uAgent ↔ AI Agent Communication Using Fetch.ai SDK and uAgents

This guide demonstrates how to enable communication between a Microservice Agent created using the uagents framework and an AI Agent created using the Fetch.ai SDK.

uAgent Script (uagent.py)

Please remember to have the uagents and the fetchai package installed in the terminal in order to create and run the agents.

This uAgent will receive a message from the AI Agent and send a response back.

uagent.py

from uagents import Agent, Context, Model

# Define the request model the uAgent will handle
class Request(Model):
    message: str

# Define the response model the uAgent will send back
class Response(Model):
    response: str

# Initialize the uAgent
uagent = Agent(
    name="Sample uAgent",
    port=8000,
    endpoint=["http://localhost:8000/submit"]
)

# Handle incoming messages with the Request model
@uagent.on_message(model=Request)
async def message_handler(ctx: Context, sender: str, msg: Request):
    ctx.logger.info(f"Received message from {sender}: {msg.message}")

    # Generate a response message
    response = Response(response=f'Hello, AI Agent! I received your message:{msg.message}')
    
    # Send the response back to the AI Agent
    await ctx.send(sender, response)

if __name__ == "__main__":
    uagent.run()

Explanation

• Agent Initialization: The uAgent listens on port 8000 for incoming messages.
• Message Handling: When a message matching the Request model is received, the agent logs it and responds with a predefined message using the Response model.

Setting Up the AI Agent

The AI Agent will send messages to the uAgent and handle the response received from the uAgent.

AI Agent Script (ai_agent.py)

ai_agent.py

import os
from flask import Flask, request, jsonify
from flask_cors import CORS
from fetchai.crypto import Identity
from fetchai.communication import send_message_to_agent, parse_message_from_agent
import logging
from dotenv import load_dotenv

# Configure logging
logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(name)s - %(levelname)s - %(message)s')
logger = logging.getLogger(__name__)

# Initialize Flask app
app = Flask(__name__)
CORS(app)

client_identity = None
agent_response = None

class Request(Model):
    message: str

# Load environment variables from .env file
load_dotenv()

def init_client():
    """Initialize and register the client agent."""
    global client_identity
    try:
        # Load the client identity from environment variables
        client_identity = Identity.from_seed("Sample AI AGENT SEED PHRASE for communication", 0)
        logger.info(f"Client agent started with address: {client_identity.address}")

        readme = """
        ![domain:innovation-lab](https://img.shields.io/badge/innovation--lab-3D8BD3)
        domain:domain-of-your-agent

        <description>This Agent can send a message to a uAgent and receive a message from a uAgent in string format.</description>
        <use_cases>
        <use_case>Send and receive messages with another uAgent.</use_case>
        </use_cases>
        <payload_requirements>
            <description>This agent can only send and receive messages in text format.</description>
            <payload>
                <requirement>
                    <parameter>message</parameter>
                    <description>The agent sends and receives messages in text format.</description>
                </requirement>
            </payload>
        </payload_requirements>
        """

        # Register the agent with Agentverse
        register_with_agentverse(
            identity=client_identity,
            url="http://localhost:5002/api/webhook",
            agentverse_token = os.getenv("AGENTVERSE_API_KEY"),
            agent_title="Sample AI Agent communication"
            readme=readme
        )

        logger.info("Client agent registration complete!")

    except Exception as e:
        logger.error(f"Initialization error: {e}")
        raise


@app.route('/request', methods=['POST'])
def send_data():
    """Send payload to the selected agent based on provided address."""
    global agent_response
    agent_response = None

    try:
        # Parse the request payload
        data = request.json
        payload = data.get('payload')  # Extract the payload dictionary

        uagent_address = "agent1qgd54rrq8ex4uhdxe6qg0sklz7h7dkacdk9rz4ec0l304wghw88sg35rfk6" #run the uagent.py copy the address and paste here
        
        # Build the Data Model digest for the Request model to ensure message format consistency between the uAgent and AI Agent
        model_digest = Model.build_schema_digest(Request)

        # Send the payload to the specified agent
        send_message_to_agent(
            client_identity,  # Frontend client identity
            uagent_address,  # Agent address where we have to send the data
            payload,  # Payload containing the data
            model_digest=model_digest
        )

        return jsonify({"status": "request_sent", "payload": payload})

    except Exception as e:
        logger.error(f"Error sending data to agent: {e}")
        return jsonify({"error": str(e)}), 500



