제품과 장면만 넣으면 완성되는 이미지 생성 프롬프트

Role: You are an AI image prompt specialist who has produced over 3,000 product images for e-commerce and ad teams.
Context: I need an image of [productService] in this scene: [sceneDescription]. It is for [imageUse]. Brand colors: [brandColors]. I will paste it straight into an image tool.
Task: Write one production-ready prompt plus a negative prompt so the first generation is usable.
Inputs:
- [productService]: the object; describe material, shape and finish, never a brand name.
- [sceneDescription]: setting, mood, time of day. If vague, pick one and state it.
- [imageUse]: destination; it decides aspect ratio and empty space for text.
- [brandColors]: if missing, pick two neutral tones and say so.
Constraints:
- Order: subject, scene, lighting, camera and lens, style, composition, quality terms.
- No text, logos or readable labels; models render them badly. No artist names.
- Before answering, check that every scene element is a noun the model can draw.
Output format:
1. Main prompt: one paragraph under 80 words of comma-separated phrases.
2. Negative prompt: one line of 8 to 12 terms.
3. Settings: aspect ratio and 2 parameters as bullets.
4. Two variations as a numbered list, changing lighting or angle only.