with one click
affiliate-marketing
// Scrape Amazon product details via browser and generate platform-optimized promotional content (Twitter/X, Instagram, WeChat) using LLM. No third-party API needed — browser-based extraction only.
// Scrape Amazon product details via browser and generate platform-optimized promotional content (Twitter/X, Instagram, WeChat) using LLM. No third-party API needed — browser-based extraction only.
从文稿内容生成专业 PPTX 演示文稿,支持用户提供的模板/参考图风格提取、AI 配图(siliconflow-img-gen)和素材图库(pexels/pixabay)。纯 Python 生成,无需 Google/ChatGPT API。
通过自媒体平台搜索内容创作者,按预设关键词和判定标准筛选潜在客户。用于 HEARTBEAT 定时任务。
维护 IR(投资人关系专员)的 SQLite 追踪数据库,记录投资人档案和接触历史,避免重复接触,跟踪进展状态。
Find local businesses on Google Maps, extract contact emails from their websites, generate personalized outreach emails with LLM, and send via SMTP. Full pipeline for B2B cold email campaigns.
Search and download copyright-free images and video clips from Pexels API. Supports both photo search (--type image) and video search (--type video, default).
Search and download copyright-free images and video clips from Pixabay API. Alternative to pexels-footage when Pexels has no suitable results. Supports both photo search (--type image) and video search (--type video, default).
| name | affiliate-marketing |
| description | Scrape Amazon product details via browser and generate platform-optimized promotional content (Twitter/X, Instagram, WeChat) using LLM. No third-party API needed — browser-based extraction only. |
| metadata | {"openclaw":{"emoji":"🛒","always":false}} |
Use this skill when:
Prerequisites: Browser session must be able to access amazon.com (international) or amazon.cn (China).
1. Navigate to https://www.amazon.com (warmup — wait for homepage to load)
2. Navigate to the affiliate product URL provided by the user
3. Wait 2–3 seconds for full page render
4. Extract the following elements:
Title:
- Find element with id="productTitle"
- text().strip()
Price:
- Try id="priceblock_ourprice" first
- Fallback: find element with class containing "a-price-whole"
- Fallback: find element with class "a-offscreen" (screen-reader price)
Rating:
- Find id="acrPopover", read the title attribute (e.g., "4.5 out of 5 stars")
- OR find element with class "a-icon-alt"
Review Count:
- Find id="acrCustomerReviewText" → text (e.g., "1,234 ratings")
Feature Bullets:
- Find id="feature-bullets"
- Extract all <li> text items (skip "Make sure this fits" disclaimer)
- Keep top 3–5 most relevant features
Main Image URL:
- Find id="landingImage" or id="imgBlkFront"
- Read the "src" or "data-old-hires" attribute
5. If any element is missing, skip it and continue with available data
6. If CAPTCHA or "To discuss automated access" appears → stop and report to user
Verify the URL already contains the affiliate tag (?tag= or &tag=). If it doesn't:
?tag={associate_tag} to the product URL (clean URL: https://www.amazon.com/dp/{ASIN}?tag={tag})Use LLM to generate platform-specific content. Call the LLM with the product data collected:
Prompt: "Write a promotional tweet for this Amazon product. Include 3 relevant hashtags.
Under 280 characters including the link placeholder [LINK].
Product: {title}
Price: {price}
Key features: {top_3_features}
Tone: enthusiastic but honest
Return ONLY the tweet text."
After generation, replace [LINK] with the actual affiliate URL.
Prompt: "Write an Instagram caption for this Amazon product.
Structure: 1 hook sentence + 3-4 feature highlights as emoji bullet points + CTA + hashtags (10-15 tags at the end).
Product: {title}, Price: {price}, Rating: {rating}
Features: {features}
Return ONLY the caption."
twitter-post skillinstagram-post skill with the product main image URL| Issue | What to do |
|---|---|
| Price shows "$0.00" or missing | Look for "See price in cart" — report to user, use "See price in cart" as placeholder |
| Feature bullets not found | Use product description instead (id="productDescription") |
| Page redirects to login | Amazon session issue — try without warmup or report to user |
| Different page layout (A+ content) | Extract from title + description only |
| CAPTCHA | Stop immediately, report to user |
browser-guide skill to perform browser actions