tws-drex-mcp 0.1.1

An MCP server providing tools to interact with the DREX document extraction AI API via LoginRadius authentication.

TWS DREX MCP Server

This MCP server provides tools to interact with the DREX document extraction AI API and LoginRadius authentication.

Installation

Install the package using pip:

pip install tws-drex-mcp

Functionality

The following tools are provided by the server:

• get_token: Obtains an authentication token from LoginRadius using a username and password.
• file_upload: Uploads files to the DREX API.
• get_status: Retrieves the processing status of a file in the DREX API.
• get_file_results: Retrieves the processed results of a file in the DREX API.

Configuration

Before using the server's tools, ensure the following environment variables are set:

• DREX_BASE_URL: The base URL of the target DREX API instance.
• LR_API_KEY: Your LoginRadius API key.
• LR_API_SECRET: Your LoginRadius API secret.

The MCP framework will automatically pick up these variables when running the server's tools.

Usage

get_token

This tool obtains an authentication token from LoginRadius using a username and password.

Example:

{
    "username": "your_username",
    "password": "your_password"
}

file_upload

This tool uploads files to the DREX API.

Example:

{
    "file_paths": ["path/to/file1.txt", "path/to/file2.png"],
    "uploaded_by": "user_name",
    "token": "your_access_token"
}

get_status

This tool retrieves the processing status of a file in the DREX API.

Example:

{
    "file_id": "file_id_1",
    "token": "your_access_token"
}

get_file_results

This tool retrieves the processed results of a file in the DREX API.

Example:

{
    "file_id": "file_id_1",
    "token": "your_access_token"
  }

LoginRadius Initialization

The LoginRadius SDK is initialized using the LR_API_KEY and LR_API_SECRET environment variables.

Dependencies

The dependencies required to run this server are listed in the mcp-server-demo/pyproject.toml file.

Running the Server

To run the server, use the following command:

uv --directory mcp-server-demo run --with mcp[cli] mcp run mcp-server-demo/server.py

To debug the server, use the following command:

mcp dev server.py

To install the server to Claude desktop, use the following command:

mcp install server.py

Example configuration:

{
  "mcpServers": {
    "DREX": {
      "command": "uv",
      "args": [
        "--directory",
        "<your-repo-path>\\mcp-server-demo",
        "run",
        "--with",
        "mcp[cli]",
        "mcp",
        "run",
        "<your-repo-path>\\mcp-server-demo\\server.py"
      ]
    }
  }
}