Night Street Costume Portrait · Open Skill 公开技能
Night Street Costume Portrait
Night Street Costume Portrait 是 Digen 上的公开 Open Skill。打开后即可了解该工作流能制作什么。
TryOn
2 0 3
https://d7qlwi7tl2u27.cloudfront.net/skill-covers/prod/4284954/2026-10-20/22pqtuha.webp
# Night Street Costume Portrait Fog, lamplight, a long coat swinging mid-stride. The same adult is recast in a full Halloween costume on a cobbled night street, hands ready for the walk. ## When to use - The user wants the adult in their photo dressed in a Halloween costume and walking down a foggy lamp-lit street at night (Halloween costume portrait, street costume photoshoot, night walk costume picture) - A variation on a previous result: velvet vampire cape Not for: Video output (use a video skill). ## Inputs and gates | Input | Required | Default | Gate | |---|---|---|---| | Photo | Yes | — | One adult, face clear; three-quarter or full body preferred so the costume can read. Ask for another photo if the face is tiny, blurred or hidden | | Description | No | tuned default below | What the user wants; merged into the prompt | | Costume | No | `horned gothic gown` | horned gothic gown (default), velvet vampire cape, witch dress with point hat, or user-described | | Mood | No | `ominous elegance` | ominous elegance (default), playful spooky, cinematic mystery | | Ratio | No | `3:4 default` | 3:4 default; 2:3, 9:16 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. - Fashion, dark and horror themes are fine as stills; refuse explicit sexual content and anything sexualizing minors. ## Steps 1. Read `references/image-edit.md` for prompt structure. 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. **Generate**: `generate_image(prompt=<PROMPT>, input_urls=[photo], model='i2i.hd', orientation='portrait', aspect_ratio='3:4', name='Night Street Costume Portrait', entity_type='skill_generation')`. Merge the user's details and options into the template. 4. **Self-check, one retry**: look at the result. If a person photo was used and the face no longer reads as the same person, regenerate with the identity clause first. If someone glancing at it would not instantly see the intended change (it reads as a light filter, the old clothes or background remain, or the style is faint), regenerate once with the change described bigger and more concretely. 5. Reply in one or two sentences, e.g. "Done! Here's your night street costume portrait. I can also try velvet vampire cape if you like." Results appear on the canvas; never paste URLs. ## Prompt templates **PROMPT** ```text The same adult from the reference: <identity line>. Same face structure, eyes, skin tone and recognizable likeness; no beautification or face reshaping. Intentional changes: the same adult walking toward camera mid-stride down a narrow **cobbled street at night** in **<costume: horned gothic gown in black lace (default), velvet vampire cape, wicked witch dress with a pointed hat, or user-described Halloween costume>** — full costume, real fabric, themed accessory, bold silhouette — wet **cobblestones** catching warm lantern pools and cool blue mist, brick façades and iron lamps; realistic night street photography, slight haze, film grain, fully clothed, non-graphic. <user details and options>. The new outfit fully replaces the original clothes with a clearly different silhouette; real fabric texture, accurate fit and drape, styling and accessories; a setting and light that suit the outfit. Lighting and materials described concretely; composition suited to portrait 3:4. Photoreal unless the style says otherwise, fully clothed, no text unless requested. ``` Quality bar: Make the new outfit read as real fabric: bold silhouette change from the original clothes, named materials and textures, hero lighting, and a scene that suits the outfit (not the selfie background). ## Failure handling - `success=false`: say what failed in one sentence; retry once only if