MCP Course #4 - Building MCP Client with Google Gemini and Python!

End-to-end MCP Course with code. GitHub Repos - Please Subscribe, allow pop-ups and then login to The AI Language website to access our GitHub Repos. Access is available only to our YouTube subscribers - https://theailanguage.com/onlySubscribers?id=mcp_client&site=github Video Tutorial for building MCP server that's used in this video https://www.youtube.com/watch?v=_veLqeCzdIQ&list=PL6tW9BrhiPTCDteflzehKS6Cn3a79-iCs 📢 Subscribe for More AI & Coding Tutorials! https://youtube.com/@theailanguage?sub_confirmation=1 Happy building! ☺️ --------------------------------------------------------------------------------------------- Udemy Course (get completion certificate, practice questions, Q&A) https://www.udemy.com/course/modelcontextprotocol/?referralCode=6FADE0F85C5DB97203C6 --------------------------------------------------------------------------------------------- 00:00 Introduction 00:13 1 minute preview - what we’ll build 01:59 Please subscribe for AI, Agents and more 02:12 Overview of MCP 02:47 Why use Gemini? 03:18 Objective 03:38 Setup Project 05:53 Get Gemini API Key 07:17 MCP client.py Code Walkthrough 07:43 Client.py Imports 08:51 MCPClient Class 09:23 Connect to Server 13:42 Process Query 17:08 Chat Loop, Cleanup, main functions 17:59 Test your MCP client 🚀 MCP Client with Gemini AI This is an MCP (Model Context Protocol) client that connects to an MCP server and integrates with Google’s Gemini AI. The client sends user queries to Gemini, allows Gemini to call external tools from the MCP server, executes them, and keeps track of the conversation history. 🔹 MCP (Model Context Protocol) allows AI models to interact with external tools and fetch information dynamically. 🔹 Google’s Gemini AI processes user queries and calls MCP tools when needed. 🔹 This client keeps a history of interactions so Gemini can remember past commands. 📌 Features ✅ Connects to an MCP server (Python or Node.js) ✅ Sends queries to Google Gemini AI ✅ Lets Gemini call external tools from the MCP server ✅ Executes MCP tool commands and returns the results ✅ Maintains conversation history, so Gemini remembers past queries 📦 Installation 1️⃣ Install the required dependencies using uv (Universal Virtualenv): uv add mcp python-dotenv google-genai 2️⃣ Clone the repository: git clone https://github.com/your-username/mcp-client-gemini.git cd mcp-client-gemini 3️⃣ Set up the project and virtual environment: uv init mcp-client cd mcp-client uv venv 4️⃣ Activate the virtual environment: # On Windows: .venv\Scripts\activate # On MacOS/Linux: source .venv/bin/activate 🔑 Setting Up the API Key To use Google Gemini AI, you need an API key. 1️⃣ Create a .env file: touch .env 2️⃣ Add your API key inside .env: GEMINI_API_KEY=your_api_key_here 3️⃣ Make sure .env is ignored in Git: echo ".env" {angled brackets} .gitignore 🚀 Running the MCP Client Start the MCP client and connect it to an MCP server: uv run client.py path/to/server.py # Use a Python server uv run client.py path/to/server.js # Use a Node.js server Example (if using a weather server): uv run client.py ./server/weather.py 🔧 How It Works 1️⃣ The user enters a query (e.g., "Create a file named test.txt"). 2️⃣ The MCP client sends the query to Gemini AI. 3️⃣ Gemini AI checks available MCP tools and calls the correct one. 4️⃣ The MCP client executes the command and returns the result. 5️⃣ Gemini remembers past interactions and adjusts responses accordingly. 📁 Project Structure mcp-client-gemini/ │── client.py # MCP Client (Main script) │── .env # Stores API Key (ignored in Git) │── README.md # Documentation │── requirements.txt # Dependencies (optional) │── server/ # Folder for MCP server scripts (e.g., weather.py) │── .gitignore # Ignores sensitive files 📢 Subscribe for More AI & Coding Tutorials! https://youtube.com/@theailanguage?sub_confirmation=1 🎯 Contributing Feel free to submit issues or contribute improvements via pull requests. 🚀 Now you're ready to build your own MCP client with Gemini AI! Let me know if you have any questions in the comments! 😊🎯