Fix: Populate the node assistant's trace service with the existing callback

[SmittieC]

resolves #1088

Description

The issue was that the trace_service gets initialized here, but only callback is being passed in as config to the runnable. When we initialize the AssistantAdapter for the pipeline (see here), we fetch a fresh instance of the tracing service, but uninitialized.

The idea behind the PR is to essentially initialize the tracing service with the existing callback that already has a trace that we can piggyback on.

We don't have to do this for the LLMResponseWithPrompt node, because it's not trying to update the trace with extra metadata as the assistant node is doing.

User Impact

Users would be able to use langfuse tracing services with pipeline assistants.

Demo

N/A

Docs

N/A

[codecov-commenter]

Codecov Report

Attention: Patch coverage is 46.66667% with 8 lines in your changes missing coverage. Please review.

✅ All tests successful. No failed tests found.

Files with missing lines	Patch %	Lines
apps/service_providers/tracing/service.py	37.50%	5 Missing ⚠️
apps/pipelines/nodes/nodes.py	25.00%	3 Missing ⚠️

Additional details and impacted files

📢 Thoughts on this report? Let us know!

[snopoke]

nit: rename

Suggested change

	def from_callback_manager(self, callback_manager: CallbackManager):
	def initialize_from_callback_manager(self, callback_manager: CallbackManager):

[sentry-io]

Suspect Issues

This pull request was deployed and Sentry observed the following issues:

• ‼️ AttributeError: 'NoneType' object has no attribute 'initialize_from_callback_manager' apps.experiments.tasks.get_response_for_webchat... View Issue

_{Did you find this useful? React with a 👍 or 👎}