AgentSkillsCN

docai-web2md

适用于需要将网页转换为 Markdown 格式时使用。采用基于优先级的方法（Jina Reader → Firecrawl → Python），轻松应对静态博客、动态页面、社交媒体以及 arXiv 论文。

SKILL.md
--- frontmatter
name: docai-web2md
description: Use when needing to convert web pages to Markdown format. Uses priority-based approach (Jina Reader → Firecrawl → Python) to handle static blogs, dynamic pages, social media, and arXiv papers.

docai:web2md

Overview

Converts web pages to Markdown with automatic priority-based method selection:

  1. •Jina Reader API (fastest, zero install) - except WeChat
  2. •Firecrawl API (advanced crawling)
  3. •Python fallback (when APIs unavailable)

Special handling: WeChat articles → Python directly (best results)

Use when: User says "convert to Markdown", "转成 Markdown", "网页转 Markdown"

Quick Action (For Claude Code)

When user provides a URL to convert:

bash
# Method 1: Jina Reader (recommended - no install)
curl https://r.jina.ai/<URL>

# Method 2: Python script (if Jina unavailable)
python skills/docai-web2md/tools/convert.py <URL>

Prerequisites (Python Method Only)

bash
# Option 1: uv (recommended)
uv pip install --system requests beautifulsoup4 markdownify pymupdf

# Option 2: pip
pip install requests beautifulsoup4 markdownify pymupdf

Usage Examples

bash
# Basic conversion
python skills/docai-web2md/tools/convert.py https://www.breezedeus.com/article/ai-agent-context-engineering

# Save to file
python skills/docai-web2md/tools/convert.py https://www.breezedeus.com/article/ai-agent-context-engineering -o article.md

# arXiv paper (auto HTML priority)
python skills/docai-web2md/tools/convert.py https://arxiv.org/abs/2601.04500v1

# Force Python method
python skills/docai-web2md/tools/convert.py <URL> --use-python

Supported Platforms

  • •✅ Static blogs & documentation (Jina/Python)
  • •✅ React/Vue dynamic pages (Python)
  • •✅ WeChat articles (微信公众号) - Python direct
  • •✅ X.com / Twitter (Python)
  • •✅ arXiv papers (HTML → PDF fallback)
  • •✅ Medium/Substack (Python)

Common Issues

  • •Missing dependencies: Install with uv pip install --system requests beautifulsoup4 markdownify pymupdf
  • •arXiv PDF: Requires pymupdf for PDF extraction
  • •Dynamic pages: May need Python method if Jina fails

Related

  • •docai-web2summary: For AI-powered summaries (conversion + analysis)