chore: refactor (chat)completions endpoints to use shared params struct (#3761)

# What does this PR do?

Converts openai(_chat)_completions params to pydantic BaseModel to
reduce code duplication across all providers.

## Test Plan
CI









---
[//]: # (BEGIN SAPLING FOOTER)
Stack created with [Sapling](https://sapling-scm.com). Best reviewed
with
[ReviewStack](https://reviewstack.dev/llamastack/llama-stack/pull/3761).
* #3777
* __->__ #3761

llama_stack/strong_typing/inspection.py

@@ -50,6 +50,10 @@ if sys.version_info >= (3, 10):
 else:
     from typing_extensions import TypeGuard
 
+
+from pydantic import BaseModel
+from pydantic.fields import FieldInfo
+
 S = TypeVar("S")
 T = TypeVar("T")
 K = TypeVar("K")
@@ -570,7 +574,8 @@ def get_class_properties(typ: type) -> Iterable[Tuple[str, type | str]]:
     elif hasattr(typ, "model_fields"):
         # Pydantic BaseModel - use model_fields to exclude ClassVar and other non-field attributes
         # Reconstruct Annotated type if discriminator exists to preserve metadata
-        from typing import Annotated, Any, cast
+        from typing import Annotated, Any
+
         from pydantic.fields import FieldInfo
 
         def get_field_type(name: str, field: Any) -> type | str:
@@ -1049,3 +1054,32 @@ def check_recursive(
         pred = lambda typ, obj: True  # noqa: E731
 
     return RecursiveChecker(pred).check(type(obj), obj)
+
+
+def is_unwrapped_body_param(param_type: Any) -> bool:
+    """
+    Check if a parameter type represents an unwrapped body parameter.
+    An unwrapped body parameter is an Annotated type with Body(embed=False)
+
+    This is used to determine whether request parameters should be flattened
+    in OpenAPI specs and client libraries (matching FastAPI's embed=False behavior).
+
+    Args:
+        param_type: The parameter type annotation to check
+
+    Returns:
+        True if the parameter should be treated as an unwrapped body parameter
+    """
+    # Check if it's Annotated with Body(embed=False)
+    if typing.get_origin(param_type) is Annotated:
+        args = typing.get_args(param_type)
+        base_type = args[0]
+        metadata = args[1:]
+
+        # Look for Body annotation with embed=False
+        # Body() returns a FieldInfo object, so we check for that type and the embed attribute
+        for item in metadata:
+            if isinstance(item, FieldInfo) and hasattr(item, "embed") and not item.embed:
+                return inspect.isclass(base_type) and issubclass(base_type, BaseModel)
+
+    return False