Python: [BREAKING] Remove deprecated kwargs compatibility paths (#4858)

* [BREAKING] Remove deprecated kwargs compatibility paths

Remove the deprecated kwargs compatibility shims across core agents, clients, tools, middleware, and telemetry.

Keep workflow kwargs behavior intact in this branch and follow up separately in #4850.

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

* Fix PR CI fallout for kwargs removal

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

* Address PR review feedback

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

* updates

* Fix Azure AI CI fallout

Remove the stale _get_current_conversation_id override from the Azure AI client after the OpenAI base helper was deleted.

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

* fixed new classes

* Fix Assistants deprecated import gating

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

* Fix integration replay regressions

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

* Switch multi-agent hosting samples to Azure chat completions

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

* Simplify Azure multi-agent sample config

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

---------

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
This commit is contained in:
Eduard van Valkenburg
2026-03-27 22:00:12 +01:00
committed by GitHub
Unverified
parent ca6cdd142e
commit b1b528e4a8
52 changed files with 1136 additions and 971 deletions
@@ -3,15 +3,21 @@
from __future__ import annotations
import sys
from collections.abc import Sequence
from typing import Any, Generic
from collections.abc import Awaitable, Callable, Mapping, Sequence
from typing import Any, Generic, Literal, cast, overload
from agent_framework import (
ChatAndFunctionMiddlewareTypes,
ChatMiddlewareLayer,
ChatOptions,
ChatResponse,
ChatResponseUpdate,
CompactionStrategy,
FunctionInvocationConfiguration,
FunctionInvocationLayer,
Message,
ResponseStream,
TokenizerProtocol,
)
from agent_framework._settings import load_settings
from agent_framework.observability import ChatTelemetryLayer
@@ -122,8 +128,8 @@ class FoundryLocalSettings(TypedDict, total=False):
'FOUNDRY_LOCAL_'.
Keys:
model_id: The name of the model deployment to use.
(Env var FOUNDRY_LOCAL_MODEL_ID)
model: The name of the model deployment to use.
(Env var FOUNDRY_LOCAL_MODEL)
"""
model: str | None
@@ -138,6 +144,78 @@ class FoundryLocalClient(
):
"""Foundry Local Chat completion class with middleware, telemetry, and function invocation support."""
@overload
def get_response(
self,
messages: Sequence[Message],
*,
stream: Literal[False] = ...,
options: ChatOptions[ResponseModelT],
compaction_strategy: CompactionStrategy | None = None,
tokenizer: TokenizerProtocol | None = None,
function_invocation_kwargs: Mapping[str, Any] | None = None,
client_kwargs: Mapping[str, Any] | None = None,
middleware: Sequence[ChatAndFunctionMiddlewareTypes] | None = None,
) -> Awaitable[ChatResponse[ResponseModelT]]: ...
@overload
def get_response(
self,
messages: Sequence[Message],
*,
stream: Literal[False] = ...,
options: FoundryLocalChatOptionsT | ChatOptions[None] | None = None,
compaction_strategy: CompactionStrategy | None = None,
tokenizer: TokenizerProtocol | None = None,
function_invocation_kwargs: Mapping[str, Any] | None = None,
client_kwargs: Mapping[str, Any] | None = None,
middleware: Sequence[ChatAndFunctionMiddlewareTypes] | None = None,
) -> Awaitable[ChatResponse[Any]]: ...
@overload
def get_response(
self,
messages: Sequence[Message],
*,
stream: Literal[True],
options: FoundryLocalChatOptionsT | ChatOptions[Any] | None = None,
compaction_strategy: CompactionStrategy | None = None,
tokenizer: TokenizerProtocol | None = None,
function_invocation_kwargs: Mapping[str, Any] | None = None,
