Nightly OSS: Incompatibility of assistant with tool_calls and tool role in the /v1/responses endpoint

NIGHTLY_CODEX_FINAL_ATTEMPT_1.md

@@ -0,0 +1,13 @@
+Implemented a small contribution for [openai/openai-python#2677](https://github.com/openai/openai-python/issues/2677).
+
+Changed:
+- Added a focused regression test in [tests/lib/responses/test_responses.py](/home/runner/work/oss-nightly-control/oss-nightly-control/target/tests/lib/responses/test_responses.py:46) that verifies `responses.create()` sends Responses-native `function_call` and `function_call_output` input items.
+- Added a minimal migration example in [examples/responses/function_call_inputs.py](/home/runner/work/oss-nightly-control/oss-nightly-control/target/examples/responses/function_call_inputs.py:1).
+- Wrote [NIGHTLY_REPORT.md](/home/runner/work/oss-nightly-control/oss-nightly-control/target/NIGHTLY_REPORT.md:1).
+
+Checks run:
+- `uv run --with pytest --with pytest-xdist --with respx --with inline-snapshot --with pytest-asyncio pytest tests/lib/responses/test_responses.py -q`
+- `uv run --with ruff ruff check tests/lib/responses/test_responses.py examples/responses/function_call_inputs.py`
+- `python -m py_compile examples/responses/function_call_inputs.py tests/lib/responses/test_responses.py`
+
+All passed. I left the working tree uncommitted as requested.

examples/responses/function_call_inputs.py

@@ -0,0 +1,30 @@
+from openai import OpenAI
+
+client = OpenAI()
+
+response = client.responses.create(
+    model="gpt-4o-mini",
+    input=[
+        {
+            "role": "developer",
+            "content": "Use the supplied function results to answer.",
+        },
+        {
+            "role": "user",
+            "content": "What is the current order status?",
+        },
+        {
+            "type": "function_call",
+            "call_id": "call_123",
+            "name": "get_order_status",
+            "arguments": '{"order_id":"order_123"}',
+        },
+        {
+            "type": "function_call_output",
+            "call_id": "call_123",
+            "output": '{"status":"shipped"}',
+        },
+    ],
+)
+
+print(response.output_text)

tests/lib/responses/test_responses.py

@@ -1,7 +1,9 @@
 from __future__ import annotations
 
+import json
 from typing_extensions import TypeVar
 
+import httpx
 import pytest
 from respx import MockRouter
 from inline_snapshot import snapshot
@@ -41,6 +43,79 @@ def test_output_text(client: OpenAI, respx_mock: MockRouter) -> None:
     )
 
 
+@pytest.mark.respx(base_url=base_url)
+def test_create_with_function_call_input_items(client: OpenAI, respx_mock: MockRouter) -> None:
+    def handler(request: httpx.Request) -> httpx.Response:
+        assert json.loads(request.content) == {
+            "model": "gpt-4o-mini",
+            "input": [
+                {
+                    "role": "developer",
+                    "content": "Use the supplied function results to answer.",
+                },
+                {
+                    "role": "user",
+                    "content": "What is the current order status?",
+                },
+                {
+                    "type": "function_call",
+                    "call_id": "call_123",
+                    "name": "get_order_status",
+                    "arguments": '{"order_id":"order_123"}',
+                },
+                {
+                    "type": "function_call_output",
+                    "call_id": "call_123",
+                    "output": '{"status":"shipped"}',
+                },
+            ],
+        }
+
+        return httpx.Response(
+            200,
+            json={
+                "id": "resp_123",
+                "object": "response",
+                "created_at": 1754925861,
+                "status": "completed",
+                "model": "gpt-4o-mini-2024-07-18",
+                "output": [],
+                "parallel_tool_calls": True,
+                "tool_choice": "auto",
+                "tools": [],
+            },
+        )
+
+    respx_mock.post("/responses").mock(side_effect=handler)
+
+    response = client.responses.create(
+        model="gpt-4o-mini",
+        input=[
+            {
+                "role": "developer",
+                "content": "Use the supplied function results to answer.",
+            },
+            {
+                "role": "user",
+                "content": "What is the current order status?",
+            },
+            {
+                "type": "function_call",
+                "call_id": "call_123",
+                "name": "get_order_status",
+                "arguments": '{"order_id":"order_123"}',
+            },
+            {
+                "type": "function_call_output",
+                "call_id": "call_123",
+                "output": '{"status":"shipped"}',
+            },
+        ],
+    )
+
+    assert response.id == "resp_123"
+
+
 @pytest.mark.parametrize("sync", [True, False], ids=["sync", "async"])
 def test_stream_method_definition_in_sync(sync: bool, client: OpenAI, async_client: AsyncOpenAI) -> None:
     checking_client: OpenAI | AsyncOpenAI = client if sync else async_client