UI はローカライズ済みです。スキルの紹介と説明文書は選択した言語へ AI が自動翻訳し、コード・コマンド・パス・専門用語は原文のままです。原作者の文書が優先されます。
スキルを探す
デザイン · SKILL.md

ai-fashion-lookbook

ai-fashion-lookbook

Generate complete fashion lookbook layouts using each::sense AI. Create cohesive multi-look editorial spreads with consistent styling, themed collections, and professional compositions for seasonal campaigns, brand launches, and fashion portfolios. Use for: fashion collections, seasonal lookbooks, brand campaigns, editorial spreads, portfolio presentation, …

Independent GitHub snapshot · Automated classification · Original documentation governs

スキルの説明

GitHub

name: ai-fashion-lookbook description: "Generate complete fashion lookbook layouts using each::sense AI. Create cohesive multi-look editorial spreads with consistent styling, themed collections, and professional compositions for seasonal campaigns, brand launches, and fashion portfolios. Use for: fashion collections, seasonal lookbooks, brand campaigns, editorial spreads, portfolio presentation, fashion marketing. Triggers: fashion lookbook, lookbook generator, fashion collection, editorial spread, seasonal collection, lookbook layout, fashion campaign, brand lookbook, collection showcase, style book" allowed-tools: Bash(curl *), WebFetch

AI Fashion Lookbook

Generate complete fashion lookbook imagery and layouts using each::sense — the intelligent AI agent that automatically selects the best model for your request.

Quick Start

Requires an each::labs API key. Get one at eachlabs.ai.

Using curl

curl -X POST https://eachsense-agent.core.eachlabs.run/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "X-API-Key: $EACHLABS_API_KEY" \
  -d '{
    "messages": [
      {
        "role": "user",
        "content": "A fashion lookbook photo: a model wearing a flowing maxi dress in terracotta, walking through a sunlit olive grove in Tuscany. Wind catching the fabric, golden hour, editorial photography, full body shot, warm color grading, Vogue style"
      }
    ],
    "stream": false
  }'

Using Python (OpenAI SDK)

from openai import OpenAI

client = OpenAI(
    api_key="YOUR_EACHLABS_API_KEY",
    base_url="https://eachsense-agent.core.eachlabs.run/v1"
)

response = client.chat.completions.create(
    model="eachsense/beta",
    messages=[{
        "role": "user",
        "content": "A fashion lookbook photo: a model wearing a flowing maxi dress in terracotta, walking through a sunlit olive grove in Tuscany. Wind catching the fabric, golden hour, editorial photography, full body shot, warm color grading, Vogue style"
    }]
)

print(response.choices[0].message.content)

With Reference Image

Provide a mood board or sample look for the collection aesthetic:

curl -X POST https://eachsense-agent.core.eachlabs.run/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "X-API-Key: $EACHLABS_API_KEY" \
  -d '{
    "messages": [
      {
        "role": "user",
        "content": [
              {"type": "text", "text": "Create a lookbook photo matching this mood and color palette. A model in a structured wool coat and turtleneck, same visual tone and lighting style as the reference. Urban winter setting, overcast sky, editorial fashion photography."},
              {"type": "image_url", "image_url": {"url": "https://example.com/mood-board.jpg"}}
            ]
          }
    ],
    "stream": false
  }'

Images are sent inside messages using the OpenAI multimodal content format. Maximum 4 images per request.

Streaming

Set "stream": true for real-time SSE responses, or "stream": false for complete result in a single response. Streaming is useful for showing progress in UIs; non-streaming is simpler for scripts and automation.

Lookbook Themes

ThemeSettingColor PaletteSeason
Mediterranean Summerolive groves, white walls, coastterracotta, white, sky blueSS
Nordic Wintersnow, concrete, birch forestsivory, charcoal, pale blueAW
Urban Nightcity lights, neon, rain-wet streetsblack, silver, electric blueAW
Desert Modernistdesert, mid-century architecturesand, rust, sageResort
Parisian Springcafes, gardens, architectureblush, cream, dusty blueSS
Industrial Rawwarehouses, steel, exposed brickgrey, khaki, raw denimPre-Fall

Prompt Engineering Tips

Prompt Structure

"A fashion lookbook photo:" + [model description] + [outfit details] + [setting/location] + [mood/atmosphere] + [photography style] + [camera details]

