product-page-designer

SKILL前端设计
前端设计 未标注权限 许可证不明确
安装这个技能

Design 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.

一句话选型结论

按资料字段整理

product-page-designer是前端设计类的 Agent Skill,安装方式未标注,权限级别未标注,许可证不明确,有可直接复制的安装命令。技能本质是给模型读的说明书,效果取决于模型与客户端;装完先用一个小任务验证,再放进正式流程。

分类前端设计 安装方式未标注 权限级别未标注权限 许可证许可证不明确
累计安装 未标注 本周 未标注
GitHub 星标 未标注 仓库 未标注
下载量 未标注 技能市场下载数
权限级别 未标注权限 安装前请确认风险范围
许可证 许可证不明确 未说明

安装

未标注
在支持 Skills 的客户端里执行,装完问一句「你有哪些技能」验证
skillhub install user_533ec4e2/product-page-designer
  1. 1
    确认客户端支持

    确认你的客户端支持 Skills(Claude、Codex、Cursor 等),并升级到较新的版本。

  2. 2
    拿到技能文件

    执行上面的命令安装;也可以把仓库里的 SKILL.md 手动放进客户端的 skills 目录。

  3. 3
    刷新技能列表

    重新打开客户端,或刷新技能列表,让新放进来的技能被加载。

  4. 4
    验证是否生效

    问一句「你有哪些技能」,确认列表里出现它,再拿一个小任务试跑。

技能档案

前端设计
字段来自技能仓库与公开索引,中文说明由本站标注
作者 / 维护方user_533ec4e2仓库 未标注 · 版本 未标注
宿主客户端未标注技能可被这些客户端加载
安装方式未标注安装包 community
权限级别未标注权限安装前请确认风险范围
许可证未说明许可证不明确,商用前请自行确认条款
仓库可用性available技能仓库当前是否可访问
最近核验未标注上游最后核验时间 · 数据更新 1791623674
正文体量10.0 KBSKILL.md 原文大小,越大说明说明书写得越细

典型使用场景

3 类
按分类整理的常见用法,具体以 SKILL.md 为准
出页面与组件

按描述生成页面结构和组件草稿。

调整样式与布局

改间距、配色与响应式布局。

生成设计说明

把设计稿整理成开发能接的说明。

选择建议

装之前先看这几条
按条目字段给出的判断,不替你决定用不用

值得选的地方

  • 有可直接复制的安装命令,不用自己翻文档找入口。

先想清楚的地方

  • 没有标注权限级别,实际能做什么要装好后确认。
  • 许可证不明确,商用或二次分发前先到仓库确认条款。
  • 没有做站内核验,内容以作者仓库为准。
  • 技能只是说明书,实际效果取决于模型与客户端,装完先用小任务试一遍。

更适合

  • 需要在前端设计场景里给模型补一套做法的人。
  • 愿意先读一遍 SKILL.md、确认它会让模型做什么的人。

不太适合

  • 对许可证要求很严、必须明确可商用的人。
  • 只想装完就自动变强、不打算看说明书的人。

同类对比

6 个同类
同分类里挑最接近的几条,字段以各自详情页为准
技能权限级别安装方式累计安装宿主客户端安装命令
product-page-designer当前 未标注权限 未标注 未标注 未标注 有
前端设计 只读 GitHub 仓库 882.3 万 Codex、Claude、Cursor 有
Vercel 组合模式 联网访问 GitHub 仓库 385.9 万 Codex、Claude、Cursor 有
Vercel React native skills 前端 只读 GitHub 仓库 246.9 万 Codex、Claude、Cursor 有
adapt 搜索 浏览器控制 GitHub 仓库 202.7 万 Codex、Claude、Cursor 有
使用我的浏览器 浏览器控制 GitHub 仓库 184.8 万 Codex、Claude、Cursor 有

SKILL.md 全文

10.0 KB
模型实际读到的说明书原文,来自技能仓库

Product Page Designer Skill

Generate complete, conversion-optimized product detail pages for consumer electronics that meet international brand standards (Apple, Nothing, Sonos level).

