phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-12-03 18:00:36 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

precommit

Browse source
This commit is contained in:
Omar Abdelwahab 2025-11-04 11:54:25 -08:00
parent d2103eb868
commit 0487496ce1
1 changed files with 13 additions and 43 deletions
Download patch file Download diff file Expand all files Collapse all files

56
src/llama_stack/providers/inline/agents/meta_reference/responses/streaming.py
View file

@ -68,9 +68,7 @@ from llama_stack.apis.inference import (
) )
from llama_stack.core.telemetry import tracing from llama_stack.core.telemetry import tracing
from llama_stack.log import get_logger from llama_stack.log import get_logger
from llama_stack.providers.utils.inference.prompt_adapter import ( from llama_stack.providers.utils.inference.prompt_adapter import interleaved_content_as_str
interleaved_content_as_str,
)
from .types import ChatCompletionContext, ChatCompletionResult from .types import ChatCompletionContext, ChatCompletionResult
from .utils import ( from .utils import (
@ -105,9 +103,7 @@ def convert_tooldef_to_chat_tool(tool_def):
""" """
from llama_stack.models.llama.datatypes import ToolDefinition from llama_stack.models.llama.datatypes import ToolDefinition
from llama_stack.providers.utils.inference.openai_compat import ( from llama_stack.providers.utils.inference.openai_compat import convert_tooldef_to_openai_tool
convert_tooldef_to_openai_tool,
)
internal_tool_def = ToolDefinition( internal_tool_def = ToolDefinition(
tool_name=tool_def.name, tool_name=tool_def.name,
@ -285,9 +281,7 @@ class StreamingResponseOrchestrator:
# add any approval requests required # add any approval requests required
for tool_call in approvals: for tool_call in approvals:
async for evt in self._add_mcp_approval_request( async for evt in self._add_mcp_approval_request(
tool_call.function.name, tool_call.function.name, tool_call.function.arguments, output_messages
tool_call.function.arguments,
output_messages,
): ):
yield evt yield evt
@ -396,12 +390,7 @@ class StreamingResponseOrchestrator:
else: else:
non_function_tool_calls.append(tool_call) non_function_tool_calls.append(tool_call)
return ( return function_tool_calls, non_function_tool_calls, approvals, next_turn_messages
function_tool_calls,
non_function_tool_calls,
approvals,
next_turn_messages,
)
def _accumulate_chunk_usage(self, chunk: OpenAIChatCompletionChunk) -> None: def _accumulate_chunk_usage(self, chunk: OpenAIChatCompletionChunk) -> None:
"""Accumulate usage from a streaming chunk into the response usage format.""" """Accumulate usage from a streaming chunk into the response usage format."""
@ -712,15 +701,12 @@ class StreamingResponseOrchestrator:
# Emit output_item.added event for the new function call # Emit output_item.added event for the new function call
self.sequence_number += 1 self.sequence_number += 1
is_mcp_tool = tool_call.function.name and tool_call.function.name in self.mcp_tool_to_server is_mcp_tool = tool_call.function.name and tool_call.function.name in self.mcp_tool_to_server
if not is_mcp_tool and tool_call.function.name not in [ if not is_mcp_tool and tool_call.function.name not in ["web_search","knowledge_search"]:
"web_search",
"knowledge_search",
]:
# for MCP tools (and even other non-function tools) we emit an output message item later # for MCP tools (and even other non-function tools) we emit an output message item later
function_call_item = OpenAIResponseOutputMessageFunctionToolCall( function_call_item = OpenAIResponseOutputMessageFunctionToolCall(
arguments="", # Will be filled incrementally via delta events arguments="", # Will be filled incrementally via delta events
call_id=tool_call.id or "", call_id=tool_call.id or "",
name=(tool_call.function.name if tool_call.function else ""), name=tool_call.function.name if tool_call.function else "",
id=tool_call_item_id, id=tool_call_item_id,
status="in_progress", status="in_progress",
) )
@ -1031,19 +1017,14 @@ class StreamingResponseOrchestrator:
sequence_number=self.sequence_number, sequence_number=self.sequence_number,
) )
async def _process_new_tools( async def _process_new_tools(self, tools: list[OpenAIResponseInputTool], output_messages: list[OpenAIResponseOutput]
