phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-12-28 02:01:59 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

fix: Restore previous responses to input list, not messages

Browse source 
This adjusts the restoration of previous responses to prepend them to
the list of Responses API inputs instead of our converted list of Chat
Completion messages. This matches the expected behavior of the
Responses API, and I misinterpreted the nuances here in the initial implementation.

Signed-off-by: Ben Browning <bbrownin@redhat.com>
This commit is contained in:
Ben Browning 2025-05-02 11:08:09 -04:00
parent 5b2e850754
commit b90bb66f28
7 changed files with 428 additions and 364 deletions
Download patch file Download diff file Expand all files Collapse all files

4
llama_stack/apis/agents/agents.py
View file

@ -31,7 +31,7 @@ from llama_stack.apis.tools import ToolDef
from llama_stack.schema_utils import json_schema_type, register_schema, webmethod
from .openai_responses import (
OpenAIResponseInputMessage,
OpenAIResponseInput,
OpenAIResponseInputTool,
OpenAIResponseObject,
OpenAIResponseObjectStream,
@ -592,7 +592,7 @@ class Agents(Protocol):
@webmethod(route="/openai/v1/responses", method="POST")
async def create_openai_response(
self,
input: str | list[OpenAIResponseInputMessage],
input: str | list[OpenAIResponseInput],
model: str,
previous_response_id: str | None = None,
store: bool | None = True,

86
llama_stack/apis/agents/openai_responses.py
View file

@ -17,6 +17,28 @@ class OpenAIResponseError(BaseModel):
message: str
@json_schema_type
class OpenAIResponseInputMessageContentText(BaseModel):
text: str
type: Literal["input_text"] = "input_text"
@json_schema_type
class OpenAIResponseInputMessageContentImage(BaseModel):
detail: Literal["low"] | Literal["high"] | Literal["auto"] = "auto"
type: Literal["input_image"] = "input_image"
# TODO: handle file_id
image_url: str | None = None
# TODO: handle file content types
OpenAIResponseInputMessageContent = Annotated[
OpenAIResponseInputMessageContentText | OpenAIResponseInputMessageContentImage,
Field(discriminator="type"),
]
register_schema(OpenAIResponseInputMessageContent, name="OpenAIResponseInputMessageContent")
@json_schema_type
class OpenAIResponseOutputMessageContentOutputText(BaseModel):
text: str
@ -31,13 +53,22 @@ register_schema(OpenAIResponseOutputMessageContent, name="OpenAIResponseOutputMe
@json_schema_type
class OpenAIResponseOutputMessage(BaseModel):
id: str
content: list[OpenAIResponseOutputMessageContent]
role: Literal["assistant"] = "assistant"
status: str
class OpenAIResponseMessage(BaseModel):
"""
Corresponds to the various Message types in the Responses API.
They are all 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.
"""
content: str | list[OpenAIResponseInputMessageContent] | list[OpenAIResponseOutputMessageContent]
role: Literal["system"] | Literal["developer"] | Literal["user"] | Literal["assistant"]
type: Literal["message"] = "message"
# The fields below are not used in all scenarios, but are required in others.
id: str | None = None
status: str | None = None
@json_schema_type
class OpenAIResponseOutputMessageWebSearchToolCall(BaseModel):
@ -47,7 +78,7 @@ class OpenAIResponseOutputMessageWebSearchToolCall(BaseModel):
OpenAIResponseOutput = Annotated[
OpenAIResponseOutputMessage | OpenAIResponseOutputMessageWebSearchToolCall,
OpenAIResponseMessage | OpenAIResponseOutputMessageWebSearchToolCall,
Field(discriminator="type"),
]
register_schema(OpenAIResponseOutput, name="OpenAIResponseOutput")
@ -89,33 +120,15 @@ OpenAIResponseObjectStream = Annotated[
register_schema(OpenAIResponseObjectStream, name="OpenAIResponseObjectStream")
@json_schema_type
class OpenAIResponseInputMessageContentText(BaseModel):
text: str
type: Literal["input_text"] = "input_text"
@json_schema_type
class OpenAIResponseInputMessageContentImage(BaseModel):
detail: Literal["low"] | Literal["high"] | Literal["auto"] = "auto"
type: Literal["input_image"] = "input_image"
# TODO: handle file_id
image_url: str | None = None
# TODO: handle file content types
OpenAIResponseInputMessageContent = Annotated[
OpenAIResponseInputMessageContentText | OpenAIResponseInputMessageContentImage,
Field(discriminator="type"),
OpenAIResponseInput = Annotated[
# Responses API allows output messages to be passed in as input
OpenAIResponseOutputMessageWebSearchToolCall
|
# Fallback to the generic message type as a last resort
OpenAIResponseMessage,
Field(union_mode="left_to_right"),
]
register_schema(OpenAIResponseInputMessageContent, name="OpenAIResponseInputMessageContent")
@json_schema_type
class OpenAIResponseInputMessage(BaseModel):
content: str | list[OpenAIResponseInputMessageContent]
role: Literal["system"] | Literal["developer"] | Literal["user"] | Literal["assistant"]
type: Literal["message"] | None = "message"
register_schema(OpenAIResponseInput, name="OpenAIResponseInput")
@json_schema_type
@ -133,18 +146,11 @@ OpenAIResponseInputTool = Annotated[
register_schema(OpenAIResponseInputTool, name="OpenAIResponseInputTool")
@json_schema_type
class OpenAIResponseInputItemMessage(OpenAIResponseInputMessage):
id: str
@json_schema_type
class OpenAIResponseInputItemList(BaseModel):
data: list[OpenAIResponseInputItemMessage]
data: list[OpenAIResponseInput]
object: Literal["list"] = "list"
@json_schema_type
class OpenAIResponsePreviousResponseWithInputItems(BaseModel):
input_items: OpenAIResponseInputItemList
response: OpenAIResponseObject