feat: Add responses and safety impl extra_body (#3781)

# What does this PR do?

Have closed the previous PR due to merge conflicts with multiple PRs
Addressed all comments from
https://github.com/llamastack/llama-stack/pull/3768 (sorry for carrying
over to this one)


## Test Plan
Added UTs and integration tests

docs/static/stainless-llama-stack-spec.yaml

@@ -1451,16 +1451,16 @@ paths:
         required: true
       deprecated: false
       x-llama-stack-extra-body-params:
-        - name: shields
+        - name: guardrails
           schema:
             type: array
             items:
               oneOf:
                 - type: string
-                - $ref: '#/components/schemas/ResponseShieldSpec'
+                - $ref: '#/components/schemas/ResponseGuardrailSpec'
           description: >-
-            List of shields to apply during response generation. Shields provide safety
-            and content moderation.
+            List of guardrails to apply during response generation. Guardrails provide
+            safety and content moderation.
           required: false
   /v1/responses/{response_id}:
     get:
@@ -5504,6 +5504,25 @@ components:
           url_citation: '#/components/schemas/OpenAIResponseAnnotationCitation'
           container_file_citation: '#/components/schemas/OpenAIResponseAnnotationContainerFileCitation'
           file_path: '#/components/schemas/OpenAIResponseAnnotationFilePath'
+    OpenAIResponseContentPartRefusal:
+      type: object
+      properties:
+        type:
+          type: string
+          const: refusal
+          default: refusal
+          description: >-
+            Content part type identifier, always "refusal"
+        refusal:
+          type: string
+          description: Refusal text supplied by the model
+      additionalProperties: false
+      required:
+        - type
+        - refusal
+      title: OpenAIResponseContentPartRefusal
+      description: >-
+        Refusal content within a streamed response part.
     "OpenAIResponseInputFunctionToolCallOutput":
       type: object
       properties:
@@ -5675,6 +5694,15 @@ components:
         under one type because the Responses API gives them all the same "type" value,
         and there is no way to tell them apart in certain scenarios.
     OpenAIResponseOutputMessageContent:
+      oneOf:
+        - $ref: '#/components/schemas/OpenAIResponseOutputMessageContentOutputText'
+        - $ref: '#/components/schemas/OpenAIResponseContentPartRefusal'
+      discriminator:
+        propertyName: type
+        mapping:
+          output_text: '#/components/schemas/OpenAIResponseOutputMessageContentOutputText'
+          refusal: '#/components/schemas/OpenAIResponseContentPartRefusal'
+    "OpenAIResponseOutputMessageContentOutputText":
       type: object
       properties:
         text:
@@ -7155,18 +7183,18 @@ components:
         - total_tokens
       title: OpenAIResponseUsage
       description: Usage information for OpenAI response.
-    ResponseShieldSpec:
+    ResponseGuardrailSpec:
       type: object
       properties:
         type:
           type: string
-          description: The type/identifier of the shield.
+          description: The type/identifier of the guardrail.
       additionalProperties: false
       required:
         - type
-      title: ResponseShieldSpec
+      title: ResponseGuardrailSpec
       description: >-
-        Specification for a shield to apply during response generation.
+        Specification for a guardrail to apply during response generation.
     OpenAIResponseInputTool:
       oneOf:
         - $ref: '#/components/schemas/OpenAIResponseInputToolWebSearch'
@@ -7472,25 +7500,6 @@ components:
       title: OpenAIResponseContentPartReasoningText
       description: >-
         Reasoning text emitted as part of a streamed response.
-    OpenAIResponseContentPartRefusal:
-      type: object
-      properties:
-        type:
-          type: string
-          const: refusal
-          default: refusal
-          description: >-
-            Content part type identifier, always "refusal"
-        refusal:
-          type: string
-          description: Refusal text supplied by the model
-      additionalProperties: false
-      required:
-        - type
-        - refusal
-      title: OpenAIResponseContentPartRefusal
-      description: >-
-        Refusal content within a streamed response part.
     OpenAIResponseObjectStream:
       oneOf:
         - $ref: '#/components/schemas/OpenAIResponseObjectStreamResponseCreated'