Document Parse

Parses documents (PDF/Word/Excel/PowerPoint) to Markdown via a local MCP stdio server.

other

5

GitHub Stars

other

Language

6 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "tianfn-mcp-parse-document": {
      "command": "uv",
      "args": [
        "tool",
        "run",
        "mcp-document-parse"
      ],
      "env": {
        "NIUTRANS_API_KEY": "${env.NIUTRANS_API_KEY}",
        "NIUTRANS_DOCUMENT_APPID": "${env.NIUTRANS_DOCUMENT_APPID}"
      }
    }
  }
}

You can use the MCP Document Parse Tool to convert a wide range of document formats into Markdown content that you can integrate into your applications. This server-like tool is designed to parse PDFs, Word, Excel, and PowerPoint files, exposing a straightforward way to extract text content for downstream use.

How to use

After you start the MCP document parse tool as a local STDIO server, you can access its endpoints to list available tools and resources. The server exposes a tool named parse_document_by_path that converts a file at a given path into Markdown text. You can also query document://supported-types to learn which file types are supported.

From your MCP client, you’ll be able to invoke the available endpoints to convert documents and to discover supported types. Use ListTools to confirm the available tool parse_document_by_path and use ListResources to read document://supported-types.

How to install

Prerequisites depend on the MCP client ecosystem. The documented installation flow uses a local MCP runtime tool. Follow these concrete steps to install and configure the MCP Document Parse Tool.

Step 1. Install the MCP tool

uv tool install mcp-document-parse

Step 2. Set required environment variables

NIUTRANS_API_KEY: Your Niutrans translation API key
NIUTRANS_DOCUMENT_APPID: Your Niutrans document API app ID

Step 3. Create MCP client configuration to run the tool as an MCP server (stdio). The following configuration runs the tool through the UV runtime and exposes the required environment variables.

{
  "mcpServers": {
    "document_parse": {
      "type": "stdio",
      "command": "uv",
      "args": [
        "tool",
        "run",
        "mcp-document-parse"
      ],
      "env": {
        "NIUTRANS_API_KEY": "${env.NIUTRANS_API_KEY}",
        "NIUTRANS_DOCUMENT_APPID": "${env.NIUTRANS_DOCUMENT_APPID}"
      }
    }
  }
}

Additional notes

This tool supports the following file formats: PDF, Word (.doc, .docx), Excel (.xls, .xlsx), and PowerPoint (.ppt, .pptx). It exposes two main capabilities: parsing a file by path into Markdown via parse_document_by_path and listing supported types via document://supported-types.

When you run the server, you can call ListTools to verify the available tools and ListResources to access the document://supported-types resource. The outputs are designed to help you integrate document parsing into your workflows.

License and contact

MIT License. If you have questions or suggestions, contact tianfengning@niutrans.com.

Tools available

parse_document_by_path – Converts a file at a given absolute path to Markdown content. Returns a status and the textual content or an error message.

document://supported-types – Returns a JSON object describing each supported file type and its description.

Available tools

parse_document_by_path

Converts a file at a given absolute path to Markdown content. Returns status and text content or an error.

document://supported-types

Provides information about all supported document types and their descriptions.