xhs-video-downloader_skill

Overview

This skill downloads videos from Xiaohongshu pages by extracting the real CDN video URL and saving the file locally. It handles pages that rely on JavaScript-rendered blob URLs by using browser automation to discover the direct MP4 link. The skill provides command-line and script options to fetch the video with proper headers to avoid access errors.

How this skill works

The skill loads the Xiaohongshu page with a headless browser so JavaScript can render and produce the real media links. It scans the page HTML for direct .mp4 CDN URLs, deduplicates results, and returns the canonical video URL. Finally, it downloads the video using an HTTP client with User-Agent and Referer headers to prevent 403 errors and saves files to a target folder.

When to use it

• You want to save a Xiaohongshu page video for offline viewing or backup.
• The browser shows a playable video but the page uses blob URLs that cannot be downloaded directly.
• You have a Xiaohongshu post URL and need the direct MP4 CDN link.
• You need to archive videos from public Xiaohongshu posts for research or backup.
• Automated downloads where authentication is not strictly required or provided.

Best practices

• Use the browser automation method for best success on JS-heavy pages.
• Include a realistic User-Agent and set Referer to https://www.xiaohongshu.com/ when downloading.
• Respect copyright and terms of service; download only content you are allowed to save.
• Check for rate limiting and avoid bulk scraping without permission.
• Save files to a clear output directory (default: ~/Downloads/xiaohongshu/).

Example use cases

• Extract the MP4 URL from a Xiaohongshu post that uses blob playback in the browser.
• Download a single public Xiaohongshu video for offline review or presentation.
• Archive videos from an account for research while preserving original filenames.
• Integrate into a backup workflow that saves publicly available media assets.