mcp-server-test

dushyant4342/mcp-server-test

3.2

If you are the rightful owner of mcp-server-test and would like to certify it and/or have it hosted online, please leave a comment on the right or send an email to henry@mcphub.com.

MCP Server is a versatile framework for integrating AI, automation, and custom logic into applications.

Tools
2
Resources
0
Prompts
0

MCP Server

Overview

MCP Server is a flexible and extensible server framework designed to expose programmable tools, APIs, and automation agents in a unified, discoverable, and secure way. MCP servers are ideal for integrating AI, automation, and custom business logic into applications, workflows, and developer tools.

Use Cases

  • AI Agent Hosting: Run AI-powered tools and assistants that can interact with users, data, or external APIs.
  • Automation Gateways: Expose automation scripts and business logic as callable tools for integration with other systems.
  • Developer Productivity: Provide code completion, code intelligence, and other developer tools via standardized interfaces.
  • Mock/Test Servers: Quickly prototype and test tool integrations before deploying to production.
  • Custom API Aggregation: Combine multiple APIs and tools into a single, discoverable endpoint.

Features

  • Tool Registration: Easily register Python functions as callable tools using decorators.
  • Async Support: Native support for asynchronous tools and workflows.
  • Transport Flexibility: Supports multiple transports (e.g., stdio, HTTP, sockets) for integration with various clients.
  • Secure and Isolated: Tools can be sandboxed and permissions can be managed per tool.
  • Extensible: Add new tools, transports, and integrations with minimal effort.
  • Discoverability: Tools are self-describing and can be discovered by clients.

Definitions

  • Tool: A function or method exposed by the MCP server that can be called remotely.
  • Transport: The mechanism by which clients communicate with the MCP server (e.g., stdio for CLI, HTTP for web).
  • Agent: An AI or automation agent that can use tools to perform tasks.
  • Tool Decorator: A Python decorator (@mcp.tool()) used to register a function as a tool.

Example Tools

  • get_alerts(state: str): Returns weather alerts for a given state.
  • get_forecast(latitude: float, longitude: float): Returns weather forecasts for a specific location.

Example: Weather MCP Server

from mcp.server.fastmcp import FastMCP

mcp = FastMCP("weather")

@mcp.tool()
async def get_alerts(state: str) -> str:
    return f"🚨 Mock alert for state: {state}"

@mcp.tool()
async def get_forecast(latitude: float, longitude: float) -> str:
    return f"🌤️ Mock forecast at lat: {latitude}, lon: {longitude}"

if __name__ == "__main__":
    print("✅ MCP Weather Server Started")
    mcp.run(transport="stdio")

MCP Server Architecture

  • Core Server: Handles tool registration, request routing, and transport management.
  • Tool Layer: User-defined Python functions registered as tools.
  • Transport Layer: Supports various communication protocols.
  • Security Layer: Manages authentication, authorization, and sandboxing.

How to Use

  1. Clone the Repository: 
    git clone https://github.com/dushyant4342/mcp-server-test.git
cd mcp-server-test
  2. Install Requirements: (Add requirements.txt as needed)
  3. Run the Example Server: 
    python weather.py

More Information

For more details, see the code examples and comments in the repository, or open an issue for questions and suggestions.