atomagent 0.1.0

A lightweight Python AI Agent framework with async support

⚛️ AtomAgent

AtomAgent is a lightweight Python AI Agent framework that supports both synchronous and asynchronous modes, providing a clean tool registration mechanism for building intelligent agents with function calling capabilities.

✨ Key Features

• 🔌 Minimal Dependencies: Only depends on pure Python libraries, no Rust dependencies
• 🎯 OpenAI API Compatible: Supports all OpenAI API compatible LLMs
• 🔧 Flexible Tool System: Easy tool registration with decorators
• 🚀 Async Support: Full async implementation for high concurrency
• 📡 Multi-Tool Calling: Supports OpenAI's latest tools protocol
• 🔄 Streaming Output: Real-time streaming responses
• 📝 Detailed Logging: Built-in hierarchical logging system

📦 Installation

# Basic installation
pip install atomagent

🚀 Quick Start

from atomagent import AsyncAgent

# Create an agent
agent = AsyncAgent(
    name="Assistant",
    model="gpt-3.5-turbo",
    api_key="your_api_key"
)

# Register a tool
@agent.tool()
def get_weather(city: str) -> str:
    """Get weather information for a city"""
    return f"The weather in {city} is sunny"

# Run the agent
import asyncio

async def main():
    response = await agent.run("What's the weather in Beijing?")
    print(response)

asyncio.run(main())

📖 Documentation

For detailed documentation and examples, visit: https://github.com/liuyunrui123/AtomAgent

🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

📄 License

This project is licensed under the MIT License.