Python: Fix merge regressions from main (typing + runtime)

Merging main into the type-checker split branch surfaced regressions that
the new five-checker test suite and unit tests caught:

Runtime fixes:
- anthropic: restore the dropped `cache_read_input_token_count` mapping in
  _parse_usage_from_anthropic (lost during merge conflict resolution).
- gemini: _get_function_calling_mode test helper returned str(enum)
  ('FunctionCallingConfigMode.AUTO') instead of the enum value ('AUTO').
- openai: _response_id_from_token test helper was an infinite self-recursion;
  return token['response_id'].
- orchestrations: reset output_events per approval iteration so the terminal
  output assertion counts only the final run.
- core: drop a stale duplicate harness test whose message ('non-negative')
  contradicted the source ('positive').
- purview: import PolicyLocation/PolicyScope/ProtectionScopeActivities/
  ExecutionMode used by the processor tests.

Type-checker fixes (tests, relaxed profile):
- core: pyright/mypy/pyrefly/ty/zuban green-ups across the harness, MCP,
  observability and types tests.
- anthropic/openai: route provider-namespaced UsageDetails keys through a
  dict cast (extra_items TypedDict unsupported by mypy/ty).
- purview: typed model constructors and cache-mock casts.
- ag-ui: annotate WorkflowContext[Any, Any] so yield_output accepts test
  payloads, guard Optional forwarded_props, and ty-ignore intentional bad args.

Source pyright (sole source checker) flagged unnecessary ignores newly
introduced by merged code in core _tools.py and declarative _declarative_base.py.

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

Author: eavanvalkenburg

python/packages/ag-ui/tests/ag_ui/test_endpoint.py

@@ -3,7 +3,7 @@
 """Tests for FastAPI endpoint creation (_endpoint.py)."""
 
 import json
-from typing import Any
+from typing import Any, cast
 
 import pytest
 from ag_ui.core import MessagesSnapshotEvent, RunStartedEvent, StateSnapshotEvent
@@ -82,7 +82,7 @@ async def test_add_endpoint_with_workflow_protocol():
     """Test adding endpoint with native Workflow support."""
 
     @executor(id="start")
-    async def start(message: Any, ctx: WorkflowContext) -> None:
+    async def start(message: Any, ctx: WorkflowContext[Any, Any]) -> None:
         await ctx.yield_output("Workflow response")  # type: ignore[arg-type]  # pyrefly: ignore[bad-argument-type]  # ty: ignore[invalid-argument-type]
 
     app = FastAPI()
@@ -695,7 +695,7 @@ async def stream_fn(messages: Any, options: Any, **kwargs: Any):
         path="/snapshots",
         state_schema={"tenant": {"type": "string"}},
         snapshot_store=store,
-        snapshot_scope_resolver=lambda request: request.forwarded_props["tenant"],
+        snapshot_scope_resolver=lambda request: cast("dict[str, Any]", request.forwarded_props)["tenant"],
     )
     client = TestClient(app)
 
@@ -994,14 +994,16 @@ async def test_workflow_endpoint_hydrates_emitted_snapshots_without_invoking_wor
     call_count = 0
 
     @executor(id="snapshotter")
-    async def snapshotter(message: Any, ctx: WorkflowContext) -> None:
+    async def snapshotter(message: Any, ctx: WorkflowContext[Any, Any]) -> None:
         nonlocal call_count
         del message
         call_count += 1
         await ctx.yield_output(StateSnapshotEvent(snapshot={"active_agent": "flights"}))
         await ctx.yield_output(
             MessagesSnapshotEvent(
-                messages=[{"id": "assistant-snapshot", "role": "assistant", "content": "Stored workflow reply"}]
+                messages=cast(
+                    Any, [{"id": "assistant-snapshot", "role": "assistant", "content": "Stored workflow reply"}]
+                )
             )
         )
 
@@ -1046,7 +1048,7 @@ async def test_workflow_endpoint_hydrates_synthesized_text_and_tool_snapshot():
     call_count = 0
 
