MCP Server Box

Description

MCP Server Box is a Python project integrating the Box API for file operations like searching, text extraction, AI querying, and data extraction. It utilizes box-sdk-gen and the Model Context Protocol (MCP) to streamline interaction with Box files and folders. MCP standardizes model interaction with data sources, enabling efficient and scalable operations on Box data using AI and machine learning. The project offers a flexible solution for managing and processing Box data.

Tools implemented

The project includes several Box tools: box_who_am_i (user info), box_authorize_app_tool (authorization), box_search_tool (file search with filtering), box_read_tool (text extraction), box_ask_ai_tool (AI querying), box_search_folder_by_name (folder location), box_ai_extract_data (AI-based data extraction), and box_list_folder_content_by_folder_id (folder content listing).

Requirements

• Python 3.13+
• Box API credentials

Installation

Installation involves cloning the repository, installing uv, creating a virtual environment, locking dependencies, and setting up a .env file with Box API credentials.

Usage

The MCP server is started using uv --directory /path-to-the-project/mcp-server-box run src/mcp_server_box.py. Claude can be used as a client by configuring claude_desktop_config.json with the server command and arguments.