# app route to get recieve the messages on the agent
@app.route('/api/webhook', methods=['POST'])
def webhook():
    """Handle incoming messages from the dashboard agent."""
    global agent_response
    try:
        # Parse the incoming webhook message
        data = request.get_data().decode("utf-8")
        logger.info("Received response")
        message = parse_message_from_agent(data)
        agent_response = message.payload
        logger.info(f"Processed response: {agent_response}")
        return jsonify({"status": "success"})
    except Exception as e:
        logger.error(f"Error in webhook: {e}")
        return jsonify({"error": str(e)}), 500

if __name__ == "__main__":
    load_dotenv()
    init_client()
    app.run(host="0.0.0.0", port=5002)

Explanation

• Data Model: The Request and Response models define the structure of messages exchanged between the AI Agent and the uAgent. A correctly defined data model is essential for sending a message to the uAgent from an SDK-based AI Agent.
• Schema Validation: The model digest ensures that messages conform to the expected schema before transmission, preventing format mismatches.
• Sending Data: The AI Agent sends a message to the uAgent using the /request endpoint.
• Handling Response: The AI Agent listens for responses using the /api/webhook endpoint.

Environment Variables

Create a .env file and add the following environment variables:

AGENTVERSE_API_KEY="YOUR_AGENTVERSE_API_KEY"
AGENT_SECRET_KEY="YOUR_SECRET_KEY_FOR_AI_AGENT"

Replace the placeholders with your actual API keys and agent secrets. Refer to this guide to get your Agentverse API Key.

Testing the Communication

Step 1: Running the uAgent

Start the uAgent by running the following command in your terminal:

python uagent.py

uAgent Logs

(venv) abhi@Fetchs-MacBook-Pro ILAgents % python3 uagent.py
INFO:     [Sample uAgent]: Starting agent with address: agent1qgd54rrq8ex4uhdxe6qg0sklz7h7dkacdk9rz4ec0l304wghw88sg35rfk6
INFO:     [Sample uAgent]: My name is Sample uAgent and my address is agent1qgd54rrq8ex4uhdxe6qg0sklz7h7dkacdk9rz4ec0l304wghw88sg35rfk6
INFO:     [Sample uAgent]: Agent inspector available at https://agentverse.ai/inspect/?uri=http%3A//127.0.0.1%3A8000&address=agent1qgd54rrq8ex4uhdxe6qg0sklz7h7dkacdk9rz4ec0l304wghw88sg35rfk6
INFO:     [Sample uAgent]: Starting server on http://0.0.0.0:8000 (Press CTRL+C to quit)
INFO:     [uagents.registration]: Registration on Almanac API successful
INFO:     [uagents.registration]: Registering on almanac contract...
INFO:     [uagents.registration]: Registering on almanac contract...complete

Copy the uAgent Address

Look for the uAgent address in the logs, which appears in this format:

agent1qgd54rrq8ex4uhdxe6qg0sklz7h7dkacdk9rz4ec0l304wghw88sg35rfk6

Copy this uAgent address and paste it into the AI Agent script at the following line:

uagent_address="PASTE YOUR UAGENT ADDRESS HERE"

Step 2: Running the AI Agent

Start the AI Agent by running the following command in your terminal:

python ai_agent.py

AI Agent Logs

(venv) abhi@Fetchs-MacBook-Pro ILAgents % python3 ai_agent.py
INFO:__main__:Client agent started with address: agent1qw7u5sw63a88kmcn5j5kxf7q326u5hgmppvy2vpxlh3re6y0yp8253ec7xl
INFO:fetchai:Registering with Almanac API
INFO:fetchai:Completed registering agent with Agentverse
INFO:__main__:Client agent registration complete!
 * Serving Flask app 'ai_agent'
 * Debug mode: off
INFO:werkzeug:WARNING: This is a development server. Do not use it in a production deployment. Use a production WSGI server instead.
 * Running on all addresses (0.0.0.0)
 * Running on http://127.0.0.1:5002
 * Running on http://172.20.10.2:5002
INFO:werkzeug:Press CTRL+C to quit

Step 3: Sending a message from Agent2 to Agent1

We will use the following curl command to send a message from the AI Agent to the uAgent:

curl -X POST http://localhost:5002/request \
-H "Content-Type: application/json" \
-d '{
  "payload": {"message": "Hello uAgent!"}
}'

