wechat-article-extractor

Category: Docs & Knowledge | Uploader: toto2016 | Downloads: 0 | Version: v1.0（Latest）

Extract full text and images from WeChat public account (微信公众号) articles and save as clean Markdown. Handles WeChat's bot-detection by finding mirror sites automatically. Use when the user shares an mp.weixin.qq.com URL and asks to save, archive, extract, or read the article.

Changelog: Source: GitHub https://github.com/toto2016/wechat-article-extractor

Directory Structure

Current level: tree/main/

📁 scripts/
- 📄 extract_wechat.py 13.2 KB
📁 tests/
- 📄 test_triggers.md 932 B
📄 .gitignore 44 B
📄 CHANGELOG.md 680 B
📄 README.md 4.8 KB
📄 SKILL.md 5.2 KB
📄 skill.yml 693 B

SKILL.md

---
name: wechat-article-extractor
description: >
  Extract full text and images from WeChat public account (微信公众号) articles
  and save as clean Markdown. Handles WeChat's bot-detection by finding mirror
  sites automatically. Use when the user shares an mp.weixin.qq.com URL and asks to
  save, archive, extract, or read the article.
triggers:
  - mp.weixin.qq.com
  - wechat article
  - 微信公众号
  - extract wechat
  - save wechat article
  - archive wechat
  - 提取公众号文章
  - 保存公众号文章
---

# WeChat Article Extractor

Extract WeChat public account articles to clean Markdown.

WeChat blocks headless browsers (环境异常 CAPTCHA) and direct HTTP requests often get empty JS-rendered pages. This skill works around these limitations by automatically finding mirror copies on aggregator sites, then extracting clean content.

## Required Agent Capabilities

Your AI agent/bot must support these capabilities for this skill to work:

| Capability | Purpose |
|:---|:---|
| **Web Fetch** | Direct article fetch attempt, read mirror pages |
| **Web Search** | Discover mirror copies when direct access is blocked |
| **Shell Exec** | Run `curl` and the Python extraction script |
| Browser automation *(optional)* | Chrome Relay fallback for unmirrored articles |

If your agent uses different tool names (e.g., `read_url` instead of `web_fetch`, `run_command` instead of `exec`), map accordingly.

## System Dependencies

| Dependency | Version | Purpose |
|:---|:---|:---|
| Python | 3.8+ | HTML-to-Markdown extraction script |
| curl | any | Download mirror pages |

No third-party Python packages needed — the script uses only the standard library.

## Workflow

### Step 1 — Try direct fetch

Fetch the WeChat article URL directly.

```
web_fetch("https://mp.weixin.qq.com/s/...")
```

**Success check:** Response contains real article paragraphs (not just nav/footer/CAPTCHA). Typically >500 chars of actual content.

→ If success: format content with header template (see below), done.

→ If failure (empty, CAPTCHA, 环境异常): proceed to Step 2.

### Step 2 — Extract article metadata

From the URL or any partial content, identify:
- Article title
- Author / account name

If metadata is unavailable, ask the user for the article title.

### Step 3 — Search for mirror copies

```
web_search("<article title> <author/account name>")
```

**Mirror site priority** (ranked by reliability):

1. **53ai.com** — full content, reliable formatting
2. **mp.ofweek.com** — tech articles
3. **juejin.cn** — developer content
4. **woshipm.com** — product/business content
5. **36kr.com** — tech/business news

If title is unknown, try: `web_search("site:53ai.com <keywords from URL>")`

### Step 4 — Download mirror and extract

Download the mirror page:

```bash
curl -s -L "<mirror_url>" -o /tmp/wechat-article.html
```

Verify file size > 10KB (smaller usually means redirect/error page).

Run the extraction script:

```bash
python3 <skill_dir>/scripts/extract_wechat.py /tmp/wechat-article.html /tmp/<output-filename>.md --source "<original_wechat_url>"
```

Replace `<skill_dir>` with the path where this skill is installed.

### Step 5 — Verify output quality

Check the output file:
- Has a title (not "WeChat Article")
- Has multiple paragraphs of real content
- Images have valid URLs
- No excessive HTML artifacts

If truncated or garbled, try a different mirror site (return to Step 3).

### Step 6 — Deliver to user

Report:
- File saved at: `<path>`
- Title: `<title>`
- Size: `<char count>` characters
- Image count: `<N>` images

## Markdown Header Template

Every extracted article must include this header:

```markdown
# <title>

**作者：** <author>
**来源：** 微信公众号「<account_name>」
**日期：** <date>
**原文：** <original_wechat_url>

---

> **摘要：** <1-2 sentence summary generated from content>

---
```

Fields that cannot be determined should be omitted.

## Scope

**Handles:**
- Extracting article text, images, and metadata from WeChat article URLs
- Finding mirror copies when direct access is blocked
- Converting HTML to clean Markdown
- Saving output as `.md` files

**Does NOT handle:**
- Publishing or syncing to note-taking apps
- Batch extraction of multiple articles
- WeChat login, authentication, or account management
- Translating article content

## Error Handling

| Problem | Detection | Action |
|:---|:---|:---|
| WeChat blocks access | Empty response or "环境异常" | Search for mirrors (Step 3) |
| No mirrors found | Search returns 0 relevant results | Tell user to manually save the page |
| Mirror content truncated | Output < 1000 chars when original is long | Try next mirror site |
| Script extraction fails | Python error or empty output | Fall back to reading mirror URL directly |
| Images broken | Image URLs return 404 | Note in output; images may expire |

## Notes

- Mirror sites typically publish within minutes of the original
- WeChat image URLs from mirrors are generally valid and render in Markdown viewers
- The `· · ·` section dividers are WeChat style — preserve them
- For very long articles (>50K chars), the script handles them fine but web_fetch may truncate
- Very new articles (< 1 hour old) or niche content may not have mirrors yet

Comments 0

Latest | Hot

Please login before commenting.

Loading comments...

wechat-article-extractor

Directory Structure

SKILL.md

Report

Notice