Install
openclaw skills install @johnsonbuilds/product-page-designerDesign and generate international-brand-quality product detail pages for cross-border e-commerce. Generates complete product page content (Hero, Features, Lifestyle, Specs, FAQ) plus AI image prompts and visual assets for any consumer electronics product. Output meets Apple/Nothing brand standards.
openclaw skills install @johnsonbuilds/product-page-designerGenerate complete, conversion-optimized product detail pages for consumer electronics that meet international brand standards (Apple, Nothing, Sonos level).
Ask the user for:
Create a complete product detail page with these sections:
Each feature card needs:
Standard features to consider (pick 4-6 relevant):
Pick scenes relevant to the product category. Each needs:
Common scenes: Office/Work, Travel, Fitness/Outdoors, Relax/Home
Three paragraphs:
Markdown table with standard specs for the category.
Standard packaging list.
8-10 questions covering: calls, compatibility, charging, returns, app, special features.
For EVERY image in the page, create a prompt following this template:
[Shot type] of [product description] in [setting/background],
[lighting description],
[composition notes],
[style keywords]:
professional commercial product photography, minimalist composition, clean studio lighting, modern aesthetic, premium feel, photorealistic, high-end consumer electronics, no visible branding, no text, no logos
Critical rules for image prompts:
no visible branding, no text, no logos to every promptprofessional commercial product photography as style anchorUse wavespeed skill to generate all images:
google/nano-banana-2/text-to-image at 2K resolutiongoogle/nano-banana-2/text-to-image at 2K resolutionbytedance/seedream-v4.5/editCommon pitfall: AI models love to generate fake brand names on products ("SONY", "Bose", "Razer", etc.) — expect 10-30% of images to need cleaning. Always:
no visible branding, no text, no logos, unbranded, generic design to every promptvision_analyze for brand textbytedance/seedream-v4.5/edit to remove it (~$0.04/edit)Before delivering, verify:
Output as a single comprehensive Markdown document with:
Standard specs: Bluetooth 5.x, ANC, 20-60h battery, IPX4-IPX7, 32Ω impedance, USB-C charging, multipoint, companion app
Standard specs: Bluetooth 5.x, 10-30h battery, IPX7, 10-40W output, stereo pairing, USB-C, built-in mic
Standard specs: AMOLED display, heart rate, SpO2, GPS, NFC, 5-14 day battery, IP68, smartphone notifications, sleep tracking
Standard specs: Bluetooth 5.x, ANC, 6-8h + case 24-32h total, IPX4-IPX5, touch controls, spatial audio, fast charge, wireless charging case
When generating product names, use this pattern:
Examples: AURA Pro X1, NOVA Ultra S2, PULSE Air SE
Always deliver in this structure:
# [Product Name] — Product Detail Page
## Brand Style
[Tone selected]
## 1. Hero Section
- Name: [product name]
- Slogan: [tagline]
- CTA: [button text]
- Visual: [description]
## 2. Key Features
| # | Feature | Title | Description | Image Concept |
|---|---------|-------|-------------|---------------|
| 1 | ... | ... | ... | ... |
## 3. Lifestyle Showcase
### Scene 1: [Name]
- Copy: [headline + subheadline]
- Image style: [description]
- Photography: [direction]
...
## 4. Product Details
[Introduction paragraph]
[Experience paragraph]
[Why buy paragraph]
## 5. Specifications
| Spec | Value |
|---|---|
## 6. What's in the Box
- Item 1
- Item 2
## 7. FAQ
**Q:** ...
**A:** ...
## 8. Image Prompts
[All prompts organized by section, ready for wavespeed]
When implementing in a Next.js project:
/product/[slug]/page.tsxScrollReveal component for scroll animationslg:[&>*:nth-child(1)]:order-2 for even items)"use client" directive on product pages (needed for interactive FAQ/cart)When AI generates brand text on product images (very common — expect 10-30% of images to need cleaning):
Prevention: Always append no visible branding, no text, no logos, unbranded, generic design to every product prompt.
Detection: After generation, check each image with vision_analyze for brand text before integrating into the page.
Fix (use bytedance/seedream-v4.5/edit — cheapest at $0.04/edit, strongest prompt adherence):
# 1. Upload original
URL=$(npx @wavespeed/cli upload ./generated.png --json | jq -r .url)
# 2. Edit to remove branding
PROMPT="Remove all text and brand logos from the image. Keep everything else identical — same lighting, pose, composition, colors. Clean unbranded surface."
npx @wavespeed/cli run bytedance/seedream-v4.5/edit \
-i prompt="$PROMPT" \
-i images="[\"$URL\"]" \
--json --download "./clean-{index}.{ext}"
Budget: ~$0.04/edit. Allow 2-3 edits per image if needed. Total brand-removal cost typically $0.10-$0.30 per product page.
Common hallucinations observed: "SONY", "Bose", "Razer", "AXIS AUDIO", "Sennheiser", "NIKE", "LEICA", "ANALOG VIBRATIONS" — on headphones, mice, vinyl sleeves, and other products.
Proven prompt for brand removal: Write the prompt to a file and read with PROMPT=$(cat ...). Be specific about WHICH object to clean (e.g., "the gaming mouse should have no visible logo", "vinyl record sleeves should have no readable text"). Generic "remove all branding" prompts are less effective than targeted ones.
This workflow was tested end-to-end in a real session and proven to work:
vision_analyze → 5. Fix branded images with seedream-v4.5/edit → 6. Build Next.js product page → 7. Deploy to Netlify