AI-MCP-Tools

anaslimem/AI-MCP-Tools

3.1

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

A simple implementation of an MCP (Model Context Protocol) server in Python, deployed with FastMCP Cloud.

Tools
4
Resources
0
Prompts
0

AI MCP Tools

A simple implementation of an MCP (Model Context Protocol) server in Python, deployed with FastMCP Cloud.

What is MCP (Model Context Protocol)?

MCP is a protocol designed for connecting and serving tools, models, and functions in a unified and context-aware way. It enables the creation of smart, extensible servers that can expose custom tools (functions) for remote or local access.

Project Overview

This project demonstrates a basic MCP server built using the FastMCP Python framework. The server is deployed and accessible publicly via FastMCP Cloud:

👉 Live Demo: https://anaslimem.fastmcp.app

Features

  • MCP Server: Runs with FastMCP and exposes custom tools.
  • Custom Tools:
    • web_search(query): Uses the Serper.dev API to perform Google searches and returns summarized results.
    • fetch_page_content(url): Fetches and extracts the main content from a web page given its URL.
    • summarize_text(text): Summarizes the given text using Gemini (Google Generative AI).
    • save_results(topic, content): Saves the summarized results to a local file on the server.
  • Environment Variables: Uses .env for sensitive keys such as the Serper API key.
  • Simple Client Example: Includes a Python client script to call the server’s tools asynchronously.

Example Usage

Web Search Tool
result = await client.call_tool("web_search", {"query": "What are the difference between chatgpt and gemini"})
# Output: "Top 3 search results with titles, snippets, and links."
Fetch Page Content Tool
content = await client.call_tool("fetch_page_content", {"url": "https://example.com"})
# Output: Main text content of the page as a string.
Summarize Text Tool
summary = await client.call_tool("summarize_text", {"text": content})
# Output: Concise summary of the provided text.

Running Locally

  1. Clone the repo:

    git clone https://github.com/anaslimem/first_mcp_server.git
cd first_mcp_server

  2. Install dependencies:

    pip install -r requirements.txt

  3. Set up your .env file:

    SERPER_API_KEY=your_serper_api_key
GOOGLE_API_KEY=your_google_api_key

  4. Start the server:

    python my_server.py

  5. Test with the included client:

    python client.py

Deployment

This server is deployed using FastMCP Cloud, making it accessible online at https://anaslimem.fastmcp.app.

Feel free to fork, experiment, and extend this project!