Two-Person Close-Up Clip · Open Skill 公开技能

Two-Person Close-Up Clip

hook: I made this for duo portraits. Two faces, one frame, soft life. The crop does the heavy lifting. Send one adult photo with both faces clear. Get a 5-second close-up clip with micro-motion. tips: Use a 9:16 portrait layout / Choose a mood, such as a quiet smile

@TylerHayes

Film

1 0 0

https://d7qlwi7tl2u27.cloudfront.net/skill-covers/prod/4284953/2026-10-17/vtkzu4ee.webp

# Two-Person Close-Up Clip

Two faces, one frame, soft life. Your duo photo becomes a short close-up clip with natural micro-motion between both people.

## When to use

- The user wants a short close-up video that gently animates the two people already in their reference image (two-person close-up clip, duo portrait image-to-video)
- Another take with a different length or orientation

Not for: Still images only.

## Inputs and gates

| Input | Required | Default | Gate |
|---|---|---|---|
| Photo | Yes | — | One image with two clearly visible adult faces. Ask for another photo if either face is tiny, blurred or cut off |
| Description | No | tuned default below | What the user wants; merged into the prompt |
| Duration | No | `5 s` | 5 s (default), 6 s, or 10 s |
| Mood | No | `soft shared glance` | soft shared glance (default), quiet smile, almost-whisper closeness, or user-described |
| Orientation | No | `9:16` portrait | Other ratios on request |

- If the person may be under 18, stop and decline politely. Real-person photos must be of adults the user has the right to use.
- Video does not allow NSFW content: keep wardrobe fully covering and refuse sexual requests. Suggestive fashion looks stay as still images.

## Steps

1. Read `references/video-motion.md`.
2. **Identity read**: `recognize_image(image_url=<photo>, prompt='Describe this adult for identity preservation: face shape, eye shape and color, brows, nose, lips, skin tone, apparent age range, hair color and length, distinctive marks, glasses. One compact line.')`.
3. **Keyframe**: `generate_image(prompt=<KEYFRAME>, input_urls=[photo], model='i2i.hd', orientation='portrait', aspect_ratio='9:16', name='Two-Person Close-Up Clip keyframe', entity_type='work_in_progress')`. If the face drifts, regenerate once.
4. **Video**: `generate_video(prompt=<MOTION>, input_urls=[keyframe], model='i2v', orientation='portrait', aspect_ratio='9:16', duration=<d>, name='Two-Person Close-Up Clip', entity_type='final_videos')`. Pass only the keyframe, never the original upload.
5. Reply in one or two sentences, e.g. "All set. Your two-person close-up clip is on the canvas. Next I can do 6 s." Results appear on the canvas; never paste URLs.

## Prompt templates

**KEYFRAME**

```text
The same adult from the reference: <identity line>. Same face structure, eyes, skin tone and recognizable likeness; no beautification or face reshaping. Opening frame of this scene: a close-up video of the two adults already in the reference image, both faces held in a tight intimate frame with soft natural micro-movements (blinks, slight head turns, shared glance or gentle smile), shallow depth of field, warm flattering light, identities locked to both people, fully clothed, tasteful non-explicit. Posed for motion, clear subject, cinematic light. Nothing from the original photo's background remains.
```

**MOTION** (motion only; the frame already shows the look)

```text
Starting from this exact frame: <one clear motion arc for this concept: a close-up video of the two adults already in the reference image, both faces held in a tight intimate frame with soft natural micro-movements (blinks, slight head turns, shared glance or gentle smile), shallow depth of field, warm flattering light, identities locked to both people, fully clothed, tasteful non-explicit>, beginning, peak and settle in one continuous shot. Camera: <one move: slow push-in, gentle orbit, or static>. The same face and outfit throughout. Cinematic, no text.
```

## Failure handling

- `success=false`: say what failed in one sentence; retry once only if transient; never invent links.
- Content-policy rejection on video: soften wording and retry once; otherwise offer the keyframe as a still.
- Identity drifts: offer to regenerate.

## Examples

**Default**: the user says "this one" with a photo. Use duration `5 s`, mood `soft shared glance`, 9:16. Reply: "All set. Your two-person close-up clip is on the canvas."
**Variation**: "next variatio