By name: hero-visual-prompt-generator description: | Generate premium AI image and video prompts for landing page hero sections. Creates cinematic, commercial-quality visual prompts optimized for tools like Nano Banana, FreePik AI, Google Veo Flow, Runway ML, and Pika Labs.
Use when user mentions:
- "Criar prompt para hero" / "Hero image prompt" / "Hero video prompt"
- "Prompt para landing page" / "Imagem para hero section"
- "Video para landing page" / "Visual premium para LP"
- "Prompt Nano Banana" / "Prompt para FreePik AI" / "Prompt para Veo Flow"
- "Animar imagem de produto"
- Any request for premium product visuals for websites
Supports: Beverages, Food, Technology, Cosmetics, Fashion, Lifestyle products. Outputs: Structured prompts in English with desktop (16:9) and mobile (9:16) versions.
Hero Visual Prompt Generator
Generate premium AI prompts for hero section images and videos.
Workflow
Phase 1: Information Gathering
Always ask before generating:
- "Qual é o produto/objeto principal da landing page?"
- "Pode enviar uma imagem de referência do produto?"
- "Estamos criando para desktop, mobile ou ambos?"
- "Qual sensação queremos transmitir? (premium, energético, natural, sofisticado)"
Required inputs:
- Product type and reference image
- Target platform (Desktop/Mobile/Both)
- Desired style/mood
- Dominant color palette (extract from image or ask)
Phase 2: Image Prompt Generation
Structure (always in English):
Subject: [Dynamic product/ingredient description in action]
Composition: [Photography type - macro, commercial, high-speed, etc.]
Key Elements: [Main visual elements - textures, particles, liquids, etc.]
Environment: [Background - monochrome void, gradient, studio, etc.]
Lighting: [Dramatic lighting - backlight, rim light, golden hour, etc.]
Style & Details: [Technical quality - hyper-realistic, 8K, cinematic, etc.]
Aspect Ratio: [Platform-based ratio]
Required technical elements:
- Hyper-realistic, 8K resolution, Ultra-sharp focus
- Cinematic color grading, Textural detail emphasis
- Professional commercial photography style
Aspect ratios:
- Desktop: 16:9 or 21:9
- Mobile: 9:16 or 4:5
Phase 3: Review and Selection
After user generates images and returns with favorites:
- Request the generated images they liked most
- Analyze composition, implied movement, highlighted elements
- Use selected images as base for video prompts
Phase 4: Video Prompt Generation
Structure (always in English):
Sequence: [Sequence type - product assembly, reveal, explosion, etc.]
Visuals:
- Start: [Initial state]
- Middle: [Main transition/action]
- End: [Final composition]
Lighting & Camera: [Light transitions and camera movement]
Style: [Technical quality - 4K, slow motion, commercial videography]
Duration: [Suggested duration - 3-5 seconds for hero]
Aspect Ratio: [Platform-based ratio]
Video aspect ratios:
- Desktop: 16:9 (1920x1080) or 21:9 (2560x1080)
- Mobile: 9:16 (1080x1920) or 4:5 (1080x1350)
References
- Category-specific formulas: See references/image-prompt-formulas.md
- Video patterns: See references/video-prompt-formulas.md
- Lighting techniques: See references/lighting-techniques.md
- Camera movements: See references/camera-movements.md
- Aspect ratio guide: See references/aspect-ratios.md
- Style keywords: See references/style-keywords.md
Output Format
For Images:
## IMAGE PROMPT - [PRODUCT] - [PLATFORM]
### Desktop Version (16:9)
[Complete prompt in English]
### Mobile Version (9:16)
[Adapted prompt for vertical ratio]
---
**Recommended tool:** Nano Banana / Midjourney / DALL-E
**Style keywords:** [keyword list]
**Technical notes:** [generation observations]
For Videos:
## VIDEO PROMPT - [PRODUCT] - [PLATFORM]
### Desktop Version (16:9)
[Complete prompt in English]
### Mobile Version (9:16)
[Adapted prompt for vertical ratio]
---
**Compatible tools:** FreePik AI, Google Veo Flow, Runway ML, Pika Labs
**Suggested duration:** [X seconds]
**Main movement:** [movement type]
**Animation notes:** [tips for best results]
Rules
- ALWAYS in English - All prompts for AI tool compatibility
- ALWAYS ask platform - Desktop, Mobile, or Both before generating
- ALWAYS request reference image - To understand product and extract colors
- ALWAYS generate dual versions - When user chooses "Both"
- ALWAYS use technical terminology - 8K, hyper-realistic, cinematic
- ALWAYS describe lighting - Crucial for premium quality
- NEVER generate generic prompts - Each prompt unique to the product
Validation
Use scripts/prompt-validator.py to check prompt completeness:
python scripts/prompt-validator.py "prompt text" --type image
python scripts/prompt-validator.py "prompt text" --type video