Integrating Agents with MCP Servers

This tutorial guides you through creating an AI-powered Weather Assistant that integrates with an MCP server to provide real-time weather information. By the end of this tutorial, you will have a personal assistant capable of delivering current weather conditions and forecast details for any location worldwide.

Prerequisites

Before you begin, ensure you have a running MCP Server connected to a weather service. For this setup, you can set up an MCP Server using the guidelines given here. This server enables effective communication between your AI agent and the weather API, allowing real-time data retrieval.

Create the AI agent

Before integrating MCP capabilities, you must first create an AI agent. Follow Steps 1 to 5 in the Introduction to Chat Agents guide to set up your agent.

For this tutorial, you can configure the agent with the following role and instructions:

Role:

Weather AI Assistant

Instructions:

You are Nova, a smart AI assistant dedicated to providing accurate and timely weather information.

Your primary responsibilities include:
- Current Weather: Provide detailed and user-friendly current weather information for a given location.
- Weather Forecast: Share reliable weather forecasts according to user preferences (e.g., hourly, daily).

Guidelines:
- Always communicate in a natural, friendly, and professional tone.
- Provide concise summaries unless the user explicitly requests detailed information.
- Confirm location details if ambiguous and suggest alternatives when data is unavailable.

Add MCP Server to the agent

By connecting to the Weather MCP server, your AI agent can access and interact with real-time weather data sources. To integrate it, follow the steps below.

Step 1: Add the MCP server

Provide the MCP server connection details.

1. In Agent Flow View, click the + button at the bottom-right of the AI Agent box.
2. Under Add Tools section, select Use MCP Server.

3. Provide the necessary configuration details, then click Save Tool.

   

Step 2: Customize the MCP server

You can further customize the MCP configuration to include additional weather tools to suit your use case.

Interact with the agent

After completing the above steps, your personal AI assistant agent is now ready to assist you with necessary tasks. WSO2 Integrator: BI provides a built-in chat interface to interact with the agent.

To start chatting with the agent:

1. Click the Chat button located at the top-left corner of the interface.
2. You will be prompted to run the integration. Click Run Integration.

3. Start chatting with your assistant.