Defining and Running Agents

Defining Agents

To define any agent system you will always use the same imports:

from any_agent import AgentConfig, AnyAgent
# In these examples, the built-in tools will be used
from any_agent.tools import search_web, visit_webpage

Check AgentConfig for more info on how to configure agents.

Single Agent

agent = AnyAgent.create(
    "openai",  # See other options under `Frameworks`
    AgentConfig(
        model_id="gpt-4.1-nano",
        instructions="Use the tools to find an answer",
        tools=[search_web, visit_webpage]
    ),
)

Multi-Agent

Warning

A multi-agent system introduces even more complexity than a single agent.

As stated before, carefully consider whether you need to adopt this pattern to solve the task.

Multi-agent can be implemented today using the A2A protocol (see A2A docs) and will be also supported with Agent-As-Tools (follow progress at https://github.com/mozilla-ai/any-agent/issues/382).

Framework Specific Arguments

The agent_args parameter in AgentConfig allows you to pass arguments specific to the underlying framework that the agent instance is built on.

Example-1: To pass the output_type parameter for structured output, when using the OpenAI Agents SDK:

from pydantic import BaseModel
from any_agent import AgentConfig, AgentFramework, AnyAgent

class BookInfo(BaseModel):
    title: str
    author: str
    publication_year: int

framework = AgentFramework.OPENAI

agent = AnyAgent.create(
    framework,
    AgentConfig(
        model_id="gpt-4.1-mini",
        instructions="Extract book information from text",
        agent_args={
            "output_type": BookInfo
        }
    )
)

Example-2: In smolagents, for structured output one needs to use the grammar parameter. Additionally, planning_interval defines the interval at which the agent will run a planning step.

from pydantic import BaseModel
from any_agent import AgentConfig, AgentFramework, AnyAgent


framework = AgentFramework.SMOLAGENTS

class WebPageInfo(BaseModel):
    title: str
    summary: str

agent = AnyAgent.create(
    framework,
    AgentConfig(
        model_id="gpt-4.1-mini",
        instructions="Extract webpage title and summary from url",
        agent_args={
            "planning_interval": 1,
            "grammar": WebPageInfo
        }
    )
)

Running Agents

agent_trace = agent.run("Which Agent Framework is the best??")
print(agent_trace.final_output)

Check AgentTrace for more info on the return type.

Async

If you are running in async context, you should use the equivalent create_async and run_async methods:

import asyncio

async def main():
    agent = await AnyAgent.create_async(
        "openai",
        AgentConfig(
            model_id="gpt-4.1-mini",
            instructions="Use the tools to find an answer",
            tools=[search_web, visit_webpage]
        )
    )

    agent_trace = await agent.run_async("Which Agent Framework is the best??")
    print(agent_trace.final_output)

if __name__ == "__main__":
    asyncio.run(main())

Cleaning up the Agent

When an AnyAgent object is deleted, the python garbage collector cleans up any resources owned by the object. However, when running or re-creating an agent in the same python process (for example, in test scripts) it may be necessary to forcefully shut down the agent to avoid unexpected side affects. For this purpose, agent.exit is available which will shut down all resources the agent was using.

For example,

agent.run("Which agent framework is the best?")
agent.exit() # cleans up the agent synchronously