Python: Add timeout parameter to FoundryAgent to fix ConnectTimeout on multi-turn conversations (#6263)

* Python: fix ConnectTimeout on multi-turn FoundryAgent conversations (#6241)

Expose a `timeout` parameter on `RawFoundryAgentChatClient`,
`_FoundryAgentChatClient`, `RawFoundryAgent`, `FoundryAgent`, and
`RawOpenAIChatClient` so callers can override the HTTP timeout used by
the underlying AsyncOpenAI client.

Root cause: `RawFoundryAgentChatClient.__init__` called
`project_client.get_openai_client()` without configuring any timeout,
inheriting the OpenAI SDK default of `httpx.Timeout(connect=5.0)`.
When connections are recycled between turns under load, the 5 s connect
timeout fires and surfaces as `openai.APITimeoutError`.

Fix:
- `load_openai_service_settings` (`_shared.py`): accept `timeout` and
  include it in `client_args` for all three `AsyncOpenAI`/
  `AsyncAzureOpenAI` construction paths.
- `RawOpenAIChatClient.__init__` (`_chat_client.py`): accept `timeout`
  and forward to `load_openai_service_settings`.
- `RawFoundryAgentChatClient.__init__` (`_agent.py`): accept `timeout`
  and set `openai_client.timeout = timeout` on the client returned by
  `get_openai_client()` before passing it to the base class.
- `_FoundryAgentChatClient`, `RawFoundryAgent`, `FoundryAgent`: accept
  and propagate `timeout` through the construction chain.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* Add timeout parameter to FoundryAgent and RawOpenAIChatClient

Expose a timeout parameter on RawFoundryAgentChatClient,
_FoundryAgentChatClient, RawFoundryAgent, FoundryAgent, and
RawOpenAIChatClient. When provided, the value is applied to the
underlying AsyncOpenAI client so that connect timeouts under load
or after connection recycling can be tuned by callers.

Previously, get_openai_client() was called without any timeout
override, so the SDK default of httpx.Timeout(connect=5.0) was
inherited and could fire on multi-turn conversations where the
underlying connection is recycled between turns.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* Python: Add `timeout` parameter to `FoundryAgent` to fix `ConnectTimeout` on multi-turn conversations

Fixes #6241

* fix(foundry): use with_options to avoid mutating shared OpenAI client timeout (#6241)

Replace direct assignment  with
 in
RawFoundryAgentChatClient.__init__.

The Azure AI Projects SDK caches and returns a shared AsyncOpenAI client
per AIProjectClient. Mutating its .timeout attribute leaked the override
to all other code paths sharing that client (other agents, user code).
with_options() returns a new client instance with the override applied,
leaving the original shared client untouched.

Update tests to assert with_options is called with the correct timeout
and that the original shared client's timeout attribute is not mutated.

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

* test(foundry): assert with_options return value flows to instance.client (#6241)

The four timeout propagation tests verified that with_options was called
but did not confirm that the returned (timeout-configured) client was
actually stored on the instance. A silent discard of the return value
would have left the tests green while the timeout had no effect.

Each test now captures the constructed instance and asserts:
  assert <instance>.client is openai_client_mock.with_options.return_value

Affected tests:
- test_raw_foundry_agent_chat_client_init_applies_timeout_to_openai_client
- test_raw_foundry_agent_chat_client_init_applies_timeout_with_preview_enabled
- test_foundry_agent_chat_client_init_propagates_timeout
- test_foundry_agent_init_propagates_timeout_to_openai_client

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

---------

Co-authored-by: Copilot <copilot@github.com>
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

python/packages/openai/agent_framework_openai/_chat_client.py

@@ -385,6 +385,7 @@ class RawOpenAIChatClient(  # type: ignore[misc]
         additional_properties: dict[str, Any] | None = None,
         env_file_path: str | None = None,
         env_file_encoding: str | None = None,
+        timeout: float | None = None,
     ) -> None:
         """Initialize a raw OpenAI Chat client.
 
@@ -406,6 +407,7 @@ class RawOpenAIChatClient(  # type: ignore[misc]
             env_file_path: Optional ``.env`` file that is checked before the process environment
                 for ``OPENAI_*`` values.
             env_file_encoding: Encoding for the ``.env`` file.
+            timeout: Optional timeout in seconds for requests.
         """
         ...
 
@@ -427,6 +429,7 @@ class RawOpenAIChatClient(  # type: ignore[misc]
         additional_properties: dict[str, Any] | None = None,
         env_file_path: str | None = None,
         env_file_encoding: str | None = None,
+        timeout: float | None = None,
     ) -> None:
         """Initialize a raw OpenAI Chat client.
 
@@ -455,6 +458,7 @@ class RawOpenAIChatClient(  # type: ignore[misc]
             env_file_path: Optional ``.env`` file that is checked before process environment
                 variables for ``AZURE_OPENAI_*`` values.
             env_file_encoding: Encoding for the ``.env`` file.
+            timeout: Optional timeout in seconds for requests.
         """
         ...
 
