fix(pydantic_ai): support pydantic-ai 2.0#538
Merged
Conversation
This makes Braintrust tracing work with pydantic-ai 2.0. Streamed agent events are traced again instead of failing because pydantic-ai changed that API. What users will see in Braintrust: stream-event spans continue to appear, completed direct streaming spans include the final model response, and errors during stream-event consumption are recorded on the span. Traces also include additional pydantic-ai 2.0 response details when pydantic-ai returns them, such as finish reason, provider URL/name, run id, and conversation id. This also removes an internal deprecation warning that could appear during normal setup_pydantic_ai() tracing.
Abhijeet Prasad (AbhiPrasad)
force-pushed
the
pydantic-2.0-support
branch
from
f252fd3 to
be77a03
Compare
Andrew Kent (realark)
approved these changes
Jun 30, 2026
|
|
||
| event = await self._iterator.__anext__() | ||
| self._wrapper._event_count += 1 | ||
| if hasattr(event, "output"): |
Contributor
There was a problem hiding this comment.
FYI, my large language friend says this is the wrong attribute to check. Not sure if it's legit:
In _AgentStreamEventsIteratorProxy.__anext__, the code checks hasattr(event, "output"), but the final event (AgentRunResultEvent) exposes .result, not .output — confirmed in both pydantic-ai 1.x and 2.0 source. As a result:
- _final_result stays None, so the span logs no output
- _maybe_create_tool_spans_from_messages is never called
Verified empirically: final_result=None, output=None. This is pre-existing (carried over verbatim from the old wrapper, not a regression), but it sits in the rewritten code and contradicts the PR's goal. To fix, the check likely needs getattr(event, "result", None) and should shape event.result. Existing tests pass only because they assert event count / span presence, not the span's output.
Member
Author
There was a problem hiding this comment.
good catch llm! will fix.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
resolves #536
AI Summary:
This makes Braintrust tracing work with pydantic-ai 2.0. Streamed agent events are traced again instead of failing because pydantic-ai changed that API.
What users will see in Braintrust: stream-event spans continue to appear, completed direct streaming spans include the final model response, and errors during stream-event consumption are recorded on the span.
Traces also include additional pydantic-ai 2.0 response details when pydantic-ai returns them, such as finish reason, provider URL/name, run id, and conversation id. This also removes an internal deprecation warning that could appear during normal setup_pydantic_ai() tracing.