     @executor(id="responder")
-    async def responder(message: Any, ctx: WorkflowContext) -> None:
+    async def responder(message: Any, ctx: WorkflowContext[Any, Any]) -> None:
         nonlocal call_count
         del message
         call_count += 1
@@ -1114,7 +1116,7 @@ async def test_workflow_endpoint_hydrates_interrupted_thread_without_invoking_wo
     call_count = 0
 
     @executor(id="requester")
-    async def requester(message: Any, ctx: WorkflowContext) -> None:
+    async def requester(message: Any, ctx: WorkflowContext[Any, Any]) -> None:
         nonlocal call_count
         del message
         call_count += 1
@@ -1167,7 +1169,7 @@ async def test_workflow_endpoint_run_error_does_not_overwrite_previous_snapshot(
     call_count = 0
 
     @executor(id="responder")
-    async def responder(message: Any, ctx: WorkflowContext) -> None:
+    async def responder(message: Any, ctx: WorkflowContext[Any, Any]) -> None:
         nonlocal call_count
         del message
         call_count += 1
@@ -1464,7 +1466,7 @@ async def test_workflow_preserves_history_across_turns():
     call_count = 0
 
     @executor(id="responder")
-    async def responder(message: Any, ctx: WorkflowContext) -> None:
+    async def responder(message: Any, ctx: WorkflowContext[Any, Any]) -> None:
         nonlocal call_count
         del message
         call_count += 1
@@ -1688,7 +1690,7 @@ async def fake_run_workflow_stream(input_data: Any, workflow: Any):
     monkeypatch.setattr(workflow_module, "run_workflow_stream", fake_run_workflow_stream)
 
     @executor(id="noop")
-    async def noop(message: Any, ctx: WorkflowContext) -> None:
+    async def noop(message: Any, ctx: WorkflowContext[Any, Any]) -> None:
         del message, ctx
 
     runner = AgentFrameworkWorkflow(
@@ -1759,7 +1761,7 @@ async def test_workflow_endpoint_snapshot_save_failure_does_not_emit_run_error()
     """A failing snapshot save after RUN_FINISHED must not emit a second terminal RUN_ERROR."""
 
     @executor(id="responder")
-    async def responder(message: Any, ctx: WorkflowContext) -> None:
+    async def responder(message: Any, ctx: WorkflowContext[Any, Any]) -> None:
         del message
         await ctx.yield_output("Workflow reply")
 
@@ -1822,7 +1824,7 @@ def test_workflow_factory_cache_is_scoped_by_snapshot_scope():
     """The same thread id under different Snapshot Scopes must not share a workflow instance."""
 
     @executor(id="noop")
-    async def noop(message: Any, ctx: WorkflowContext) -> None:
+    async def noop(message: Any, ctx: WorkflowContext[Any, Any]) -> None:
         del message, ctx
 
     def factory(thread_id: str) -> Any:

python/packages/ag-ui/tests/ag_ui/test_snapshots.py

@@ -151,10 +151,10 @@ async def test_in_memory_snapshot_store_rejects_invalid_keys() -> None:
     with pytest.raises(ValueError):
         await store.save(scope="tenant-a", thread_id="", snapshot=snapshot)
     with pytest.raises(TypeError):
-        await store.save(scope=123, thread_id="thread-1", snapshot=snapshot)  # type: ignore[arg-type]
+        await store.save(scope=123, thread_id="thread-1", snapshot=snapshot)  # type: ignore[arg-type]  # ty: ignore[invalid-argument-type]
     with pytest.raises(ValueError):
         await store.get(scope="tenant-a", thread_id="")
     with pytest.raises(TypeError):
-        await store.delete(scope=None, thread_id="thread-1")  # type: ignore[arg-type]
+        await store.delete(scope=None, thread_id="thread-1")  # type: ignore[arg-type]  # ty: ignore[invalid-argument-type]
     with pytest.raises(ValueError):
         await store.clear(scope="")

