AutoGen Integration

AutoGen Integration

Microsoft AutoGen supports OpenAI-compatible endpoints, making it straightforward to use Forge as your LLM backend for multi-agent conversations.

Setup

pip install autogen-agentchat

Configure

import autogen

config_list = [
    {
        "model": "auto",
        "api_key": "forge_sk_your_key",
        "base_url": "https://api.optima-forge.com/v1",
    }
]

llm_config = {"config_list": config_list, "temperature": 0.7}

assistant = autogen.AssistantAgent(
    name="assistant",
    llm_config=llm_config,
    system_message="You are a helpful AI assistant.",
)

user_proxy = autogen.UserProxyAgent(
    name="user_proxy",
    human_input_mode="NEVER",
    max_consecutive_auto_reply=3,
)

user_proxy.initiate_chat(
    assistant,
    message="Write a Python function that calculates Fibonacci numbers.",
)

Multi-Agent Conversations

AutoGen's group chat feature works with Forge. Each agent in the group can be routed to different providers based on their role and the complexity of their contributions. Forge tracks the full conversation through Langfuse tracing.