WeChat Article Pipeline

Problem

WeChat articles are often laid out as mixed text and images, and copying plain text loses headings, lists, tables, code blocks, and image positions. To archive a mp.weixin.qq.com article as an offline-readable, printable document, an engineer usually has to open the page, identify the article body, download images, adjust typography, and render a PDF manually. This pipeline automates that repetitive workflow.

How It Works

The skill turns the process into a structured workflow:

• Import and verify: After receiving an article link, it writes the URL into an IMA knowledge base and reads media information to confirm the article entry point.
• Fetch and download: It uses WebFetch to extract the full body, image URLs, lists, tables, and code blocks; images are downloaded into a local working directory.
• Clean and typeset: It removes author metadata, source labels, ads, preview tags, and follow-up prompts, then generates a styled HTML document.
• PDF output: It renders the HTML into an A4 PDF using Playwright/Chromium and delivers the file.

Limits

The skill is best suited for WeChat Official Account article links and requires the relevant IMA setup plus a browser rendering environment. Image downloads may hit anti-hotlink protection, requiring a retry with the https://mp.weixin.qq.com/ Referer. Color emoji may fall back to plain symbols, and the generated PDF usually renders Chinese visually correctly even if text extraction is affected by CID encoding.