48 lines
1.3 KiB
Markdown
48 lines
1.3 KiB
Markdown
# Gemini MCP Integration
|
|
|
|
This directory contains a boilerplate setup for integrating Google Gemini with the Model Context Protocol (MCP).
|
|
|
|
## Structure
|
|
|
|
- **`simple_server.py`**: A simple MCP server using `FastMCP` that exposes two tools (`add_numbers`, `get_current_time`).
|
|
- **`gemini_client.py`**: A Python script that acts as an MCP client. It connects to `simple_server.py`, lists the available tools, and provides guidance on integrating them with the `google-genai` SDK.
|
|
- **`setup.sh`**: A provision script that sets up the Python virtual environment and installs dependencies.
|
|
- **`requirements.txt`**: List of Python dependencies.
|
|
|
|
## Prerequisites
|
|
|
|
- Python 3
|
|
- Node.js & npm (for MCP Inspector)
|
|
|
|
## Setup
|
|
|
|
1. Run the setup script:
|
|
```bash
|
|
./setup.sh
|
|
```
|
|
2. Activate the virtual environment:
|
|
```bash
|
|
source venv/bin/activate
|
|
```
|
|
3. Set your Google API Key in `.env`:
|
|
```bash
|
|
# Edit .env file
|
|
GOOGLE_API_KEY=your_api_key_here
|
|
```
|
|
|
|
## Running the Example
|
|
|
|
Run the client script to see the connection to the server and the list of tools:
|
|
|
|
```bash
|
|
python gemini_client.py
|
|
```
|
|
|
|
## Using MCP Inspector
|
|
|
|
You can use the official MCP Inspector to interact with the `simple_server.py` visually.
|
|
|
|
```bash
|
|
npx @modelcontextprotocol/inspector python simple_server.py
|
|
```
|