name: wechat-article
description: Fetch and extract content from WeChat public account articles. Use when user shares a WeChat URL or asks to read a WeChat article.
user_invocable: false

WeChat Article Extractor

Extract content from WeChat public account articles (mp.weixin.qq.com).

When to Use

• User shares a WeChat article URL
• User asks to read/summarize a WeChat article
• Need to extract content from mp.weixin.qq.com

URL Formats

WeChat articles have two URL formats:

Extraction Workflow

Step 1: Identify URL Format

import re

def is_short_url(url: str) -> bool:
    """Check if URL is short format (directly usable)."""
    # Short format: /s/ followed by alphanumeric ID (no query params)
    return bool(re.match(r'https?://mp\.weixin\.qq\.com/s/[A-Za-z0-9_-]+$', url))

Step 2: Extract Content

For short URLs — Use wechat.imagenie.us API:

curl -X POST "https://wechat.imagenie.us/extract" \
  -H "Content-Type: application/json" \
  -d '{"url":"https://mp.weixin.qq.com/s/ABC123","format":"markdown"}'

Response:

{
  "success": true,
  "content": "# Article Title\n\nArticle content in markdown..."
}

For long URLs — Convert first:

1. Extract article title from the URL (if available) or fetch page title
2. Search for short URL: WebSearch: "article title" site:mp.weixin.qq.com/s/
3. Use the short URL with wechat.imagenie.us

Step 3: Fallback Options

If wechat.imagenie.us fails:

1. Mirror sites — Search for reposts:
   WebSearch: "article title" site:163.com OR site:zhihu.com OR site:csdn.net

2. Serper scrape — Try mirror URL directly:
   ```bash
   # OpenClaw
   mcporter call serper.scrape url="[mirror URL]"

# Claude Code
mcp__serper__scrape
```

Complete Workflow

User provides WeChat URL
         │
         ▼
   Is short URL?
    (/s/ABC123)
         │
    ┌────┴────┐
   Yes        No
    │          │
    ▼          ▼
  Use API   Search for
  directly  short URL
    │          │
    ▼          ▼
  Success?  Found short URL?
    │          │
   ┌┴┐        ┌┴┐
  Yes No    Yes  No
   │   │     │    │
   ▼   │     ▼    ▼
Return │   Use   Search
content│   API   mirrors
       │     │    │
       └──┬──┘    │
          ▼       ▼
       Search   Scrape
       mirrors  mirror

Example Usage

Input: https://mp.weixin.qq.com/s/nnysYJCNQA-SPUefOPBwZw

Action:

curl -s -X POST "https://wechat.imagenie.us/extract" \
  -H "Content-Type: application/json" \
  -d '{"url":"https://mp.weixin.qq.com/s/nnysYJCNQA-SPUefOPBwZw","format":"markdown"}'

Output: Full article in Markdown format.

API Reference

wechat.imagenie.us

POST /extract body:

{
  "url": "https://mp.weixin.qq.com/s/ABC123",
  "format": "markdown"  // or "json", "html"
}

Headers for GET:
- Accept: text/markdown — Returns Markdown instead of HTML

Error Handling

Limitations

• Only short URLs work with the API
• Some articles may be deleted or restricted
• Rate limits apply to the extraction service
• Mirror sites may have outdated versions

Related

• See web-search skill for mirror site search strategies
• See evaluate-article skill for article analysis workflow