimate voiceover.
- The user wants to redo or tweak one step: change the scene/visual, change the voiceover line, regenerate the clip, or mix the final video.

## Defaults
- Orientation: **portrait (9:16)** for short-form; use landscape only if the user asks.
- Default visual prompt (use as-is unless the user changes the scene):
  "Close-up shot of an attractive curvy Japanese model standing near a window during golden hour, sun glinting through her hair. Soft atmospheric lighting, romantic atmosphere, detailed facial expressions, subtle smile, warm tone grading, 85mm portrait lens, photorealistic 4K."
- Default voiceover line (use as-is unless the user provides their own):
  "You think you can handle me today? You talk a big game, but whenever we're this close, you get so quiet... tell me what you really want to do."

## Steps
1. Confirm the scene and the voiceover line. If the user only says "make it" or "redo it," use the defaults above. Ask only if the user's request is ambiguous.
2. **Generate the image** with `generate_image`:
   - `model='t2i'`, `orientation='portrait'`, `resolution='1080P'`
   - `prompt` = the scene description (default above, or the user's custom scene)
   - `name='golden_hour_shot'`, `entity_type='storyboard_images'`
3. **Generate the voiceover** with `generate_tts`:
   - `model='tts.design'`
   - `text` = the voiceover line (default above, or the user's custom line)
   - `voice_instruction` (in English) = a warm, breathy, playful, intimate female voice, soft and flirty, slow seductive delivery, gentle smile in the tone
   - `name='romance_voiceover'`, `entity_type='voiceovers'`
4. **Generate the animated clip** with `generate_video`:
   - `model='ai2v'` (image + audio → video)
   - `input_urls=[<golden_hour_shot asset_id>]`, `audio_url=<romance_voiceover asset_id>`
   - `orientation='portrait'`, `resolution='720P'`
   - `prompt` = subtle natural motion: gentle hair movement in the golden light, soft breathing, a slow subtle smile, warm light flickering across her face
   - `name='golden_hour_clip'`, `entity_type='storyboard_videos'`
5. **Optional — add romantic background music**: generate with `generate_music` (`model='music'`, `name='romance_music'`, `entity_type='music'`), a warm, intimate, slow romantic instrumental (soft piano / strings), then mix.
6. **Mix the final video** with `mix_clips`:
   - `clips=[{ video: { url: <golden_hour_clip asset_id> }, audios: [ { url: <romance_voiceover asset_id> }, { url: <romance_music asset_id>, volume: 0.35 } ] }]`
   - `orientation='portrait'`, `name='golden_hour_final'`, `entity_type='final_videos'`
   - If no music was generated, mix with just the voiceover audio.
7. Reply in natural language with the final video; offer to tweak the scene, the line, or the mix.

## Notes
- Use the asset_id returned by each generation step as the input for the next step; never fabricate URLs.
- If a generation fails or returns an error, report it honestly and retry once with a slightly adjusted prompt.
- Video generation does not support NSFW/adult content on any channel — keep the scene romantic and tasteful (clothed, suggestive but not explicit).
- Intermediate products (image, voiceover, clip, music) will appear on the canvas alongside the final video; that's expected.
- When a stage completes, call `set_guidance` to suggest next actions (e.g. "change the voiceover line", "make it landscape", "redo the scene").