MCP File Parser

A file parser using Kimi AI. However, all of this is free.

Installation

pip install mcp-fileparser

Usage

As a Python Library

1. Set up your environment variables.

   Create a .env file in your project root and add your Kimi API key:

   KIMI_API_KEY="your_api_key"
   

2. Use the parse_file function.

   from mcp_fileparser import parse_file
   
   content = parse_file("/path/to/your/file.xlsx")
   print(content)
   

As an MCP Server

This package can also be used as an MCP (Model Context Protocol) server, providing file parsing capabilities to MCP clients.

Installation

1. Install the package:

   pip install mcp-fileparser
   

2. Set up your environment variables:

   Create a .env file in your project root and add your Kimi API key:

   KIMI_API_KEY="your_api_key"
   

Configuration

Add the MCP server to your MCP client configuration. For example, in Claude Desktop or other MCP-compatible clients:

{
  "mcpServers": {
    "mcp-fileparser": {
      "command": "npx",
      "args": [
        "--from",
        "mcp-fileparser",
        "parse_file"
      ],
      "env": {
        "KIMI_API_KEY": "your_api_key"
      }
    }
  }
}

Available Tools

Once connected, the MCP server provides the following tool:

• parse_file: Parse a file and extract its content using Kimi AI
  • Parameters:
    • fpath (string): The path to the file you want to parse
  • Returns: The extracted content as a string

Example Usage

In an MCP-compatible client, you can use the parse_file tool like this:

# Example of how the tool would be called by an MCP client
result = await client.call_tool("parse_file", {
  "fpath": "/path/to/your/document.pdf"
})
print(result.content)

Supported File Types

The file parser supports various file types including:

• PDF documents
• Word documents (.docx)
• Excel spreadsheets (.xlsx)
• Text files (.txt)
• And many other formats supported by Kimi AI

Running the Server Manually

You can also run the MCP server manually for testing:

# Set your API key
export KIMI_API_KEY="your_api_key"

# Run the server
fileparser

The server will start and listen for MCP connections via stdio transport.

License

This project is licensed under the MIT License.