Expected Logs on the uAgent Terminal

(venv) abhi@Fetchs-MacBook-Pro ILAgents % python3 uagent.py
INFO:     [Sample uAgent]: Starting agent with address: agent1qgd54rrq8ex4uhdxe6qg0sklz7h7dkacdk9rz4ec0l304wghw88sg35rfk6
INFO:     [Sample uAgent]: My name is Sample uAgent and my address is agent1qgd54rrq8ex4uhdxe6qg0sklz7h7dkacdk9rz4ec0l304wghw88sg35rfk6
INFO:     [Sample uAgent]: Agent inspector available at https://agentverse.ai/inspect/?uri=http%3A//127.0.0.1%3A8000&address=agent1qgd54rrq8ex4uhdxe6qg0sklz7h7dkacdk9rz4ec0l304wghw88sg35rfk6
INFO:     [Sample uAgent]: Starting server on http://0.0.0.0:8000 (Press CTRL+C to quit)
INFO:     [uagents.registration]: Registration on Almanac API successful
INFO:     [uagents.registration]: Registering on almanac contract...
INFO:     [uagents.registration]: Registering on almanac contract...complete
INFO:     [Sample uAgent]: Received message from agent1qw7u5sw63a88kmcn5j5kxf7q326u5hgmppvy2vpxlh3re6y0yp8253ec7xl: Hello uAgent!

Expected Logs on the AI Agent Terminal

(venv) abhi@Fetchs-MacBook-Pro ILAgents % python3 ai_agent.py
INFO:__main__:Client agent started with address: agent1qw7u5sw63a88kmcn5j5kxf7q326u5hgmppvy2vpxlh3re6y0yp8253ec7xl
INFO:fetchai:Registering with Almanac API
INFO:fetchai:Completed registering agent with Agentverse
INFO:__main__:Client agent registration complete!
 * Serving Flask app 'ai_agent'
 * Debug mode: off
INFO:werkzeug:WARNING: This is a development server. Do not use it in a production deployment. Use a production WSGI server instead.
 * Running on all addresses (0.0.0.0)
 * Running on http://127.0.0.1:5002
 * Running on http://172.20.10.2:5002
INFO:werkzeug:Press CTRL+C to quit
{"version":1,"sender":"agent1qw7u5sw63a88kmcn5j5kxf7q326u5hgmppvy2vpxlh3re6y0yp8253ec7xl","target":"agent1qgd54rrq8ex4uhdxe6qg0sklz7h7dkacdk9rz4ec0l304wghw88sg35rfk6","session":"c28d03fd-ad42-4c09-80b8-2bb8df9d7c3e","schema_digest":"model:14d760ab9a6127711e530c0f1bd84d5caa48c6bc6566ca489581d6918e6dff85","protocol_digest":"proto:a03398ea81d7aaaf67e72940937676eae0d019f8e1d8b5efbadfef9fd2e98bb2","payload":"eyJtZXNzYWdlIjoiaGV5In0=","expires":null,"nonce":null,"signature":"sig14ukfdrc98924wvx66xa2c32pk2wqyn69cepkvcpwahlygv3tc2t8hrzfkuzc9earscpnasdt8txpu99cyw24gmyspfdhqkxsn7a3lnq60mpaq"}
INFO:fetchai:Got response looking up agent endpoint
http://localhost:8000/submit
INFO:fetchai:Sent message to agent
INFO:werkzeug:127.0.0.1 - - [30/Jan/2025 14:21:58] "POST /request HTTP/1.1" 200 -
INFO:__main__:Received response
INFO:__main__:Processed response: {'response': 'Hello, AI Agent! I received your message: Hello uAgent!'}
INFO:werkzeug:127.0.0.1 - - [30/Jan/2025 14:21:58] "POST /api/webhook HTTP/1.1" 200 -

Explanation of Communication Flow

• The AI Agent sends a message using the send_message_to_agent function.
• The uAgent receives the message via the @uagent.on_message handler.
• The uAgent processes the message and responds using await ctx.send().
• The AI Agent receives the response through the /api/webhook endpoint.

Key Takeaways

• The uAgent handles structured messages using Fetch.ai’s uAgents framework.
• The AI Agent utilizes Fetch.ai’s SDK for message transmission and parsing.
• Communication between the two agents follows a request-response pattern using their respective handlers.

This setup can be extended to build more complex agent interactions involving dynamic data exchange, service orchestration, and autonomous decision-making.