Workflow

Step 1: Gather Product Info

Ask the user for:
- Product name (e.g., "Bluetooth Earbuds", "Smart Watch")
- Product category (headphones, earbuds, speaker, watch, etc.)
- Key specs (battery, connectivity, waterproof rating, etc.) — if unknown, use standard industry specs for the category
- Target price range (budget/mid/premium) — affects copy tone
- Brand voice preference (minimalist, energetic, technical, lifestyle) — default: minimalist

Step 2: Generate Product Page Content

Create a complete product detail page with these sections:

A. Hero Section

  • Product name (original, no brand infringement)
  • One-line slogan (max 5 words, punchy)
  • CTA with price
  • Visual direction: product on dark background, dramatic lighting, no clutter

B. Key Features (4-6 points)

Each feature card needs:
- Icon suggestion
- Title (2-4 words, benefit-driven)
- One-line description (max 20 words)
- Image concept description

Standard features to consider (pick 4-6 relevant):
- ANC / Sound isolation
- Hi-Res / Audio quality
- Battery life
- Connectivity (Bluetooth version, multipoint)
- Latency / Gaming mode
- Waterproof rating
- Comfort / Fit
- Smart features (app, EQ, spatial audio)

C. Lifestyle Showcase (4 scenes)

Pick scenes relevant to the product category. Each needs:
- Scene name
- Image style description
- Photography direction
- Overlay copy (headline + subheadline)

Common scenes: Office/Work, Travel, Fitness/Outdoors, Relax/Home

D. Product Details

Three paragraphs:
1. Introduction (what it is, why it matters)
2. Experience (how it feels to use)
3. Why buy (differentiators, warranty, trial)

E. Specifications

Markdown table with standard specs for the category.

F. What's in the Box

Standard packaging list.

G. FAQ

8-10 questions covering: calls, compatibility, charging, returns, app, special features.

Step 3: Generate AI Image Prompts

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:
1. ALWAYS append no visible branding, no text, no logos to every prompt
2. ALWAYS include professional commercial product photography as style anchor
3. ALWAYS specify aspect ratio and resolution
4. Product must be described generically (matte black, premium build) — never use real brand names
5. For lifestyle shots, describe the person/product relationship, not the person's identity

Step 4: Generate Images with WaveSpeed

Use wavespeed skill to generate all images:

  1. Hero + Lifestyle images → google/nano-banana-2/text-to-image at 2K resolution
  2. Feature cards → google/nano-banana-2/text-to-image at 2K resolution
  3. Brand removal (if AI generates fake logos) → bytedance/seedream-v4.5/edit

Common pitfall: AI models love to generate fake brand names on products ("SONY", "Bose", "Razer", etc.) — expect 10-30% of images to need cleaning. Always:
- Add no visible branding, no text, no logos, unbranded, generic design to every prompt
- After generation, check each image with vision_analyze for brand text
- If found, use bytedance/seedream-v4.5/edit to remove it (~$0.04/edit)
- Budget extra time and credits for the brand removal pass

Step 5: Quality Review Checklist

Before delivering, verify:

  • [ ] All copy is original (no brand infringement)
  • [ ] Slogan is ≤ 5 words and memorable
  • [ ] Feature descriptions are benefit-driven, not spec-driven
  • [ ] Lifestyle copy matches the scene emotion
  • [ ] All images are unbranded (no text, no logos)
  • [ ] Image style is consistent across all shots
  • [ ] Specs table is complete and accurate
  • [ ] FAQ covers real customer concerns
  • [ ] Overall tone matches the chosen brand voice

Step 6: Deliver

Output as a single comprehensive Markdown document with:
1. Complete product page content (ready for web developers)
2. All image prompts organized by section
3. Generated image URLs (if wavespeed was used)
4. Layout suggestions for each section

Product Category Templates

Headphones / Earbuds

Standard specs: Bluetooth 5.x, ANC, 20-60h battery, IPX4-IPX7, 32Ω impedance, USB-C charging, multipoint, companion app

