address feedback

src/llama_stack/providers/remote/files/s3/files.py

@@ -55,6 +55,9 @@ def _create_s3_client(config: S3FilesImplConfig) -> S3Client:
                 }
             )
 
+        # Both cast and type:ignore are needed here:
+        # - cast tells mypy the return type for downstream usage (S3Client vs generic client)
+        # - type:ignore suppresses the call-overload error from boto3's complex overloaded signatures
         return cast("S3Client", boto3.client("s3", **s3_config))  # type: ignore[call-overload]
 
     except (BotoCoreError, NoCredentialsError) as e:

src/llama_stack/providers/remote/inference/gemini/gemini.py

@@ -37,21 +37,21 @@ class GeminiInferenceAdapter(OpenAIMixin):
         Override embeddings method to handle Gemini's missing usage statistics.
         Gemini's embedding API doesn't return usage information, so we provide default values.
         """
-        # Build kwargs conditionally to avoid NotGiven/Omit type mismatch
-        kwargs: dict[str, Any] = {
+        # Build request params conditionally to avoid NotGiven/Omit type mismatch
+        request_params: dict[str, Any] = {
             "model": await self._get_provider_model_id(params.model),
             "input": params.input,
         }
         if params.encoding_format is not None:
-            kwargs["encoding_format"] = params.encoding_format
+            request_params["encoding_format"] = params.encoding_format
         if params.dimensions is not None:
-            kwargs["dimensions"] = params.dimensions
+            request_params["dimensions"] = params.dimensions
         if params.user is not None:
-            kwargs["user"] = params.user
+            request_params["user"] = params.user
         if params.model_extra:
-            kwargs["extra_body"] = params.model_extra
+            request_params["extra_body"] = params.model_extra
 
-        response = await self.client.embeddings.create(**kwargs)
+        response = await self.client.embeddings.create(**request_params)
 
         data = []
         for i, embedding_data in enumerate(response.data):

src/llama_stack/providers/utils/inference/openai_mixin.py

@@ -351,22 +351,22 @@ class OpenAIMixin(NeedsRequestProviderData, ABC, BaseModel):
         """
         Direct OpenAI embeddings API call.
         """
-        # Build kwargs conditionally to avoid NotGiven/Omit type mismatch
+        # Build request params conditionally to avoid NotGiven/Omit type mismatch
         # The OpenAI SDK uses Omit in signatures but NOT_GIVEN has type NotGiven
-        kwargs: dict[str, Any] = {
+        request_params: dict[str, Any] = {
             "model": await self._get_provider_model_id(params.model),
             "input": params.input,
         }
         if params.encoding_format is not None:
-            kwargs["encoding_format"] = params.encoding_format
+            request_params["encoding_format"] = params.encoding_format
         if params.dimensions is not None:
-            kwargs["dimensions"] = params.dimensions
+            request_params["dimensions"] = params.dimensions
         if params.user is not None:
-            kwargs["user"] = params.user
+            request_params["user"] = params.user
         if params.model_extra:
-            kwargs["extra_body"] = params.model_extra
+            request_params["extra_body"] = params.model_extra
 
-        response = await self.client.embeddings.create(**kwargs)
+        response = await self.client.embeddings.create(**request_params)
 
         data = []
         for i, embedding_data in enumerate(response.data):