MCP Multi-Server Client

Example Use Cases

• 1

  Execute and capture results from arbitrary Python code securely via a dedicated Python executor server.

• 2

  Perform file I/O and directory listing tasks using a remote/local file processor server.

• 3

  Run shell commands or simple calculations by routing requests to specialized MCP servers while leveraging OpenAI models for orchestration.

Description

This project provides a multi-server client implementing the Model Control Protocol (MCP) to run and manage multiple local server tools (calculator, Python executor, file processor, shell processor) concurrently. It includes intelligent tool routing to find appropriate servers for tasks, an interactive CLI, and configurable OpenAI API integration for model-based decision making. Configurations can be loaded from JSON files or environment variables (.env), and the system supports code execution, file operations, and shell command execution for agent use cases. The architecture is intended to make building and testing tool-enabled agents straightforward and configurable.