Model Context Protocol (MCP)

Name: Model Context Protocol (MCP)
Author: drkhan107

@drkhan107

リポジトリ

Model Context Protocol (MCP) について

A working pattern for SSE-based MCP clients and servers using Gemini LLM

基本情報

カテゴリ

AI とエージェント

ライセンス

MIT

ランタイム

jupyter notebook

トランスポート

stdio

公開者

drkhan107

設定

以下の設定を使って、このサーバーを MCP 対応クライアントに追加してください。

{
  "mcpServers": {
    "mcp_gemini": {
      "command": "python",
      "args": [
        "sse_server.py"
      ]
    }
  }
}

ツール

ツールは検出されませんでした

ツールは README から自動的に抽出されます。メンテナーは ## Tools という見出しの下に記載することで、このタブに反映できます。

概要

What is Model Context Protocol (MCP)?

A working demo of the Model Context Protocol integrated with Google’s Gemini. It provides a server that uses SSE (Server-Sent Events) to expose MCP endpoints, alongside a FastAPI backend and a Streamlit GUI for interacting with the system.

How to use Model Context Protocol (MCP)?

Clone the repository, set a GOOGLE_API_KEY in a .env file, install dependencies with pip install -r requirements.txt, then start the MCP server with python sse_server.py (defaults to http://localhost:8080/sse). Optionally launch the SSE client with python ssc_client.py http://localhost:8080/sse, start the FastAPI server with python fastapp.py, and finally run the Streamlit app with streamlit run app.py (opens on localhost:8501). Click “Connect to MCP server” in the browser.

Key features of Model Context Protocol (MCP)

Integrates with Google’s Gemini.
Uses SSE (Server-Sent Events) transport.
Includes a FastAPI server for a GUI backend.
Provides a Streamlit web interface.
Simple setup with a .env file for API key.
Fully functional demo ready to run locally.

Use cases of Model Context Protocol (MCP)

Experimenting with MCP and Gemini in a local environment.
Building a chat-like GUI that connects to an MCP server.
Prototyping applications that combine context management with Google’s generative AI.
Learning how to set up an MCP server with SSE and a client.

FAQ from Model Context Protocol (MCP)

What transport does the MCP server use?

The server uses SSE (Server-Sent Events) and runs by default on http://localhost:8080/sse.

Do I need an API key to run this server?

Yes, you must add a valid Google API key to a .env file as GOOGLE_API_KEY.

How do I start the GUI?

First ensure the MCP server and FastAPI server are running, then launch the Streamlit app with streamlit run app.py and open localhost:8501 in your browser.

What dependencies are required?

All dependencies are listed in requirements.txt and can be installed with pip install -r requirements.txt.

「AI とエージェント」の他のコンテンツ

21st.dev Magic AI Agent

21st-dev

It's like v0 but in your Cursor/WindSurf/Cline. 21st dev Magic MCP server for working with your frontend like Magic

5.3K

Model Context Protocol for Unreal Engine

chongdashu

Enable AI assistant clients like Cursor, Windsurf and Claude Desktop to control Unreal Engine through natural language using the Model Context Protocol (MCP).

2.0K