Markdownify MCP Server

Name: Markdownify MCP Server
Author: zcaceres

@zcaceres

リポジトリ

Markdownify MCP Server について

A Model Context Protocol server for converting almost anything to Markdown

#markdown #ocr #tools #ai #mcp #anthropic #anthropic-claude #anthropic-ai #model-context-protocol

基本情報

カテゴリ

その他

ライセンス

MIT

ランタイム

node

トランスポート

stdio

公開者

zcaceres

設定

以下の設定を使って、このサーバーを MCP 対応クライアントに追加してください。

{
  "mcpServers": {
    "markdownify-mcp": {
      "command": "docker",
      "args": [
        "build",
        "-t",
        "markdownify-mcp",
        "."
      ]
    }
  }
}

ツール

Convert YouTube videos to Markdown

Convert PDF files to Markdown

Convert Bing search results to Markdown

Convert web pages to Markdown

Convert images to Markdown with metadata

Convert audio files to Markdown with transcription

Convert DOCX files to Markdown

Convert XLSX files to Markdown

Convert PPTX files to Markdown

Retrieve an existing Markdown file. File extension must end with: *.md, *.markdown.

概要

What is Markdownify MCP Server?

Markdownify is a Model Context Protocol (MCP) server that converts various file types and web content to Markdown format. It provides tools to transform PDFs, images, audio files, web pages, and more into readable, shareable Markdown text.

How to use Markdownify MCP Server?

Clone the repository, install dependencies with bun install, build with bun run build, and start with bun start. To integrate with a desktop app, add the server configuration to your app's mcpServers object. Optionally use Docker; mount host directories and set MD_ALLOWED_PATHS to enforce read boundaries.

Key features of Markdownify MCP Server

Convert PDF, images, audio, DOCX, XLSX, and PPTX to Markdown
Transcribe audio files to Markdown
Extract YouTube video transcripts as Markdown
Fetch Bing search results and general web pages as Markdown
Retrieve existing Markdown files from allowed directories
Restrict file access with the MD_ALLOWED_PATHS environment variable

Use cases of Markdownify MCP Server

Transform PDF reports into editable Markdown documents
Transcribe meeting recordings or podcasts into searchable notes
Capture YouTube video summaries as Markdown for quick reference
Convert office files (DOCX, XLSX, PPTX) to plain Markdown
Scrape web pages or search results into structured Markdown content

FAQ from Markdownify MCP Server

What file types can Markdownify MCP Server convert?

It converts PDF, images, audio (with transcription), DOCX, XLSX, PPTX, YouTube videos, Bing search results, and general web pages. It can also retrieve existing Markdown files.

Does the Docker image support all conversion features?

No. The published Docker image installs only markitdown[pdf]. Audio transcription and image OCR require the [all] extras and will fail in the slim image. Use a local install (bun install) for full functionality.

How do I restrict which files the server can read?

Set the MD_ALLOWED_PATHS environment variable to a colon-separated list of allowed directories. When set, all file‑input tools reject paths outside these directories.

Where is the `markitdown` executable located?

By default, the server looks for it in the project’s .venv/bin/markitdown, then falls back to any markitdown on PATH. Override with the MARKITDOWN_PATH environment variable if you have installed it system‑wide (e.g. via pipx).

What are the runtime dependencies?

The server requires Node.js (via node in the desktop configuration), Bun for building and running, and a Python virtual environment (created automatically by bun install) for the markitdown utility. For Docker, only Node.js and the slim markitdown are included.