Currency Agent

An AI agent implementation built on Google Cloud, featuring MCP (Model Context Protocol) for tool integration, ADK (Agent Development Kit) for orchestration, and A2A (Agent-to-Agent) protocol for inter-agent communication.

Setup & Google Cloud Configuration

1. Authenticate & Verify Project

Open the Cloud Shell and verify your active account:

gcloud auth list

Display the active Google Cloud project ID:

gcloud config list project

2. Configure Environment Variables

Export your Project ID to an environment variable and configure gcloud:

export PROJECT_ID="<YOUR_PROJECT_ID>"
gcloud config set project $PROJECT_ID

3. Enable Required Google Cloud APIs

Enable all required Google Cloud service APIs for the project:

gcloud services enable \
  cloudresourcemanager.googleapis.com \
  servicenetworking.googleapis.com \
  run.googleapis.com \
  cloudbuild.googleapis.com \
  artifactregistry.googleapis.com \
  aiplatform.googleapis.com \
  compute.googleapis.com

Related MCP server: FrankfurterMCP

Installation & Local Setup

This project requires Python 3.10+. Verify your Python version in Cloud Shell:

python3 --version

1. Clone the Repository

Clone the repository to your environment:

git clone https://github.com/jackwotherspoon/currency-agent.git
cd currency-agent

2. Install Dependency Manager (uv)

This project uses uv to manage Python dependencies. Install uv using the following curl command:

curl -LsSf https://astral.sh/uv/install.sh | bash

To verify the installation in a new terminal, check the version and executable location:

uv --version
which uv

3. Configure Environment Variables

Create and configure your .env file to supply the necessary variables:

echo "GOOGLE_GEN_AI_USE_VERTEXAI=True" >> .env \
  && echo "GOOGLE_CLOUD_LOCATION=us-central1" >> .env \
  && echo "GOOGLE_CLOUD_PROJECT=$PROJECT_ID" >> .env

Verify that the .env file has been populated correctly:

cat .env

Running and Testing the Server Locally

1. Start the MCP Server

Run the following command in your Cloud Shell terminal to start the server:

uv run mcp-server/server.py

2. Test the MCP Server

Open a new Cloud Shell terminal and run the test script. This will query the server's API and output the result:

uv run mcp-server/test_server.py

Deployment to Google Cloud Run

To make the tool permanent, reliable, and benefit from automatic scaling, deploy the MCP server to Google Cloud Run.

1. Deploy the Service

Stop the local server and run the following deployment command from within the mcp-server directory:

gcloud run deploy mcp-server \
  --no-allow-unauthenticated \
  --region=us-central1 \
  --source .

2. Authenticate and Proxy Deployed Calls

To securely access and test the live Cloud Run service locally, start a local proxy tunnel:

gcloud run services proxy mcp-server --region=us-central1

This establishes a secure tunnel between your local machine (or Cloud Shell) and the Cloud Run service.

3. Test the Deployed Service

With the proxy running, open a new Cloud Shell terminal and run the test script:

uv run mcp-server/test_server.py

4. View Deployment Logs

To monitor the service and verify requests, stream the logs from the command line:

gcloud run services logs read mcp-server --region=us-central1 --limit=5

You are now ready to build and run agents interacting with your deployed tools!

External APIs & Data Sources

Currency Converter Endpoint

The agent utilizes the Frankfurter API for real-time exchange rates:

• Endpoint: https://api.frankfurter.dev/v2/latest

• Sample Request:

  GET https://api.frankfurter.dev/v2/latest?base=USD&symbols=INR,EUR