python/packages/anthropic/agent_framework_anthropic/_chat_client.py

@@ -1027,6 +1027,7 @@ def _parse_usage_from_anthropic(self, usage: BetaUsage | BetaMessageDeltaUsage |
             usage_details["cache_creation_input_token_count"] = usage.cache_creation_input_tokens
         if usage.cache_read_input_tokens is not None:
             usage_details["anthropic.cache_read_input_tokens"] = usage.cache_read_input_tokens
+            usage_details["cache_read_input_token_count"] = usage.cache_read_input_tokens
         return usage_details
 
     def _parse_contents_from_anthropic(

python/packages/anthropic/tests/test_anthropic_client.py

@@ -2350,10 +2350,11 @@ def test_parse_usage_with_cache_tokens(mock_anthropic_client: MagicMock) -> None
     result = client._parse_usage_from_anthropic(mock_usage)
 
     assert result is not None
+    result_dict = cast("dict[str, Any]", result)
     assert result["output_token_count"] == 50
     assert result["input_token_count"] == 100
-    assert result["anthropic.cache_creation_input_tokens"] == 20
-    assert result["anthropic.cache_read_input_tokens"] == 30
+    assert result_dict["anthropic.cache_creation_input_tokens"] == 20
+    assert result_dict["anthropic.cache_read_input_tokens"] == 30
     assert result["cache_creation_input_token_count"] == 20
     assert result["cache_read_input_token_count"] == 30
 
@@ -2371,9 +2372,10 @@ def test_parse_usage_preserves_zero_cache_tokens(mock_anthropic_client: MagicMoc
     result = client._parse_usage_from_anthropic(mock_usage)
 
     assert result is not None
-    assert result["anthropic.cache_creation_input_tokens"] == 0
+    result_dict = cast("dict[str, Any]", result)
+    assert result_dict["anthropic.cache_creation_input_tokens"] == 0
     assert result["cache_creation_input_token_count"] == 0
-    assert result["anthropic.cache_read_input_tokens"] == 0
+    assert result_dict["anthropic.cache_read_input_tokens"] == 0
     assert result["cache_read_input_token_count"] == 0
 
 

python/packages/core/agent_framework/_tools.py

@@ -1730,7 +1730,7 @@ async def _try_execute_function_calls(
                 id=fcc.call_id,  # type: ignore[arg-type]
                 function_call=fcc,
             )
-            tool_name = fcc.name  # type: ignore[attr-defined]
+            tool_name = fcc.name
             if tool_name is None:
                 visible_requests.append(approval_request)
                 continue

python/packages/core/tests/core/test_feature_stage.py

@@ -46,7 +46,7 @@ class HelperReleaseCandidateFeature(str, Enum):
 
 
 @pytest.fixture(autouse=True)
-def clear_feature_warning_state() -> Generator[None]:  # type: ignore[misc]  # pyrefly: ignore[bad-return]  # ty: ignore[invalid-return-type]
+def clear_feature_warning_state() -> Generator[None]:  # type: ignore[misc]  # pyrefly: ignore[bad-return]
     _WARNED_FEATURES.clear()
     yield
     _WARNED_FEATURES.clear()

python/packages/core/tests/core/test_function_invocation_logic.py

@@ -635,11 +635,11 @@ def func_with_approval(arg1: str) -> str:
 
     messages: list[Any]
     if not streaming:
-        response = await chat_client_base.get_response([Message(role="user", contents=["hello"])], options=options)  # type: ignore[call-overload]  # pyrefly: ignore[no-matching-overload]  # ty: ignore[no-matching-overload]
+        response = await chat_client_base.get_response([Message(role="user", contents=["hello"])], options=options)  # type: ignore[call-overload]  # pyrefly: ignore[no-matching-overload]
         messages = response.messages
     else:
         updates = []
-        async for update in chat_client_base.get_response(  # type: ignore[call-overload]  # pyrefly: ignore[no-matching-overload]  # ty: ignore[no-matching-overload]
+        async for update in chat_client_base.get_response(  # type: ignore[call-overload]  # pyrefly: ignore[no-matching-overload]
             [Message(role="user", contents=["hello"])], options=options, stream=True
         ):
             updates.append(update)