feat: add Chinese UI, Responses models, and mirrored Windows setup

Localize the web console and persist custom model and capability settings.
Use OpenAI Responses exclusively and move runtime limits into environment configuration.
Bootstrap uv and the virtual environment on Windows, support optional China mirrors,
and package source distributions without environments, build output, or credentials.

Update regression tests and repository quality checks for the new behavior.

tests/unit/tools/test_object_detection_tool.py

@@ -66,14 +66,14 @@ def test_object_detection_tool_subclass():
     assert operator_object_detection.args_schema == OperatorObjectDetectionArgs
 
     # GenAI FunctionDeclaration export
-    declaration = object_detection.to_genai_declaration()
+    declaration = object_detection.to_tool_declaration()
     assert declaration.name == "object_detection"
-    assert "image_path" in declaration.parameters.properties
-    assert "queries" in declaration.parameters.properties
+    assert "image_path" in declaration.parameters["properties"]
+    assert "queries" in declaration.parameters["properties"]
 
-    op_declaration = operator_object_detection.to_genai_declaration()
+    op_declaration = operator_object_detection.to_tool_declaration()
     assert op_declaration.name == "object_detection"
-    assert "queries" in op_declaration.parameters.properties
+    assert "queries" in op_declaration.parameters["properties"]
 
     # Wrappers
     assert object_detection_wrapper.tool_fn_getter == get_object_detector_tool