SKILL.md - China Shopping

Overview

China Shopping is an OpenClaw skill that recommends the best Chinese shopping websites based on product categories. This v1.0.1 version provides straightforward recommendations for common product types like electronics, clothing, groceries, and more.

Security Updates (v1.0.1)

• Fixed potential command injection vulnerabilities in shell scripts
• Added input validation and sanitization for product names
• Used --arg parameter in jq queries to safely pass variables
• Added length limits and dangerous character filtering

Description

When users input a product name (e.g., "手机", "衣服"), the skill recommends the most suitable Chinese e-commerce platforms for purchasing that product, helping users find the best shopping destinations quickly. For live page verification or cross-platform product checks, pair it with the shared browser-commerce-base workflow.

Trigger Phrases & Usage

Trigger Phrases

• "china-shopping"
• "中国购物"
• "推荐购物网站"
• "买[产品]去哪"
• "哪里买[产品]"

Command Usage

# Basic recommendation
china-shopping recommend "手机"

# Alternative syntax
china-shopping 推荐 "衣服"

# Show all categories
china-shopping categories

# Help
china-shopping help

In Chat Context

• "帮我推荐买手机的网站"
• "买衣服应该去哪个网站？"
• "中国购物推荐：笔记本电脑"

Installation & Dependencies

This skill requires no external dependencies. It's a pure CLI tool with a simple data file.

Files

• china-shopping (executable script)
• categories.json (product-category mapping)
• recommendations.json (category-website recommendations)

Permissions

• Read access to data files
• Write access to logs (optional)

Browser Workflow Upgrade

When users move from “推荐平台” to “帮我实际看看页面/价格/店铺”, switch to the shared browser-commerce-base workflow:

• use this skill to narrow candidate platforms first
• then inspect the live site with openclaw browser on public pages
• use user only when logged-in carts, coupons, or member prices matter
• compare candidate platforms with the common commerce extraction schema

Recommended next step after platform recommendation:

1. pick 2–3 candidate sites
2. open live search/product pages
3. extract price, seller/store type, shipping, coupon, and risk signals
4. return a ranked comparison with screenshots when useful

How It Works

1. Input Parsing: Takes a product name as input
2. Category Mapping: Maps product to predefined categories (electronics, clothing, groceries, etc.)
3. Recommendation Lookup: Returns recommended websites for that category
4. Output Formatting: Presents results in a user-friendly format

Example Output

$ china-shopping recommend "手机"
📱 手机 (Electronics) 推荐购物网站：
1. 京东 (JD.com) - 正品保障，物流快
2. 天猫 (Tmall.com) - 品牌官方旗舰店
3. 苏宁易购 (Suning.com) - 家电数码专业
4. 小米商城 (Mi.com) - 小米生态链产品

💡 提示：购买电子产品建议选择官方旗舰店或自营平台，注意查看用户评价和售后服务。

Configuration

No configuration required for basic usage.

Advanced Features (Future)

• Price comparison across platforms
• User preference learning
• Seasonal/realtime recommendations
• Coupon/price alert integration

Notes

• This is v1.0.0 simple version
• Recommendations are based on general market consensus
• Actual shopping experience may vary
• Always verify seller reputation before purchase

Author & License

OpenClaw Skill - MIT License