Consistency Across Looks

For a cohesive lookbook, keep these elements constant across all prompts:

  • Location type (same setting or setting category)
  • Lighting (golden hour, overcast, studio)
  • Color grading (warm, cool, desaturated)
  • Photography style (editorial, commercial, street)
  • Model description (same height, hair, build)

Examples

Look 1 — Hero Shot

curl -X POST https://eachsense-agent.core.eachlabs.run/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "X-API-Key: $EACHLABS_API_KEY" \
  -d '{
    "messages": [
      {
        "role": "user",
        "content": "A fashion lookbook hero shot: a tall female model with slicked-back dark hair wearing a sculptural white blazer with exaggerated shoulders over a matching wide-leg trouser. Standing on a rooftop overlooking a city at dusk. Dramatic side lighting, strong shadow, high fashion editorial, shot from low angle, desaturated teal and warm amber color grade"
      }
    ],
    "stream": false
  }'

Look 2 — Detail Close-Up

curl -X POST https://eachsense-agent.core.eachlabs.run/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "X-API-Key: $EACHLABS_API_KEY" \
  -d '{
    "messages": [
      {
        "role": "user",
        "content": "A fashion lookbook detail shot: close-up on hands adjusting the cuff of a cream silk blouse, delicate pearl button visible, soft natural light from a window, shallow depth of field, minimal warm-toned editorial photography, macro lens feel"
      }
    ],
    "stream": false
  }'

Look 3 — Lifestyle Context

curl -X POST https://eachsense-agent.core.eachlabs.run/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "X-API-Key: $EACHLABS_API_KEY" \
  -d '{
    "messages": [
      {
        "role": "user",
        "content": "A fashion lookbook lifestyle shot: a model in a camel cashmere overcoat and cream scarf, walking through a misty Parisian street with bare autumn trees. Candid mid-stride movement, film photography aesthetic with slight grain, warm desaturated tones, full body shot"
      }
    ],
    "stream": false
  }'

Collection Workflow: Seasonal Lookbook

COLLECTION_THEME="Autumn/Winter collection, Nordic minimalism, muted earth tones, overcast natural light, editorial fashion photography"

LOOKS=(
  "Look 1: oversized charcoal wool coat over cream cable-knit sweater and dark brown leather trousers, standing in a birch forest"
  "Look 2: camel turtleneck dress belted at the waist with cognac leather boots, seated on a concrete bench in an urban park"
  "Look 3: layered outfit — grey knit vest over white poplin shirt with wide olive trousers, walking along a misty lakefront"
  "Look 4: black tailored suit with an open-neck silk shirt in ivory, leaning against a steel railing at a modernist building"
  "Look 5: chunky cream fisherman sweater with a long pleated skirt in heather grey, standing on weathered wooden pier"
)

for LOOK in "${LOOKS[@]}"; do
  curl -X POST https://eachsense-agent.core.eachlabs.run/v1/chat/completions \
    -H "Content-Type: application/json" \
    -H "X-API-Key: $EACHLABS_API_KEY" \
    -d "{
      \"messages\": [{\"role\": \"user\", \"content\": \"A fashion lookbook photo: female model with short auburn hair, $LOOK. $COLLECTION_THEME, full body shot, Vogue Scandinavia style\"}],
      \"stream\": false
    }"
  echo "=== $LOOK ==="
done

Shot List for a Complete Lookbook

  1. Hero / Cover — full body, dramatic composition, establishes the collection mood
  2. Full Length Front — clean front-facing shot showing complete outfit
  3. Three-Quarter Turn — shows garment from an angled perspective
  4. Detail Close-Up — fabric texture, buttons, stitching, jewelry
  5. Back View — reveals back design elements
  6. Lifestyle / Movement — model walking, turning, interacting with environment
  7. Group / Pairing — two looks styled together for contrast

Common Pitfalls

  • Inconsistent model across shots — repeat the exact same model description in every prompt.
  • Mixed photography styles — keep the same camera, lighting, and grading keywords throughout.
  • Overly complex backgrounds that distract from the garments. The setting should complement, not compete.
  • Ignoring negative space — lookbooks need breathing room for text overlays in final layout.

Related Skills

Related Models

Documentation

保存したスキル