test: fix five tests left stale by #41311, #41337, #39996 and #41310

Every one of these fails on main's own scheduled CircleCI run with the same
assertion as on any PR, and each traces to a merged behavior change that
never updated the test that pinned the old behavior

- tests/integration/_support/client.py: #41311 made /key/info serve deleted
  keys from the archive with status deleted, so the scenario teardown asserts
  the live row is gone and the readback reports deleted instead of a 404.
  This alone accounts for nine integration-management and one
  integration-providers failure
- tests/integration/authorization/test_warmed_policy.py: #39996 made team
  admins unable to edit any team field unless a proxy admin allow-lists it,
  and tpm_limit is the only field it accepts today. The demotion test now
  enables tpm_limit for the scenario and edits that instead of team_alias
- tests/llm_responses_api_testing/test_base_responses_api_streaming_iterator.py:
  #41337 reads usage off the terminal response and copies the event when it
  is missing, which a Mock(spec=ResponsesAPIResponse) cannot survive. The
  four mocks now carry a usage object
- tests/test_openai_endpoints.py: #41310 lengthened the access-denied
  message, and the test matched against the ExceptionInfo repr, which
  saferepr truncates in the middle. It now matches the exception text
- tests/local_testing/test_text_completion.py: Together no longer serves
  Qwen2-1.5B serverless, the cheapest cost-map row. The test mocks the
  completions call and asserts the request litellm builds, so a vendor
  catalog rotation cannot fail it again

test_router_fallbacks_with_cooldowns_and_dynamic_credentials is deliberately
untouched: it passes and fails on main with identical code, and the failing
path is a product question about whether dynamic-credential 429s cool down

tests/integration/_support/client.py

@@ -132,8 +132,10 @@ class Scenario:
 
     def delete_key(self, token: str) -> None:
         self.gateway.post("/key/delete", {"keys": [token]})
-        response: Final = self.gateway.request("GET", "/key/info", params={"key": sha256(token.encode()).hexdigest()})
-        assert response.status_code == 404, f"Deleted key remains readable: {response.status_code}"
+        hashed: Final = sha256(token.encode()).hexdigest()
+        assert read_rows('SELECT token FROM "LiteLLM_VerificationToken" WHERE token = %s', (hashed,)) == []
+        info: Final = object_value(self.gateway.get("/key/info", {"key": hashed})["info"])
+        assert info["status"] == "deleted", f"Deleted key still served as live: {info['status']}"
 
     def delete_model(self, identity: str) -> None:
         self.gateway.post("/model/delete", {"id": identity})

tests/integration/authorization/test_warmed_policy.py

@@ -1,10 +1,12 @@
-from contextlib import ExitStack
+from collections.abc import Iterator
+from contextlib import ExitStack, contextmanager
 from hashlib import sha256
 from typing import Final
 import os
 
 import psycopg
 import pytest
+from pydantic import JsonValue
 from hypothesis import strategies as st
 from hypothesis.stateful import RuleBasedStateMachine, invariant, rule, run_state_machine_as_test
 
@@ -134,37 +136,58 @@ def test_scim_deactivation_blocks_null_and_false_keys_but_preserves_other_owners
                 assert_serving(gateway, model, token, 200)
 
