xiaohongshu-downloader
# Xiaohongshu Video Downloader
Download videos from Xiaohongshu (小红书) by extracting the real CDN URL from the page.
## How It Works
Xiaohongshu videos use blob URLs in the browser, which cannot be downloaded directly. This skill:
1. Uses browser automation to load the page (required for JS rendering)
2. Extracts the real CDN video URL from the page HTML
3. Downloads the video with proper headers
## Method 1: Browser Automation (Recommended)
Use the browser tool to extract the video URL:
1. Navigate to the Xiaohongshu page:
```
browser action=navigate targetUrl="https://www.xiaohongshu.com/explore/NOTE_ID"
```
2. Extract the video URL with JavaScript:
```javascript
(() => {
const html = document.documentElement.outerHTML;
const mp4Matches = html.match(/https?:\/\/[^"\s]+\.mp4[^"\s]*/g);
if (mp4Matches) return [...new Set(mp4Matches)];
return null;
})()
```
3. Download with curl:
```bash
curl -L -o output.mp4 "<VIDEO_URL>" \
-H "User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36" \
-H "Referer: https://www.xiaohongshu.com/"
```
## Method 2: Python Script
For pages that don't require authentication:
```bash
python3 scripts/download_video.py "https://www.xiaohongshu.com/explore/NOTE_ID"
```
Note: The Python script may not work for all pages due to:
- JavaScript rendering requirements
- Authentication/login requirements
- Rate limiting
## Notes
- Videos are typically hosted on `xhscdn.com` CDN
- User-Agent and Referer headers are required to avoid 403 errors
- Output directory: `~/Downloads/xiaohongshu/`
- Video URLs follow pattern: `https://sns-video-*.xhscdn.com/stream/.../*.mp4`
标签
skill
ai