A programming framework for agentic AI
Make DockerCommandLineCodeExecutor the default for MagenticOne team (#6684)
Copilotca3d693b19e4
diff --git a/python/packages/autogen-ext/src/autogen_ext/teams/magentic_one.py b/python/packages/autogen-ext/src/autogen_ext/teams/magentic_one.pyindex ec2d5192a..12de81328 100644--- a/python/packages/autogen-ext/src/autogen_ext/teams/magentic_one.py+++ b/python/packages/autogen-ext/src/autogen_ext/teams/magentic_one.py@@ -14,11 +14,61 @@ from autogen_ext.agents.web_surfer import MultimodalWebSurferfrom autogen_ext.code_executors.local import LocalCommandLineCodeExecutorfrom autogen_ext.models.openai._openai_client import BaseOpenAIChatCompletionClient+# Docker imports for default code executor+try:+ import docker+ from docker.errors import DockerException++ from autogen_ext.code_executors.docker import DockerCommandLineCodeExecutor++ _docker_available = True+except ImportError:+ docker = None # type: ignore+ DockerException = Exception # type: ignore+ DockerCommandLineCodeExecutor = None # type: ignore+ _docker_available = False+SyncInputFunc = Callable[[str], str]AsyncInputFunc = Callable[[str, Optional[CancellationToken]], Awaitable[str]]InputFuncType = Union[SyncInputFunc, AsyncInputFunc]+def _is_docker_available() -> bool:+ """Check if Docker is available and running."""+ if not _docker_available:+ return False++ try:+ if docker is not None:+ client = docker.from_env()+ client.ping() # type: ignore+ return True+ except DockerException:+ return False++ return False+++def _create_default_code_executor() -> CodeExecutor:+ """Create the default code executor, preferring Docker if available."""+ if _is_docker_available() and DockerCommandLineCodeExecutor is not None:+ try:+ return DockerCommandLineCodeExecutor()+ except Exception:+ # Fallback to local if Docker fails to initialize+ pass++ # Issue warning and use local executor if Docker is not available+ warnings.warn(+ "Docker is not available or not running. Using LocalCommandLineCodeExecutor instead of the recommended DockerCommandLineCodeExecutor. "+ "For security, it is recommended to install Docker and ensure it's running before using MagenticOne. "+ "To install Docker, visit: https://docs.docker.com/get-docker/",+ UserWarning,+ stacklevel=3,+ )+ return LocalCommandLineCodeExecutor()++class MagenticOne(MagenticOneGroupChat):"""MagenticOne is a specialized group chat class that integrates various agents@@ -77,7 +127,7 @@ class MagenticOne(MagenticOneGroupChat):async def example_usage():client = OpenAIChatCompletionClient(model="gpt-4o")- m1 = MagenticOne(client=client)+ m1 = MagenticOne(client=client) # Uses DockerCommandLineCodeExecutor by defaulttask = "Write a Python script to fetch data from an API."result = await Console(m1.run_stream(task=task))print(result)@@ -89,20 +139,22 @@ class MagenticOne(MagenticOneGroupChat):.. code-block:: python- # Enable human-in-the-loop mode+ # Enable human-in-the-loop mode with explicit Docker executorimport asynciofrom autogen_ext.models.openai import OpenAIChatCompletionClientfrom autogen_ext.teams.magentic_one import MagenticOne+ from autogen_ext.code_executors.docker import DockerCommandLineCodeExecutorfrom autogen_agentchat.ui import Consoleasync def example_usage_hil():client = OpenAIChatCompletionClient(model="gpt-4o")- # to enable human-in-the-loop mode, set hil_mode=True- m1 = MagenticOne(client=client, hil_mode=True)- task = "Write a Python script to fetch data from an API."- result = await Console(m1.run_stream(task=task))- print(result)+ # Explicitly specify Docker code executor for better security+ async with DockerCommandLineCodeExecutor() as code_executor:+ m1 = MagenticOne(client=client, hil_mode=True, code_executor=code_executor)+ task = "Write a Python script to fetch data from an API."+ result = await Console(m1.run_stream(task=task))+ print(result)if __name__ == "__main__":@@ -134,11 +186,11 @@ class MagenticOne(MagenticOneGroupChat):if code_executor is None:warnings.warn(- "Instantiating MagenticOne without a code_executor is deprecated. Provide a code_executor to clear this warning (e.g., code_executor=LocalCommandLineCodeExecutor() ).",+ "Instantiating MagenticOne without a code_executor is deprecated. Provide a code_executor to clear this warning (e.g., code_executor=DockerCommandLineCodeExecutor() ).",DeprecationWarning,stacklevel=2,)- code_executor = LocalCommandLineCodeExecutor()+ code_executor = _create_default_code_executor()fs = FileSurfer("FileSurfer", model_client=client)ws = MultimodalWebSurfer("WebSurfer", model_client=client)diff --git a/python/packages/autogen-ext/tests/teams/__init__.py b/python/packages/autogen-ext/tests/teams/__init__.pynew file mode 100644index 000000000..5d2a4d079--- /dev/null+++ b/python/packages/autogen-ext/tests/teams/__init__.py@@ -0,0 +1 @@+"""Init file for teams tests."""diff --git a/python/packages/autogen-ext/tests/teams/test_magentic_one.py b/python/packages/autogen-ext/tests/teams/test_magentic_one.pynew file mode 100644index 000000000..d0c443d52--- /dev/null+++ b/python/packages/autogen-ext/tests/teams/test_magentic_one.py@@ -0,0 +1,138 @@+"""Tests for MagenticOne team."""