MCP with Gemini Tutorial

This repository contains the complete code for the tutorial on building Model Context Protocol (MCP) servers with Google's Gemini 2.0 model, as described in this blog post.

What is Model Context Protocol (MCP)?

MCP is an open standard developed by Anthropic that enables AI models to seamlessly access external tools and resources. It creates a standardized way for AI models to interact with tools, access the internet, run code, and more, without needing custom integrations for each tool or model.

Key benefits include:

• Interoperability: Any MCP-compatible model can use any MCP-compatible tool
• Modularity: Add or update tools without changing model integrations
• Standardization: Consistent interface reduces integration complexity
• Separation of Concerns: Clean division between model capabilities and tool functionality

Project Overview

This tutorial demonstrates how to:

• Build a complete MCP server with Brave Search integration
• Connect it to Google's Gemini 2.0 model
• Create a flexible architecture for AI-powered applications

Getting Started

Prerequisites

• Bun (for fast TypeScript execution)
• Brave Search API key
• Google API key for Gemini access

Installation

# Clone the repository
git clone https://github.com/GuiBibeau/mcp-gemini-tutorial.git
cd mcp-tutorial

# Install dependencies
bun install

Environment Setup

Create a `.env` file with your API keys:

BRAVE_API_KEY="your_brave_api_key"
GOOGLE_API_KEY="your_google_api_key"

Usage

Running the Basic Client

bun examples/basic-client.ts

Running the Gemini Integration

bun examples/gemini-tool-function.ts

Project Structure

• `src/` - Core implementation of the MCP server and tools
• `examples/` - Example clients demonstrating how to use the MCP server
• `tests/` - Test files for the project

Tools Implemented

This MCP server exposes two main tools:

1. Web Search: For general internet searches via Brave Search

2. Local Search: For finding businesses and locations via Brave Search

Extending the Project

You can add your own tools by:

1. Defining a new tool with a schema

2. Implementing the functionality

3. Registering it with the MCP server

Learn More

• Official MCP Documentation
• Google Gemini API Documentation
• Brave Search API Documentation

License

MIT

This project was created using `bun init` in bun v1.1.37. Bun is a fast all-in-one JavaScript runtime.