markdown-utils-mcp

Markdown Utils MCP server. Convert to HTML, extract headings/links, word count.

No external dependencies. Pure Regex processing.

Install

npm install -g markdown-utils-mcp

Usage

Claude Desktop / Cursor / OpenCode

Add to your MCP config:

{
  "mcpServers": {
    "markdown-utils": {
      "command": "npx",
      "args": ["-y", "markdown-utils-mcp"]
    }
  }
}

Tools

to_html — Markdown to HTML

Convert Markdown text to HTML.

Input:

{
  "markdown": "# Hello\n\nThis is **bold** and *italic*.\n\n[Link](https://example.com)"
}

Output:

<h1>Hello</h1>
<p>This is <strong>bold</strong> and <em>italic</em>.</p>
<p><a href="https://example.com">Link</a></p>

extract_headings — Extract Headings

Extract all headings (H1-H6) from Markdown text.

Input:

{
  "markdown": "# Title\n## Section\n### Subsection"
}

Output:

{
  "count": 3,
  "headings": [
    { "level": 1, "text": "Title" },
    { "level": 2, "text": "Section" },
    { "level": 3, "text": "Subsection" }
  ]
}

extract_links — Extract Links

Extract all links from Markdown text.

Input:

{
  "markdown": "[Google](https://google.com)\n[GitHub](https://github.com)"
}

Output:

{
  "count": 2,
  "links": [
    { "text": "Google", "url": "https://google.com" },
    { "text": "GitHub", "url": "https://github.com" }
  ]
}

word_count — Word Count

Count words, characters, paragraphs, and lines in Markdown text.

Input:

{
  "markdown": "# Hello World\n\nThis is a paragraph with some words."
}

Output:

{
  "words": 9,
  "characters": 55,
  "charactersNoSpaces": 46,
  "paragraphs": 2,
  "lines": 2
}

Supported Markdown Syntax

| Syntax | Description | |--------|-------------| | # H1 - ###### H6 | Headings | | **bold** | Bold text | | *italic* | Italic text | | [text](url) | Links | | `code` | Inline code | | ```lang\ncode\n``` | Code blocks | | --- | Horizontal rules | | > quote | Blockquotes | | - item | Unordered lists | | 1. item | Ordered lists |

Design

| Feature | Why | |---------|-----| | Zero runtime deps | Only @modelcontextprotocol/sdk and zod | | Pure Regex | No Markdown parsing library needed | | Fast conversion | Simple pattern matching, no AST | | Structured output | JSON in → JSON/HTML out |

License

MIT