client_kwargs: Mapping[str, Any] | None = None,
middleware: Sequence[ChatAndFunctionMiddlewareTypes] | None = None,
) -> ResponseStream[ChatResponseUpdate, ChatResponse[Any]]: ...
def get_response(
self,
messages: Sequence[Message],
*,
stream: bool = False,
options: FoundryLocalChatOptionsT | ChatOptions[Any] | None = None,
compaction_strategy: CompactionStrategy | None = None,
tokenizer: TokenizerProtocol | None = None,
function_invocation_kwargs: Mapping[str, Any] | None = None,
client_kwargs: Mapping[str, Any] | None = None,
middleware: Sequence[ChatAndFunctionMiddlewareTypes] | None = None,
) -> Awaitable[ChatResponse[Any]] | ResponseStream[ChatResponseUpdate, ChatResponse[Any]]:
"""Get a response from the Foundry Local chat client with all standard layers enabled."""
super_get_response = cast(
"Callable[..., Awaitable[ChatResponse[Any]] | ResponseStream[ChatResponseUpdate, ChatResponse[Any]]]",
super().get_response,
)
effective_client_kwargs = dict(client_kwargs) if client_kwargs is not None else {}
if middleware is not None:
effective_client_kwargs["middleware"] = middleware
return super_get_response(
messages=messages,
stream=stream,
options=options,
compaction_strategy=compaction_strategy,
tokenizer=tokenizer,
function_invocation_kwargs=function_invocation_kwargs,
client_kwargs=effective_client_kwargs,
)
def __init__(
self,
model: str | None = None,
@@ -182,7 +260,7 @@ class FoundryLocalClient(
# Create a FoundryLocalClient with a specific model ID:
from agent_framework.foundry import FoundryLocalClient
client = FoundryLocalClient(model_id="phi-4-mini")
client = FoundryLocalClient(model="phi-4-mini")
agent = client.as_agent(
name="LocalAgent",
@@ -192,7 +270,7 @@ class FoundryLocalClient(
response = await agent.run("What's the weather like in Seattle?")
# Or you can set the model id in the environment:
os.environ["FOUNDRY_LOCAL_MODEL_ID"] = "phi-4-mini"
os.environ["FOUNDRY_LOCAL_MODEL"] = "phi-4-mini"
client = FoundryLocalClient()
# A FoundryLocalManager is created and if set, the service is started.
@@ -205,12 +283,12 @@ class FoundryLocalClient(
from foundry_local.models import DeviceType
client = FoundryLocalClient(
model_id="phi-4-mini",
model="phi-4-mini",
device=DeviceType.GPU,
)
# and choosing if the model should be prepared on initialization:
client = FoundryLocalClient(
model_id="phi-4-mini",
model="phi-4-mini",
prepare_model=False,
)
# Beware, in this case the first request to generate a completion
@@ -230,7 +308,7 @@ class FoundryLocalClient(
class MyOptions(FoundryLocalChatOptions, total=False):
my_custom_option: str
client: FoundryLocalClient[MyOptions] = FoundryLocalClient(model_id="phi-4-mini")
client: FoundryLocalClient[MyOptions] = FoundryLocalClient(model="phi-4-mini")
response = await client.get_response("Hello", options={"my_custom_option": "value"})
Raises:
@@ -1,5 +1,6 @@
# Copyright (c) Microsoft. All rights reserved.
import inspect
from unittest.mock import MagicMock, patch
import pytest
@@ -66,6 +67,15 @@ def test_foundry_local_client_init(mock_foundry_local_manager: MagicMock) -> Non
assert isinstance(client, SupportsChatGetResponse)
def test_foundry_local_client_get_response_uses_explicit_runtime_buckets() -> None:
"""Foundry Local should expose explicit runtime buckets instead of raw kwargs."""
signature = inspect.signature(FoundryLocalClient.get_response)
assert "client_kwargs" in signature.parameters
assert "function_invocation_kwargs" in signature.parameters
assert all(parameter.kind != inspect.Parameter.VAR_KEYWORD for parameter in signature.parameters.values())
def test_foundry_local_client_init_with_bootstrap_false(mock_foundry_local_manager: MagicMock) -> None:
"""Test FoundryLocalClient initialization with bootstrap=False."""
with patch(