 
+def _set_team_admin_editable_fields(gateway: Gateway, fields: list[JsonValue]) -> None:
+    response: Final = gateway.request("PATCH", "/update/ui_settings", {"team_admin_editable_team_fields": fields})
+    assert response.status_code == 200, response.text
+
+
+@contextmanager
+def _team_admins_may_edit(gateway: Gateway, fields: list[JsonValue]) -> Iterator[None]:
+    original: Final = object_value(gateway.get("/get/ui_settings")["values"]).get("team_admin_editable_team_fields")
+    _set_team_admin_editable_fields(gateway, fields)
+    try:
+        yield
+    finally:
+        _set_team_admin_editable_fields(gateway, original if isinstance(original, list) else [])
+
+
 @pytest.mark.covers("mgmt.team.member_update.demoted_role_cannot_write")
 def test_warmed_team_role_demotion_prevents_later_management_writes(gateway: Gateway) -> None:
-    with gateway.scenario() as scenario:
+    with gateway.scenario() as scenario, _team_admins_may_edit(gateway, ["tpm_limit"]):
         model: Final = scenario.model()
         user: Final = scenario.user(user_role="internal_user")
-        team: Final = scenario.team(models=[model], members_with_roles=[{"user_id": user, "role": "admin"}])
-        control_team: Final = scenario.team(models=[model])
+        team: Final = scenario.team(
+            models=[model], tpm_limit=1000, members_with_roles=[{"user_id": user, "role": "admin"}]
+        )
+        control_team: Final = scenario.team(models=[model], tpm_limit=1000)
         caller: Final = scenario.key(
             user_id=user, team_id=team, models=[model], allowed_routes=["/team/update", "/v1/chat/completions"]
         )
         gateway.chat(model, key=caller)
-        changed: Final = gateway.request("POST", "/team/update", {"team_id": team, "team_alias": "permitted"}, key=caller)
+        changed: Final = gateway.request("POST", "/team/update", {"team_id": team, "tpm_limit": 5000}, key=caller)
         assert changed.status_code == 200, changed.text
+        assert read_rows('SELECT tpm_limit FROM "LiteLLM_TeamTable" WHERE team_id = %s', (team,)) == [
+            {"tpm_limit": 5000}
+        ]
         unrelated_before: Final = read_rows(
-            'SELECT team_alias FROM "LiteLLM_TeamTable" WHERE team_id = %s', (control_team,)
+            'SELECT tpm_limit FROM "LiteLLM_TeamTable" WHERE team_id = %s', (control_team,)
         )
         unrelated: Final = gateway.request(
-            "POST", "/team/update", {"team_id": control_team, "team_alias": "must-not-persist"}, key=caller
+            "POST", "/team/update", {"team_id": control_team, "tpm_limit": 7000}, key=caller
         )
         assert unrelated.status_code == 403, unrelated.text
         assert read_rows(
-            'SELECT team_alias FROM "LiteLLM_TeamTable" WHERE team_id = %s', (control_team,)
+            'SELECT tpm_limit FROM "LiteLLM_TeamTable" WHERE team_id = %s', (control_team,)
         ) == unrelated_before
         gateway.post("/team/member_update", {"team_id": team, "user_id": user, "role": "user"})
         for target in (team, control_team):
-            before: Final = read_rows('SELECT team_alias FROM "LiteLLM_TeamTable" WHERE team_id = %s', (target,))
+            before: Final = read_rows('SELECT tpm_limit FROM "LiteLLM_TeamTable" WHERE team_id = %s', (target,))
             denied: Final = gateway.request(
-                "POST", "/team/update", {"team_id": target, "team_alias": "must-not-persist"}, key=caller
+                "POST", "/team/update", {"team_id": target, "tpm_limit": 9000}, key=caller
             )
             assert denied.status_code == 403, denied.text
-            assert read_rows('SELECT team_alias FROM "LiteLLM_TeamTable" WHERE team_id = %s', (target,)) == before
+            after: Final = read_rows('SELECT tpm_limit FROM "LiteLLM_TeamTable" WHERE team_id = %s', (target,))
+            assert after == before
         roster: Final = read_rows('SELECT members_with_roles FROM "LiteLLM_TeamTable" WHERE team_id = %s', (team,))
         members: Final = roster[0]["members_with_roles"]
         assert isinstance(members, list)

tests/llm_responses_api_testing/test_base_responses_api_streaming_iterator.py

@@ -26,6 +26,7 @@ from litellm.llms.base_llm.responses.transformation import BaseResponsesAPIConfi
 from litellm.responses.streaming_iterator import BaseResponsesAPIStreamingIterator
 from litellm.responses.utils import ResponsesAPIRequestUtils
 from litellm.types.llms.openai import (
+    ResponseAPIUsage,
     ResponseCompletedEvent,
     ResponseFailedEvent,
     ResponseIncompleteEvent,
@@ -69,6 +70,7 @@ class TestBaseResponsesAPIStreamingIterator:
 
