MCP Sample Application (Stateless Architecture)

A Model Context Protocol (MCP) sample project implemented in Python using FastMCP and mcp.client. This project demonstrates a stateless MCP server exposing mathematical tools over stdio transport along with an asynchronous client.

Features

• Stateless MCP Server (server.py): Uses standard stdio transport without maintaining session states.

• Asynchronous Client (client.py): Connects to the MCP server, discovers available tools dynamically, and executes tool calls.

Related MCP server: mcp-poc

Setup & Installation

Prerequisites

• Python 3.10+ (Python 3.14 supported)

• Virtual Environment (.venv)

1. Create and Activate Virtual Environment

python3 -m venv .venv
source .venv/bin/activate

2. Install Dependencies

pip install -r requirements.txt

Running the Application

Option A: Using the Runner Script

Execute the provided bash script to automatically use the environment's Python interpreter:

chmod +x run_sample.sh
./run_sample.sh

Option B: Running Manually

Activate your virtual environment and run the client:

python3 client.py

Example Output

When running client.py or ./run_sample.sh, the client will:

1. Start the MCP server via stdio.

2. Discover and print all available registered tools.

3. Invoke the soma tool with arguments a=15 and b=27.

Starting MCP Server (Stateless Mode 2026-07-28 Spec)...
Connected to MCP Server (Stateless Mode).

--- Discovered Tools (Stateless) ---
{
  "name": "soma",
  "description": "Soma dois números fornecidos pelo usuário statelessly.",
  ...
}
------------------------------------

Calling tool: soma(15, 27)...
Result: O resultado de 15.0 + 27.0 é 42.0

Client finished. Stateless server connection closed.