Record tool response in tool run span

[alexmojaki]

Closes #2004

[Copilot]

Pull Request Overview

This PR ensures that the tool response is recorded within the tool run span, addressing issue #2004.

• Updated tests to assert a more comprehensive snapshot of tool attributes when content is included.
• Modified the tool execution flow in the tools module to capture and record the tool response in the current trace span.

Reviewed Changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 1 comment.

File	Description
tests/test_logfire.py	Updated snapshot assertions to check additional tool fields
pydantic_ai_slim/pydantic_ai/tools.py	Revised span handling to record the tool response within the span

Comments suppressed due to low confidence (1)

tests/test_logfire.py:548

• [nitpick] The updated snapshot assertion now compares multiple keys. Please ensure that dynamic fields like 'gen_ai.tool.call.id' and the JSON schema remain stable across test runs and consider adding comments to clarify the expected snapshot structure.

    assert result.output == snapshot('{"add_numbers":84}')

[github-actions]

Docs Preview

commit:	40324fa
Preview URL:	https://62ad36fa-pydantic-ai-previews.pydantic.workers.dev

[DouweM]

@alexmojaki Thanks, makes sense, approved!