wechat-article-extractor_skill

Overview

This skill extracts the full text, images, and metadata from a WeChat public account article URL and saves the result as a clean Markdown file. It handles WeChat bot-detection by searching for mirror copies on aggregator sites and falls back to a browser relay when needed. The output includes a structured header, full article content, and preserved image URLs.

How this skill works

First the skill attempts a direct fetch of the mp.weixin.qq.com URL and checks whether the fetched content contains real paragraphs. If direct access is blocked or returns truncated content, it extracts title and author metadata and searches mirror sites (priority list includes 53ai, ofweek, juejin, woshipm, 36kr). When a mirror is found the page is downloaded and a Python extraction script converts HTML to Markdown. If no mirror is available, a Chrome Extension Relay snapshot is used as a fallback to capture the rendered page.

When to use it

• You have a mp.weixin.qq.com article URL and want a local Markdown copy.
• The article page shows bot-detection, CAPTCHA, or returns empty JS-rendered content.
• You need images preserved with working URLs inside Markdown.
• You want a consistent metadata header (title, author, source, date, original URL).
• You need a single-article archive or backup (not bulk processing).

Best practices

• Provide the exact mp.weixin.qq.com URL; include title if metadata is missing.
• Use the default filename unless you need a specific location; the skill defaults to /tmp/ and kebab-case title.
• If a mirror is not found, open the article in Chrome and enable the browser relay extension to capture the snapshot.
• Verify the saved Markdown: check title, multiple paragraphs, and image URLs after extraction.
• If output looks truncated, ask to try an alternate mirror or provide the article snapshot.

Example use cases

• Save a single WeChat article to Obsidian or a local notes folder as Markdown.
• Archive a technical article that WeChat blocks for long-term reference.
• Extract images and text from a product announcement posted to a WeChat public account.
• Convert a developer tutorial reposted on a mirror site into clean Markdown for offline reading.
• Provide a readable copy for accessibility or content-review purposes.