@@ -476,6 +480,7 @@ class RawOpenAIChatClient(  # type: ignore[misc]
         additional_properties: dict[str, Any] | None = None,
         env_file_path: str | None = None,
         env_file_encoding: str | None = None,
+        timeout: float | None = None,
     ) -> None:
         """Initialize a raw OpenAI Chat client.
 
@@ -511,6 +516,8 @@ class RawOpenAIChatClient(  # type: ignore[misc]
                 variables. The same file is used for both ``OPENAI_*`` and ``AZURE_OPENAI_*``
                 lookups.
             env_file_encoding: Encoding for the ``.env`` file.
+            timeout: HTTP timeout in seconds for requests. When not provided, the
+                OpenAI SDK default is used (connect: 5s, total: 600s).
 
         Notes:
             Environment resolution and routing precedence are:
@@ -541,6 +548,7 @@ class RawOpenAIChatClient(  # type: ignore[misc]
             openai_model_fields=("chat_model", "model"),
             azure_model_fields=("chat_model", "model"),
             responses_mode=True,
+            timeout=timeout,
         )
 
         self.client = client

python/packages/openai/agent_framework_openai/_shared.py

@@ -162,6 +162,7 @@ def load_openai_service_settings(
     openai_model_fields: Sequence[OpenAIModelSettingName] = ("model",),
     azure_model_fields: Sequence[OpenAIModelSettingName] = ("model",),
     responses_mode: bool = False,
+    timeout: float | None = None,
 ) -> tuple[dict[str, Any], AsyncOpenAI, bool]:
     """Load OpenAI settings, including Azure OpenAI model aliases.
 
@@ -218,6 +219,8 @@ def load_openai_service_settings(
             }
             if base_url := openai_settings.get("base_url"):
                 client_args["base_url"] = base_url
+            if timeout is not None:
+                client_args["timeout"] = timeout
             return openai_settings, AsyncOpenAI(**client_args), False  # type: ignore[return-value]
         checked_openai = True
     azure_settings = load_settings(
@@ -299,8 +302,12 @@ def load_openai_service_settings(
             openai_args["api_key"] = _ensure_async_token_provider(client_args["azure_ad_token_provider"])
         elif "api_key" in client_args:
             openai_args["api_key"] = client_args["api_key"]
+        if timeout is not None:
+            openai_args["timeout"] = timeout
         return azure_settings, AsyncOpenAI(**openai_args), True  # type: ignore[return-value]
 
+    if timeout is not None:
+        client_args["timeout"] = timeout
     return azure_settings, AsyncAzureOpenAI(**client_args), True  # type: ignore[return-value]
 
 

python/packages/openai/tests/openai/test_openai_chat_client.py

@@ -36,7 +36,7 @@ from agent_framework.exceptions import (
     ChatClientInvalidRequestException,
     SettingNotFoundError,
 )
-from openai import BadRequestError
+from openai import AsyncOpenAI, BadRequestError
 from openai.types.responses.response_reasoning_item import Summary
 from openai.types.responses.response_reasoning_summary_text_delta_event import (
     ResponseReasoningSummaryTextDeltaEvent,
@@ -55,7 +55,7 @@ from pydantic import BaseModel
 from pytest import param
 
 from agent_framework_openai import OpenAIChatClient
-from agent_framework_openai._chat_client import OPENAI_LOCAL_SHELL_CALL_ITEM_ID_KEY
+from agent_framework_openai._chat_client import OPENAI_LOCAL_SHELL_CALL_ITEM_ID_KEY, RawOpenAIChatClient
 from agent_framework_openai._exceptions import OpenAIContentFilterException
 
 skip_if_openai_integration_tests_disabled = pytest.mark.skipif(
@@ -194,6 +194,26 @@ def test_init_uses_explicit_parameters() -> None:
     assert all(parameter.kind != inspect.Parameter.VAR_KEYWORD for parameter in signature.parameters.values())
 
 
+def test_raw_openai_chat_client_init_uses_explicit_parameters() -> None:
+    signature = inspect.signature(RawOpenAIChatClient.__init__)
+
+    assert "additional_properties" in signature.parameters
+    assert "compaction_strategy" in signature.parameters
+    assert "tokenizer" in signature.parameters
+    assert "timeout" in signature.parameters
+    assert all(parameter.kind != inspect.Parameter.VAR_KEYWORD for parameter in signature.parameters.values())
+
+
+def test_raw_openai_chat_client_accepts_preconfigured_client_with_timeout() -> None:
+    """Test that timeout is accepted without error when async_client is pre-provided."""
+
+    mock_client = MagicMock(spec=AsyncOpenAI)
+    mock_client.timeout = 5.0
+
+    client = RawOpenAIChatClient(async_client=mock_client, timeout=30.0)
+    assert client is not None
+
+
 def test_openai_chat_client_supports_all_tool_protocols() -> None:
     assert isinstance(OpenAIChatClient, SupportsCodeInterpreterTool)
     assert isinstance(OpenAIChatClient, SupportsWebSearchTool)