multi-swarm 0.1.2

A framework for creating collaborative AI agents using multiple LLM providers

Multi-Swarm: Multi-LLM Agent Framework

A framework for creating collaborative AI agents using multiple LLM providers (Google's Gemini and Anthropic's Claude).

Features

• Multi-LLM Support: Leverage different LLM providers for specialized tasks
• Flexible Agent Architecture: Create custom agents with specific roles and capabilities
• Structured Communication: Define clear communication flows between agents
• Easy Integration: Simple API for creating and running agent swarms

Installation

1. Clone the repository
2. Install dependencies:

pip install -r requirements.txt

Environment Setup

Create a .env file in your project root with your API keys:

GOOGLE_API_KEY=your_gemini_api_key
ANTHROPIC_API_KEY=your_claude_api_key

Usage

Basic Example

from multi_swarm import Agency, CEOAgent, TrendsAnalyst

# Initialize agents
ceo = CEOAgent()  # Uses Gemini 2.0 Pro
analyst = TrendsAnalyst()  # Uses Claude 3.5 Sonnet

# Create agency with communication flows
agency = Agency(
    agents=[
        ceo,  # CEO is the entry point
        [ceo, analyst],  # CEO can delegate to analyst
    ],
    shared_instructions="agency_manifesto.md"
)

# Run the agency
agency.run_demo()

Custom Agent Creation

1. Create a new agent class:

from multi_swarm import BaseAgent

class CustomAgent(BaseAgent):
    def __init__(self):
        super().__init__(
            name="Custom Agent",
            description="Description of the agent's role",
            instructions="path/to/instructions.md",
            tools_folder="path/to/tools",
            model="model-name",  # gemini-2.0-pro or claude-3.5-sonnet
            temperature=0.7
        )

2. Create instructions for your agent in a markdown file
3. Add any custom tools in the agent's tools folder
4. Integrate the agent into your agency's communication flow

License

MIT License - see LICENSE file for details