feat: add list responses API (#2233)

# What does this PR do?
This is not part of the official OpenAI API, but we'll use this for the
logs UI.
In order to support more filtering options, I'm adopting the newly
introduced sql store in in place of the kv store.

## Test Plan
Added integration/unit tests.

llama_stack/providers/inline/agents/meta_reference/agents.py

@@ -20,9 +20,11 @@ from llama_stack.apis.agents import (
     AgentTurnCreateRequest,
     AgentTurnResumeRequest,
     Document,
+    ListOpenAIResponseObject,
     OpenAIResponseInput,
     OpenAIResponseInputTool,
     OpenAIResponseObject,
+    Order,
     Session,
     Turn,
 )
@@ -39,6 +41,7 @@ from llama_stack.apis.tools import ToolGroups, ToolRuntime
 from llama_stack.apis.vector_io import VectorIO
 from llama_stack.providers.utils.kvstore import InmemoryKVStoreImpl, kvstore_impl
 from llama_stack.providers.utils.pagination import paginate_records
+from llama_stack.providers.utils.responses.responses_store import ResponsesStore
 
 from .agent_instance import ChatAgent
 from .config import MetaReferenceAgentsImplConfig
@@ -66,15 +69,17 @@ class MetaReferenceAgentsImpl(Agents):
         self.tool_groups_api = tool_groups_api
 
         self.in_memory_store = InmemoryKVStoreImpl()
-        self.openai_responses_impl = None
+        self.openai_responses_impl: OpenAIResponsesImpl | None = None
 
     async def initialize(self) -> None:
         self.persistence_store = await kvstore_impl(self.config.persistence_store)
+        self.responses_store = ResponsesStore(self.config.responses_store)
+        await self.responses_store.initialize()
         self.openai_responses_impl = OpenAIResponsesImpl(
-            self.persistence_store,
             inference_api=self.inference_api,
             tool_groups_api=self.tool_groups_api,
             tool_runtime_api=self.tool_runtime_api,
+            responses_store=self.responses_store,
         )
 
     async def create_agent(
@@ -323,3 +328,12 @@ class MetaReferenceAgentsImpl(Agents):
         return await self.openai_responses_impl.create_openai_response(
             input, model, instructions, previous_response_id, store, stream, temperature, tools
         )
+
+    async def list_openai_responses(
+        self,
+        after: str | None = None,
+        limit: int | None = 50,
+        model: str | None = None,
+        order: Order | None = Order.desc,
+    ) -> ListOpenAIResponseObject:
+        return await self.openai_responses_impl.list_openai_responses(after, limit, model, order)

llama_stack/providers/inline/agents/meta_reference/config.py

@@ -10,10 +10,12 @@ from pydantic import BaseModel
 
 from llama_stack.providers.utils.kvstore import KVStoreConfig
 from llama_stack.providers.utils.kvstore.config import SqliteKVStoreConfig
+from llama_stack.providers.utils.sqlstore.sqlstore import SqliteSqlStoreConfig, SqlStoreConfig
 
 
 class MetaReferenceAgentsImplConfig(BaseModel):
     persistence_store: KVStoreConfig
+    responses_store: SqlStoreConfig
 
     @classmethod
     def sample_run_config(cls, __distro_dir__: str) -> dict[str, Any]:
@@ -21,5 +23,9 @@ class MetaReferenceAgentsImplConfig(BaseModel):
             "persistence_store": SqliteKVStoreConfig.sample_run_config(
                 __distro_dir__=__distro_dir__,
                 db_name="agents_store.db",
-            )
+            ),
+            "responses_store": SqliteSqlStoreConfig.sample_run_config(
+                __distro_dir__=__distro_dir__,
+                db_name="responses_store.db",
+            ),
         }

llama_stack/providers/inline/agents/meta_reference/openai_responses.py

@@ -12,7 +12,9 @@ from typing import Any, cast
 from openai.types.chat import ChatCompletionToolParam
 from pydantic import BaseModel
 