Bluetooth Speakers

Standard specs: Bluetooth 5.x, 10-30h battery, IPX7, 10-40W output, stereo pairing, USB-C, built-in mic

Smart Watches

Standard specs: AMOLED display, heart rate, SpO2, GPS, NFC, 5-14 day battery, IP68, smartphone notifications, sleep tracking

Wireless Earbuds (TWS)

Standard specs: Bluetooth 5.x, ANC, 6-8h + case 24-32h total, IPX4-IPX5, touch controls, spatial audio, fast charge, wireless charging case

Naming Convention

When generating product names, use this pattern:
- Brand prefix: Abstract, tech-sounding (AURA, NOVA, PULSE, VELO, ORBIT, LUMA, ZEN, CORE, FLUX, NEXUS)
- Product type: Pro, Ultra, Air, Lite, Max, SE
- Number: X1, Z3, S2 (optional, adds specificity)

Examples: AURA Pro X1, NOVA Ultra S2, PULSE Air SE

Tone Guidelines

Minimalist (Apple/Nothing style)

  • Short sentences. Period. Like this.
  • Numbers speak louder than adjectives
  • "40 hours. Not 39. Not 41."
  • Focus on feeling, not features

Energetic (Nike/Adidas style)

  • Action verbs. Movement. Energy.
  • "Push harder. Go further."
  • Emojis in social copy only
  • Bold claims, backed by specs

Technical (Sony/Bose style)

  • Precise language. Measurements. Certifications.
  • "40mm custom neodymium drivers"
  • "Hybrid ANC with 45dB reduction"
  • Data-driven comparisons

Output Format

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]

Implementation Notes

Next.js Product Page Pattern

When implementing in a Next.js project:
- Create route at /product/[slug]/page.tsx
- Use ScrollReveal component for scroll animations
- Use Tailwind CSS with custom nav-* color tokens (defined in globals.css)
- Hero: 2-column grid, product image + CTA
- Features: 3-column grid with image overlay cards
- Lifestyle: alternating 2-column blocks (image left/right, use lg:[&>*:nth-child(1)]:order-2 for even items)
- Specs: table with alternating row styling (use divs instead of table for better Tailwind control)
- FAQ: accordion pattern with React useState
- Sticky bottom bar for CTA
- All images should be generated via wavespeed and hosted on CDN (cloudfront)
- Use "use client" directive on product pages (needed for interactive FAQ/cart)
- Move interactive state (useState) to the top-level component, not sub-components, to avoid "Cannot find name" errors
- Toast notifications: create DOM element dynamically rather than relying on global functions

Brand Removal Workflow

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.

End-to-End Session Pattern (Validated)

This workflow was tested end-to-end in a real session and proven to work:
1. Design product page content in Markdown → 2. Write image prompts to files → 3. Generate images in parallel batches (8-10 at a time) using background mode → 4. Evaluate images with vision_analyze → 5. Fix branded images with seedream-v4.5/edit → 6. Build Next.js product page → 7. Deploy to Netlify

常见问题

5 问
答案由本站依据条目字段整理
product-page-designer怎么安装?

在支持 Skills 的客户端里执行:skillhub install user_533ec4e2/product-page-designer。安装方式为未标注。装完刷新技能列表或重开客户端。

product-page-designer支持哪些客户端?

资料没有标注宿主客户端,但只要客户端支持 Skills(Claude、Codex、Cursor 等),一般都能加载。

product-page-designer权限高吗,安全吗?

权限级别为未标注权限。安装前先看它会读取或写入哪些内容。

product-page-designer收费吗?

本站不代收费用,也没有标注具体价格。是否收费取决于技能本身和所用的模型或客户端,正式使用前按作者说明确认。

product-page-designer用什么许可证?

许可证标注为 未说明(许可证不明确)。商用或二次分发前请到仓库确认条款原文。

热门专题

按场景挑技能
全部专题 →

技能内容以作者仓库为准;分类、权限级别与许可证状态的中文说明为中文整理,最近核验 未标注。