AgentSys (experimental, educational)

AgentSys is currently an experimental sample framework intended to explore ergonomic interfaces for multi-agent systems. It is not intended to be used in production, and therefore has no official support. (This also means we will not be reviewing PRs or issues!)

The primary goal of AgentSys is to showcase the handoff & routines patterns explored in the Orchestrating Agents: Handoffs & Routines cookbook. It is not meant as a standalone library, and is primarily for educational purposes.

Installation

SSH

pip install git+ssh://git@github.com/lifsys/agentsys.git

HTTPS

pip install git+https://github.com/lifsys/agentsys.git

Project Structure

The project is organized into several key modules:

• agentsys.orchestration: Core orchestration functionality
  • Swarm: Main orchestration class for managing agent interactions
  • Agent: Base class for defining agents
• agentsys.models: Model interfaces and implementations
  • BaseModel: Abstract base class for model implementations
  • OpenAIModel: OpenAI model implementation
• agentsys.config: Configuration management
  • Settings: Configuration settings and utilities
• agentsys.types: Type definitions and data structures
• agentsys.util: Utility functions and helpers

Basic Usage

from agentsys import Swarm, Agent

client = Swarm()
agent = Agent(
    name="test",
    instructions="You are a helpful assistant.",
    model="gpt-4",
)

messages = [{"role": "user", "content": "Hello!"}]
response = client.run(agent, messages)
print(response.messages[-1]["content"])

Function Calling

AgentSys supports function calling with both OpenAI's function calling and tool calling APIs:

def get_weather(location: str) -> str:
    """Get the weather for a location."""
    return f"The weather in {location} is sunny!"

agent = Agent(
    name="weather",
    instructions="You can help users check the weather.",
    model="gpt-4",
    functions=[get_weather],
)

messages = [{"role": "user", "content": "What's the weather in San Francisco?"}]
response = client.run(agent, messages)
print(response.messages[-1]["content"])

Configuration

You can configure AgentSys using environment variables or by passing a config object:

from agentsys import Settings

settings = Settings(
    openai_api_key="your-api-key",
    temperature=0.7,
)

Swarm Client

The Swarm class is the main entry point for interacting with agents. It handles:

• Message routing
• Function calling
• Response streaming
• Error handling

Contributing

This is an experimental project and we are not accepting contributions at this time.

License

MIT