Skip to main content

Lab 16: Implementing the "Greeting Router" Challenge

Goal​

In this lab, you will implement the multi-agent "Greeting Router" system. You will create a main router agent and a specialist spanish_greeter agent, and configure them to work together using the coordinator/dispatcher pattern.

Step 1: Create the Project Structure​

  1. Create the agent project: 
    adk create greeting-agent
    cd greeting-agent

Step 2: Create the Specialist Agent​

Exercise: Create a new file named spanish_greeter_agent.py. In this file, use the skeleton below to define the configuration for the spanish_greeter_agent.

# In spanish_greeter_agent.py
from google.adk.agents import LlmAgent

agent = LlmAgent(
    name="spanish_greeter_agent",
    model="gemini-2.5-flash",
    description="""
# TODO: Write a clear description that explains this agent's capability
# (e.g., "An expert at providing friendly greetings in Spanish.").
# This is critical for the router to find it.
""",
    instruction="""
# TODO: Write an instruction that tells the agent its only job is to
# provide a warm greeting in Spanish.
"""
)

Step 3: Configure the Coordinator (Router) Agent​

Exercise: Now, open the main agent.py file and use the skeleton below to configure it to act as the router.

# In agent.py
from google.adk.agents import LlmAgent

# TODO: Import the `agent` object from your `spanish_greeter_agent` module.

root_agent = LlmAgent(
    name="router_agent",
    model="gemini-2.5-flash",
    description="The main greeter agent that routes to language specialists.",
    instruction="""
# TODO: Write an instruction that tells the agent its job is to delegate
# to the `spanish_greeter_agent` for Spanish greetings and that it
# should *not* greet the user itself.
""",
    # TODO: Add the `sub_agents` list and include the imported spanish greeter agent.
    sub_agents=[]
)

Step 4: Test the Multi-Agent System​

  1. Set up your .env file.
  2. Navigate to the parent directory (cd ..) and start the Dev UI: 
    adk web greeting-agent
  3. Interact with the router:
    • Test Case 1: "Say hello to me in Spanish."
    • Test Case 2: "Say hello in French."
  4. Examine the Trace: For the first test case, you should see the router_agent run, call transfer_to_agent, and then the spanish_greeter_agent run to produce the final output. For the second case, the router should respond itself that it cannot handle the request.

Having Trouble?​

If you get stuck, you can find the complete, working Python code in the lab-solution.md file.

Lab Summary​

You have successfully built your first multi-agent system! You have learned to:

  • Create separate Python files for different agents.
  • Use the sub_agents key to establish a parent-child hierarchy in Python.
  • Write a clear description for a specialist agent so the coordinator can understand its function.
  • Write an instruction for a coordinator agent that tells it how to delegate tasks.
  • Verify the agent transfer process by inspecting the Trace View.

Self-Reflection Questions​

  • What do you think would happen if you forgot to add the description to the spanish_greeter_agent? How would the router_agent behave?
  • In the router_agent's instruction, why is it important to explicitly tell it not to greet the user itself?
  • How does breaking the logic into a router and a specialist make the system easier to extend in the future (e.g., to add a french_greeter_agent)?

🕵️ Hidden Solution 🕵️​

Looking for the solution? Here's a hint (Base64 decode me): L2RvYy1hZGstdHJhaW5pbmcvbW9kdWxlMTYtY29vcmRpbmF0b3ItYWdlbnQvbGFiLXNvbHV0aW9u

The direct link is: Lab Solution