Read WeChat Official Account Articles

Trigger Words

• /read-gzh <link>

• /read-gzh <link1> <link2> ...

  (Batch processing)
• "Help me read this official account article"
• "Summarize this article"

Features

1. Article Capture — Title, author, main text, images
2. Image Recognition — Download images, extract text/information from images
3. Structured Summary — Core viewpoints, key information, golden sentences extraction
4. Batch Processing — Support processing multiple links at once

Execution Process

Step 1: Capture Article

scripts/fetch_wechat_article.py

• Title, author
• Full text (graphic article) or abstract (Video Account)
• All image links

python3 .claude/skills/read-gzh/scripts/fetch_wechat_article.py "<WeChat Official Account link>"

Step 2: Download and Recognize Images

1. Create a temporary directory to store images
2. Download all images
3. Read each image and extract text or key information from it
4. Incorporate image content into the summary

# Download image
curl -o /tmp/article_img_1.jpg "<image link>"

Step 3: Generate Summary

# 📄 Article Summary

## Basic Information
- **Title**: xxx
- **Author**: xxx
- **Type**: Graphic Article / Video Account
- **Images**: x pieces

## Core Viewpoints
1. xxx
2. xxx
3. xxx

## Key Information
- xxx
- xxx

## Golden Sentences Excerpt
> "xxx"
> "xxx"

## Image Content
- Image 1: xxx
- Image 2: xxx

## Reflections/Iteration Points
- What insights does this give me?
- What can be learned from this?

Batch Processing

# 📚 Batch Article Summary

## Article 1: xxx
[Summary content]

## Article 2: xxx
[Summary content]

---

## Summary Insights
- Common Theme: xxx
- Differences: xxx

Notes

1. Video Account Articles: Only the introduction can be obtained, video content cannot be accessed
2. Image-only Articles: Focus on recognizing text in images
3. Content requiring login: Cannot be accessed