</>MCP Agents Market
Skill

huashu-md-html

by alchaincyf904CSSUpdated 2026-08-25

md/html 双向流水线 · 三个能力一站式:万物→md / md→精美html / html→md。封装 markitdown + Pandoc + html-to-markdown + trafilatura。4 套反 AI slop 主题。Claude Code skill.

Claude CodeCursorCodexOpenClawHermes

huashu-md-html is an agent skill that enables AI assistants to convert documents bidirectionally between markdown, HTML, and other formats. It provides four core capabilities: converting any file (PDF, DOCX, PPTX, images, audio, YouTube videos, web pages) to markdown, rendering markdown into professionally-styled HTML with four publication-grade themes, extracting clean markdown from published HTML pages, and generating publisher-ready DOCX files with automated formatting. The skill wraps markitdown, Pandoc, html-to-markdown, and trafilatura into a unified conversational interface, allowing developers to process documents through natural language commands in their AI coding environment.

Key Features

Universal-to-markdown conversion supporting PDF, DOCX, PPTX, XLSX, EPUB, images, audio, YouTube videos, and web URLs via markitdown
Four anti-AI-slop HTML themes (article, report, reading, interactive) inspired by Tufte, Pentagram, and Medium design systems
Clean HTML-to-markdown extraction with automatic removal of navigation, sidebars, and advertisements using trafilatura
Publisher-grade DOCX generation with automatic cover pages, table of contents, embedded images, headers, footers, and styled tables
Self-contained single-CSS HTML output with no external CDN dependencies
Conversational interface for Claude Code, Cursor, Codex, and other agent clients—no GUI required
Design tokens enforcing typography standards: 17-18px body text, 1.75-1.85 line height for Chinese, JetBrains Mono for code blocks
Dual-path URL processing: markitdown for structured data retention, trafilatura for content-focused extraction

Use Cases

  • 01Converting research PDFs or whitepapers into clean markdown, then rendering them as beautiful HTML reports for publication
  • 02Archiving published blog posts from URLs back into version-controlled markdown source files
  • 03Extracting YouTube video transcripts to markdown and transforming them into essay-format HTML
  • 04Generating publisher-ready DOCX manuscripts from markdown chapters with automatic formatting, image embedding, and table of contents
  • 05Scraping technical documentation or product pages into structured markdown for analysis or reuse
  • 06Comparing multiple HTML theme outputs (article vs. interactive) to choose the best presentation style for long-form content

Related Skills

View more

huashu-md-html — FAQ

What is huashu-md-html?+

huashu-md-html is an agent skill that provides document conversion pipelines for AI coding assistants, enabling conversational transformation between markdown, HTML, DOCX, PDF, and other formats. It wraps four Python tools into a unified interface accessible through natural language commands in Claude Code, Cursor, and similar clients.

How do I install huashu-md-html?+

Install it using the skills CLI with 'npx skills add alchaincyf/huashu-md-html'. You'll also need to install Python dependencies: markitdown, Pandoc, html-to-markdown, trafilatura, python-docx, and Pillow using pip.

Which AI clients work with this skill?+

The skill is agent-agnostic and works with Claude Code, Cursor, Codex, OpenClaw, Hermes, and other clients that support the skills.sh standard. You interact with it through natural language commands rather than a graphical interface.

Does huashu-md-html require API keys?+

No API keys are required. The skill uses local Python tools (markitdown, Pandoc, trafilatura, python-docx) that run entirely on your machine without external API calls.

Is huashu-md-html free to use?+

Yes, it is released under the MIT License for both personal and commercial use at no cost. All underlying tools (markitdown, Pandoc, etc.) are also open-source.

What's the difference between the two URL conversion paths?+

For structured pages (product details, API docs, technical documentation), use markitdown to preserve metadata and field structure. For content-focused pages (blog posts, news articles, essays), use trafilatura to automatically strip navigation, ads, and sidebars, leaving only the main text.

How do I install huashu-md-html?+

Open the source repository on GitHub and follow its README. huashu-md-html is a skill — MCP Agents Market links you directly to the official repo.

Is huashu-md-html free?+

huashu-md-html is an open-source project hosted on GitHub. Check the repository for its license and any usage requirements.

Related searches