Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.galadriel.com/llms.txt

Use this file to discover all available pages before exploring further.

This tutorial guides you through creating a multi-agent system using Galadriel, demonstrating a manager-worker pattern where a manager agent coordinates with a specialized worker agent to accomplish tasks. This setup allows for efficient delegation and specialization, improving overall system performance.

​
Prerequisites

Before you begin, ensure you have:
  • Python 3.10 or higher
  • Galadriel installed
  • An OpenAI API key

​
Setup

Create a project directory: 
mkdir multi-agents-example
cd multi-agents-example
Set up a virtual environment: 
python3 -m venv venv
source venv/bin/activate
Install Galadriel: 
pip install galadriel

​
Code Implementation

​
1. Project Structure

Create the following files:
  • agent.py: Contains the agent definitions and runtime setup.
  • template.env: Stores environment variables.

​
2. Define Environment Variables

Create a .env file with your OpenAI API key: 
echo "OPENAI_API_KEY=your_api_key_here" > .env

​
3. Implement the Agents

Open agent.py and add the following code: 
import asyncio
import os
from pathlib import Path
from dotenv import load_dotenv
from galadriel import AgentRuntime, CodeAgent
from galadriel.clients import SimpleMessageClient
from galadriel.core_agent import LiteLLMModel, DuckDuckGoSearchTool

# Load environment variables
load_dotenv(dotenv_path=Path(".") / ".env", override=True)

model = LiteLLMModel(model_id="gpt-4o", api_key=os.getenv("OPENAI_API_KEY"))

# Create the worker agent
managed_web_agent = CodeAgent(
    tools=[DuckDuckGoSearchTool()],
    model=model,
    name="web_search",
    description="Runs web searches for you. Give it your query as an argument.",
)

# Create the manager agent
manager_agent = CodeAgent(tools=[], model=model, managed_agents=[managed_web_agent])

# Set up the client
client = SimpleMessageClient("What's the most recent of Daige on X (twitter)?")

# Set up the runtime
runtime = AgentRuntime(
    agent=manager_agent,
    inputs=[client],
    outputs=[client],
)

# Run the agent
asyncio.run(runtime.run())

​
Run the Example

Set the OPENAI_API_KEY in your environment: 
export OPENAI_API_KEY=your_api_key_here
Run the agent: 
python agent.py

​
Expected Output

The agent will perform a web search to find the most recent tweets from Daige on X (Twitter) and provide a summary.

​
Conclusion

This tutorial demonstrated how to create a multi-agent system using Galadriel. By creating a manager agent that delegates tasks to a specialized worker agent, you can build more efficient and modular AI systems.