External Tool Execution Async Responses

This example demonstrates external tool execution using OpenAI Responses API with gpt-4.1-mini model. It shows how to handle tool-call IDs and execute multiple external tools in a loop until completion.

Code

external_tool_execution_async_responses.py

"""🤝 Human-in-the-Loop with OpenAI Responses API (gpt-4.1-mini)

This example mirrors the external tool execution async example but uses
OpenAIResponses with gpt-4.1-mini to validate tool-call id handling.

Run `pip install openai agno` to install dependencies.
"""

import asyncio
import subprocess

from agno.agent import Agent
from agno.models.openai import OpenAIResponses
from agno.tools import tool
from agno.utils import pprint


# We have to create a tool with the correct name, arguments and docstring
# for the agent to know what to call.
@tool(external_execution=True)
def execute_shell_command(command: str) -> str:
    """Execute a shell command.

    Args:
        command (str): The shell command to execute

    Returns:
        str: The output of the shell command
    """
    if (
        command.startswith("ls ")
        or command == "ls"
        or command.startswith("cat ")
        or command.startswith("head ")
    ):
        return subprocess.check_output(command, shell=True).decode("utf-8")
    raise Exception(f"Unsupported command: {command}")


agent = Agent(
    model=OpenAIResponses(id="gpt-4.1-mini"),
    tools=[execute_shell_command],
    markdown=True,
)

run_response = asyncio.run(agent.arun("What files do I have in my current directory?"))

# Keep executing externally-required tools until the run completes
while (
    run_response.is_paused and len(run_response.tools_awaiting_external_execution) > 0
):
    for external_tool in run_response.tools_awaiting_external_execution:
        if external_tool.tool_name == execute_shell_command.name:
            print(
                f"Executing {external_tool.tool_name} with args {external_tool.tool_args} externally"
            )
            result = execute_shell_command.entrypoint(**external_tool.tool_args)
            external_tool.result = result
        else:
            print(f"Skipping unsupported external tool: {external_tool.tool_name}")

    run_response = asyncio.run(agent.acontinue_run(run_response=run_response))

pprint.pprint_run_response(run_response)


# Or for simple debug flow
# agent.print_response("What files do I have in my current directory?")

Usage

1. Create a virtual environment

   Open the Terminal and create a python virtual environment.

   Mac

   python3 -m venv .venv
   source .venv/bin/activate
   
2. Install libraries

   pip install -U agno openai
   
3. Export your OpenAI API key

   Mac/Linux

     export OPENAI_API_KEY="your_openai_api_key_here"
   
4. Create a Python file

   Create a Python file and add the above code.

   touch external_tool_execution_async_responses.py
   
5. Run Agent

   Mac

   python external_tool_execution_async_responses.py
   
6. Find All Cookbooks

   Explore all the available cookbooks in the Agno repository. Click the link below to view the code on GitHub:

   Agno Cookbooks on GitHub