++import os+import warnings+from unittest.mock import Mock, patch++import pytest+from autogen_agentchat.agents import CodeExecutorAgent+from autogen_core.models import ChatCompletionClient+from autogen_ext.code_executors.local import LocalCommandLineCodeExecutor+from autogen_ext.teams.magentic_one import MagenticOne+++def docker_tests_enabled() -> bool:+ """Check if Docker tests should be enabled."""+ if os.environ.get("SKIP_DOCKER", "unset").lower() == "true":+ return False++ try:+ import docker+ from docker.errors import DockerException+ except ImportError:+ return False++ try:+ client = docker.from_env()+ client.ping() # type: ignore+ return True+ except DockerException:+ return False+++def _is_docker_available() -> bool:+ """Local implementation of Docker availability check."""+ return docker_tests_enabled()+++def mock_chat_client() -> Mock:+ """Create a mock chat completion client."""+ mock_client = Mock(spec=ChatCompletionClient)+ mock_client.model_info = {"function_calling": True, "json_output": True, "vision": True}+ return mock_client++[email protected](not docker_tests_enabled(), reason="Docker is not available")+def test_magentic_one_uses_docker_by_default(mock_chat_client: Mock) -> None:+ """Test that MagenticOne uses Docker code executor by default when Docker is available."""+ from autogen_ext.code_executors.docker import DockerCommandLineCodeExecutor++ with warnings.catch_warnings():+ warnings.simplefilter("ignore", DeprecationWarning)++ m1 = MagenticOne(client=mock_chat_client)++ # Find the CodeExecutorAgent in the participants list+ code_executor_agent = None+ for agent in m1._participants: # type: ignore[reportPrivateUsage]+ if isinstance(agent, CodeExecutorAgent):+ code_executor_agent = agent+ break++ assert code_executor_agent is not None, "CodeExecutorAgent not found"+ assert isinstance(+ code_executor_agent._code_executor, # type: ignore[reportPrivateUsage]+ DockerCommandLineCodeExecutor, # type: ignore[reportPrivateUsage]+ ), f"Expected DockerCommandLineCodeExecutor, got {type(code_executor_agent._code_executor)}" # type: ignore[reportPrivateUsage]+++def test_docker_availability_check() -> None:+ """Test the Docker availability check function."""+ # This test should pass regardless of Docker availability+ result = _is_docker_available()+ assert isinstance(result, bool)+++@patch("autogen_ext.teams.magentic_one._is_docker_available")+def test_magentic_one_falls_back_to_local_when_docker_unavailable(+ mock_docker_check: Mock, mock_chat_client: Mock+) -> None:+ """Test that MagenticOne falls back to local executor when Docker is not available."""+ mock_docker_check.return_value = False++ with warnings.catch_warnings(record=True) as w:+ warnings.simplefilter("always")++ m1 = MagenticOne(client=mock_chat_client)++ # Find the CodeExecutorAgent in the participants list+ code_executor_agent = None+ for agent in m1._participants: # type: ignore[reportPrivateUsage]+ if isinstance(agent, CodeExecutorAgent):+ code_executor_agent = agent+ break++ assert code_executor_agent is not None, "CodeExecutorAgent not found"+ assert isinstance(+ code_executor_agent._code_executor, # type: ignore[reportPrivateUsage]+ LocalCommandLineCodeExecutor, # type: ignore[reportPrivateUsage]+ ), f"Expected LocalCommandLineCodeExecutor, got {type(code_executor_agent._code_executor)}" # type: ignore[reportPrivateUsage]++ # Check that appropriate warnings were issued+ warning_messages = [str(warning.message) for warning in w]+ docker_warning_found = any("Docker is not available" in msg for msg in warning_messages)+ deprecated_warning_found = any(+ "Instantiating MagenticOne without a code_executor is deprecated" in msg for msg in warning_messages+ )++ assert docker_warning_found, f"Docker unavailable warning not found in: {warning_messages}"+ assert deprecated_warning_found, f"Deprecation warning not found in: {warning_messages}"+++def test_magentic_one_with_explicit_code_executor(mock_chat_client: Mock) -> None:+ """Test that MagenticOne uses the provided code executor when explicitly given."""+ explicit_executor = LocalCommandLineCodeExecutor()++ with warnings.catch_warnings(record=True) as w:+ warnings.simplefilter("always")++ m1 = MagenticOne(client=mock_chat_client, code_executor=explicit_executor)++ # Find the CodeExecutorAgent in the participants list+ code_executor_agent = None+ for agent in m1._participants: # type: ignore[reportPrivateUsage]+ if isinstance(agent, CodeExecutorAgent):+ code_executor_agent = agent+ break++ assert code_executor_agent is not None, "CodeExecutorAgent not found"+ assert code_executor_agent._code_executor is explicit_executor, "Expected the explicitly provided code executor" # type: ignore[reportPrivateUsage]++ # No deprecation warning should be issued when explicitly providing a code executor+ warning_messages = [str(warning.message) for warning in w]+ deprecated_warning_found = any(+ "Instantiating MagenticOne without a code_executor is deprecated" in msg for msg in warning_messages+ )++ assert not deprecated_warning_found, f"Unexpected deprecation warning found: {warning_messages}"