SKILL

default

markdown

---
name: wechat-mp-reader
description: Read WeChat official account articles. Use the built-in browser tool to open the page and extract body text. Always append ?scene=1 to the URL.
---
# WeChat Article Reader

## URL Normalization

**Critical**: The URL must end with `?scene=1` (not `&scene=1`), otherwise a CAPTCHA will be triggered.

Rules:
- No query params → append `?scene=1`
- Has existing `?` params → parse and rewrite query to include `scene=1`

## Steps

1. Open the page with `browser open "<url>?scene=1"`
2. Wait for content with `browser wait "#js_content" --load networkidle`
3. Extract body text with `browser evaluate --fn "() => document.querySelector('#js_content')?.innerText || document.querySelector('.rich_media_content')?.innerText || document.body.innerText"`
4. Return plain text content
5. Close the tab with `browser close <tabId>`

## Troubleshooting

- CAPTCHA → verify the URL has `?scene=1`
- Empty content → page may not have fully loaded, retry `browser wait`
- Deleted article → the page will display a notice

Uploading file...

Edit message:

Cancel

Editing SKILL