Luxury Close-Up Magazine Cover · skill pública Open Skill
Luxury Close-Up Magazine Cover
Luxury Close-Up Magazine Cover é uma Open Skill pública no Digen. Abra para ver o que este fluxo cria.
Custom
0 0 0
https://d7qlwi7tl2u27.cloudfront.net/skill-covers/prod/4150474/2026-10-15/yt5znbq6.jpg
# Luxury Close-Up Magazine Cover
Produces one luxury magazine cover photo per request: a single woman in an extreme close-up, black-and-gold beauty lighting, luxury wardrobe and elegant gold serif cover typography. Input is a short text brief plus an optional reference portrait that personalizes her identity; a text-only brief works as well.
There is no structured project state: everything needed for a second render (theme, cover wording, identity notes, the last asset) lives in the conversation or on the canvas.
## When to use
- The user asks for a magazine cover, a luxury / black-and-gold cover portrait, or a cover of a named theme (holiday, awards season, bridal, noir…).
- The user provides a portrait and asks to turn it into a cover, or asks to keep that person's face.
- The user asks to redo an earlier cover with a different theme, wardrobe or cover wording.
Not for: full-body fashion editorials, multi-person group covers, posters/logos, or plain portrait retouching.
## Fixed template (template-owned — never change unless the user explicitly asks)
- Layout: 3:4 vertical, extreme close-up, the subject fills most of the frame (head and shoulders or tighter), eyes on the upper third, quiet dark space across the top reserved for the masthead.
- Palette & materials: obsidian black base with champagne gold, antique brass and warm amber accents; liquid satin, silk, velvet, gold lamé, sequins, polished metal.
- Light: beauty lighting with a warm gold rim light on hair and shoulders, luminous specular highlights on cheekbones and collarbone, big soft catchlights in both eyes, realistic skin texture with visible pores, shadows falling off into black.
- Typography: high-contrast serif masthead in gold foil across the top, only 1–3 words per line, clean real letterforms — never glyph soup.
- Only wardrobe, pose, scene, hair and the cover wording follow the theme; the subject's identity never changes.
Read the full style spec before writing the first prompt of a session:
`read_skill_file('luxury-cover-portrait', 'references/luxury-cover-style.md')`.
## Steps
1. Read the brief and extract three things: the theme/mood, the subject (identity from the reference image, or features described by the user), and the exact cover wording the user asked for. If no wording is given, pick a 1–3 word masthead that matches the theme. If the user attached several people or a non-portrait image, ask which single subject to use.
2. Read `references/luxury-cover-style.md` (first cover of the session, or whenever the theme changes) and take wardrobe, hair & makeup, jewelry and background for that theme from its theme table.
3. Compose the English prompt from the fixed template plus the variable slots — subject, theme, wardrobe, hair & makeup, jewelry, cover line. Master skeleton to fill in:
`"Luxury magazine cover portrait, extreme close-up of {SUBJECT}. {WARDROBE}, {HAIR_AND_MAKEUP}, {JEWELRY}. Obsidian black and champagne-gold palette, {THEME} mood, liquid satin and polished gold with specular highlights, beauty lighting with a warm gold rim light on hair and shoulders, big soft catchlights in the eyes, glossy lips, realistic skin texture with visible pores. 3:4 vertical face-filling crop, head and shoulders, eyes on the upper third, dark background falling into black, clean negative space across the top for an elegant gold serif masthead reading '{COVER_LINE}'. Photorealistic, 85mm portrait lens, shallow depth of field, high-end fashion magazine cover."`
When a reference image is used, prepend: `"Keep the exact face, features, skin tone and identity of the woman in the reference image; change only wardrobe, pose, hairstyle and scene to match the theme."`
4. Generate with the channel that matches the input:
- Reference portrait provided: `generate_image(prompt=..., model='i2i', input_urls=['<reference asset_id or uri>'], orientation='portrait', aspect_ratio='3:4', name='<short cover title>', entity_type='magazine_covers')`
- Text o