         mock_responses_api_response = Mock(spec=ResponsesAPIResponse)
         mock_responses_api_response.id = "resp_u2028"
+        mock_responses_api_response.usage = ResponseAPIUsage(input_tokens=3, output_tokens=2, total_tokens=5)
         mock_completed_event = Mock(spec=ResponseCompletedEvent)
         mock_completed_event.type = ResponsesAPIStreamEvents.RESPONSE_COMPLETED
         mock_completed_event.response = mock_responses_api_response
@@ -123,6 +125,7 @@ class TestBaseResponsesAPIStreamingIterator:
         # Mock the _update_responses_api_response_id_with_model_id method
         updated_response = Mock(spec=ResponsesAPIResponse)
         updated_response.id = "updated_response_id"
+        updated_response.usage = ResponseAPIUsage(input_tokens=3, output_tokens=2, total_tokens=5)
 
         # Create the iterator instance
         iterator = BaseResponsesAPIStreamingIterator(
@@ -524,7 +527,7 @@ class TestBaseResponsesAPIStreamingIterator:
             "type": "server_error",
             "message": "The model encountered an error",
         }
-        mock_responses_api_response.usage = None
+        mock_responses_api_response.usage = ResponseAPIUsage(input_tokens=3, output_tokens=2, total_tokens=5)
 
         mock_failed_event = Mock(spec=ResponseFailedEvent)
         mock_failed_event.type = ResponsesAPIStreamEvents.RESPONSE_FAILED
@@ -604,7 +607,7 @@ class TestBaseResponsesAPIStreamingIterator:
         mock_responses_api_response = Mock(spec=ResponsesAPIResponse)
         mock_responses_api_response.id = "resp_incomplete_123"
         mock_responses_api_response.incomplete_details = {"reason": "max_output_tokens"}
-        mock_responses_api_response.usage = None
+        mock_responses_api_response.usage = ResponseAPIUsage(input_tokens=3, output_tokens=2, total_tokens=5)
 
         mock_incomplete_event = Mock(spec=ResponseIncompleteEvent)
         mock_incomplete_event.type = ResponsesAPIStreamEvents.RESPONSE_INCOMPLETE

tests/local_testing/test_text_completion.py

@@ -12,7 +12,6 @@ from unittest.mock import MagicMock, patch
 import pytest
 
 import litellm
-from tests._live_test_helpers import cheapest_together_chat_model
 from litellm import (
     RateLimitError,
     TextCompletionResponse,
@@ -4023,27 +4022,27 @@ def test_async_text_completion():
     asyncio.run(test_get_response())
 
 
-@pytest.mark.flaky(retries=6, delay=1)
 def test_async_text_completion_together_ai():
-    litellm.set_verbose = True
-    print("test_async_text_completion")
+    from openai import AsyncOpenAI
 
-    async def test_get_response():
-        try:
+    client = AsyncOpenAI(api_key="my-fake-key")
+
+    async def run_call():
+        with patch.object(client.completions.with_raw_response, "create", side_effect=mock_post) as mock_call:
             response = await litellm.atext_completion(
-                model=cheapest_together_chat_model(),
+                model="together_ai/Qwen/Qwen2-1.5B-Instruct",
                 prompt="good morning",
                 max_tokens=10,
+                client=client,
             )
-            print(f"response: {response}")
-        except litellm.RateLimitError as e:
-            print(e)
-        except litellm.Timeout as e:
-            print(e)
-        except Exception as e:
-            pytest.fail("An unexpected error occurred")
+        return response, mock_call.call_args.kwargs
 
-    asyncio.run(test_get_response())
+    response, sent = asyncio.run(run_call())
+    assert sent["model"] == "Qwen/Qwen2-1.5B-Instruct"
+    assert sent["prompt"] == "good morning"
+    assert sent["max_tokens"] == 10
+    assert response.choices[0].text == ") might be faster than then answering, and the added time it takes for the"
+    assert response.usage.total_tokens == 18
 
 
 # test_async_text_completion()

tests/test_openai_endpoints.py

@@ -307,7 +307,7 @@ async def test_chat_completion():
                 model="gpt-4",
                 messages=[{"role": "user", "content": "Hello!"}],
             )
-        assert "is not available for this API key" in str(e)
+        assert "is not available for this API key" in str(e.value)
 
 
 @pytest.mark.asyncio