self,
tools: list[OpenAIResponseInputTool],
output_messages: list[OpenAIResponseOutput],
) -> AsyncIterator[OpenAIResponseObjectStream]: ) -> AsyncIterator[OpenAIResponseObjectStream]:
"""Process all tools and emit appropriate streaming events.""" """Process all tools and emit appropriate streaming events."""
from openai.types.chat import ChatCompletionToolParam from openai.types.chat import ChatCompletionToolParam
from llama_stack.apis.tools import ToolDef from llama_stack.apis.tools import ToolDef
from llama_stack.models.llama.datatypes import ToolDefinition from llama_stack.models.llama.datatypes import ToolDefinition
from llama_stack.providers.utils.inference.openai_compat import ( from llama_stack.providers.utils.inference.openai_compat import convert_tooldef_to_openai_tool
convert_tooldef_to_openai_tool,
)
def make_openai_tool(tool_name: str, tool: ToolDef) -> ChatCompletionToolParam: def make_openai_tool(tool_name: str, tool: ToolDef) -> ChatCompletionToolParam:
tool_def = ToolDefinition( tool_def = ToolDefinition(
@ -1080,9 +1061,7 @@ class StreamingResponseOrchestrator:
raise ValueError(f"Llama Stack OpenAI Responses does not yet support tool type: {input_tool.type}") raise ValueError(f"Llama Stack OpenAI Responses does not yet support tool type: {input_tool.type}")
async def _process_mcp_tool( async def _process_mcp_tool(
self, self, mcp_tool: OpenAIResponseInputToolMCP, output_messages: list[OpenAIResponseOutput]
mcp_tool: OpenAIResponseInputToolMCP,
output_messages: list[OpenAIResponseOutput],
) -> AsyncIterator[OpenAIResponseObjectStream]: ) -> AsyncIterator[OpenAIResponseObjectStream]:
"""Process an MCP tool configuration and emit appropriate streaming events.""" """Process an MCP tool configuration and emit appropriate streaming events."""
from llama_stack.providers.utils.tools.mcp import list_mcp_tools from llama_stack.providers.utils.tools.mcp import list_mcp_tools
@ -1203,10 +1182,7 @@ class StreamingResponseOrchestrator:
return True return True
async def _add_mcp_approval_request( async def _add_mcp_approval_request(
self, self, tool_name: str, arguments: str, output_messages: list[OpenAIResponseOutput]
tool_name: str,
arguments: str,
output_messages: list[OpenAIResponseOutput],
) -> AsyncIterator[OpenAIResponseObjectStream]: ) -> AsyncIterator[OpenAIResponseObjectStream]:
mcp_server = self.mcp_tool_to_server[tool_name] mcp_server = self.mcp_tool_to_server[tool_name]
mcp_approval_request = OpenAIResponseMCPApprovalRequest( mcp_approval_request = OpenAIResponseMCPApprovalRequest(
@ -1233,9 +1209,7 @@ class StreamingResponseOrchestrator:
) )
async def _add_mcp_list_tools( async def _add_mcp_list_tools(
self, self, mcp_list_message: OpenAIResponseOutputMessageMCPListTools, output_messages: list[OpenAIResponseOutput]
mcp_list_message: OpenAIResponseOutputMessageMCPListTools,
output_messages: list[OpenAIResponseOutput],
) -> AsyncIterator[OpenAIResponseObjectStream]: ) -> AsyncIterator[OpenAIResponseObjectStream]:
# Add the MCP list message to output # Add the MCP list message to output
output_messages.append(mcp_list_message) output_messages.append(mcp_list_message)
@ -1268,15 +1242,11 @@ class StreamingResponseOrchestrator:
) )
async def _reuse_mcp_list_tools( async def _reuse_mcp_list_tools(
self, self, original: OpenAIResponseOutputMessageMCPListTools, output_messages: list[OpenAIResponseOutput]
original: OpenAIResponseOutputMessageMCPListTools,
output_messages: list[OpenAIResponseOutput],
) -> AsyncIterator[OpenAIResponseObjectStream]: ) -> AsyncIterator[OpenAIResponseObjectStream]:
for t in original.tools: for t in original.tools:
from llama_stack.models.llama.datatypes import ToolDefinition from llama_stack.models.llama.datatypes import ToolDefinition
from llama_stack.providers.utils.inference.openai_compat import ( from llama_stack.providers.utils.inference.openai_compat import convert_tooldef_to_openai_tool
convert_tooldef_to_openai_tool,
)
# convert from input_schema to map of ToolParamDefinitions... # convert from input_schema to map of ToolParamDefinitions...
tool_def = ToolDefinition( tool_def = ToolDefinition(