+from llama_stack.apis.agents import Order
 from llama_stack.apis.agents.openai_responses import (
+    ListOpenAIResponseObject,
     OpenAIResponseInput,
     OpenAIResponseInputFunctionToolCallOutput,
     OpenAIResponseInputItemList,
@@ -53,7 +55,7 @@ from llama_stack.apis.tools.tools import ToolGroups, ToolInvocationResult, ToolR
 from llama_stack.log import get_logger
 from llama_stack.models.llama.datatypes import ToolDefinition, ToolParamDefinition
 from llama_stack.providers.utils.inference.openai_compat import convert_tooldef_to_openai_tool
-from llama_stack.providers.utils.kvstore import KVStore
+from llama_stack.providers.utils.responses.responses_store import ResponsesStore
 
 logger = get_logger(name=__name__, category="openai_responses")
 
@@ -169,34 +171,27 @@ class OpenAIResponsePreviousResponseWithInputItems(BaseModel):
 class OpenAIResponsesImpl:
     def __init__(
         self,
-        persistence_store: KVStore,
         inference_api: Inference,
         tool_groups_api: ToolGroups,
         tool_runtime_api: ToolRuntime,
+        responses_store: ResponsesStore,
     ):
-        self.persistence_store = persistence_store
         self.inference_api = inference_api
         self.tool_groups_api = tool_groups_api
         self.tool_runtime_api = tool_runtime_api
-
-    async def _get_previous_response_with_input(self, id: str) -> OpenAIResponsePreviousResponseWithInputItems:
-        key = f"{OPENAI_RESPONSES_PREFIX}{id}"
-        response_json = await self.persistence_store.get(key=key)
-        if response_json is None:
-            raise ValueError(f"OpenAI response with id '{id}' not found")
-        return OpenAIResponsePreviousResponseWithInputItems.model_validate_json(response_json)
+        self.responses_store = responses_store
 
     async def _prepend_previous_response(
         self, input: str | list[OpenAIResponseInput], previous_response_id: str | None = None
     ):
         if previous_response_id:
-            previous_response_with_input = await self._get_previous_response_with_input(previous_response_id)
+            previous_response_with_input = await self.responses_store.get_response_object(previous_response_id)
 
             # previous response input items
-            new_input_items = previous_response_with_input.input_items.data
+            new_input_items = previous_response_with_input.input
 
             # previous response output items
-            new_input_items.extend(previous_response_with_input.response.output)
+            new_input_items.extend(previous_response_with_input.output)
 
             # new input items from the current request
             if isinstance(input, str):
@@ -216,8 +211,17 @@ class OpenAIResponsesImpl:
         self,
         response_id: str,
     ) -> OpenAIResponseObject:
-        response_with_input = await self._get_previous_response_with_input(response_id)
-        return response_with_input.response
+        response_with_input = await self.responses_store.get_response_object(response_id)
+        return OpenAIResponseObject(**{k: v for k, v in response_with_input.model_dump().items() if k != "input"})
+
+    async def list_openai_responses(
+        self,
+        after: str | None = None,
+        limit: int | None = 50,
+        model: str | None = None,
+        order: Order | None = Order.desc,
+    ) -> ListOpenAIResponseObject:
+        return await self.responses_store.list_responses(after, limit, model, order)
 
     async def create_openai_response(
         self,
@@ -360,15 +364,9 @@ class OpenAIResponsesImpl:
                     else:
                         input_items_data.append(input_item)
 
-            input_items = OpenAIResponseInputItemList(data=input_items_data)
-            prev_response = OpenAIResponsePreviousResponseWithInputItems(
-                input_items=input_items,
-                response=response,
-            )
-            key = f"{OPENAI_RESPONSES_PREFIX}{response.id}"
-            await self.persistence_store.set(
-                key=key,
-                value=prev_response.model_dump_json(),
+            await self.responses_store.store_response_object(
+                response_object=response,
+                input=input_items_data,
             )
 
         if stream: