fix(openai-agents): Patch correct run_single_turn_streamed reference

[alexander-alderman-webb]

Summary

The run_single_turn_streamed() function is used in agents.run_internal.run_loop, unlike run_single_turn(). See

https://github.com/openai/openai-agents-python/blob/19e364c17344905ce6d17f41ea4fe084cba20388/src/agents/run_internal/run_loop.py#L1231

The integration previously created a reference that doesn't otherwise exist in openai-agents.

Add assertions on the Invoke Agent span in test_run_streamed_tool_execution_span().

Add a new test_run_streamed_error_handling() test, adapted from the test_error_handling() test with AgentRunner.run() replaced by AgentRunner.run_streamed().

Issues

Reminders

• Please add tests to validate your changes, and lint your code using uv run ruff.
• Add GH Issue ID & Linear ID (if applicable)
• PR title should use conventional commit style (feat:, fix:, ref:, meta:)
• For external contributors: CONTRIBUTING.md, Sentry SDK development docs, Discord community

[github-actions]

Codecov Results 📊

✅ 99162 passed | ⏭️ 6479 skipped | Total: 105641 | Pass Rate: 93.87% | Execution Time: 352m 33s

📊 Comparison with Base Branch

Metric	Change
Total Tests	📈 +60
Passed Tests	📈 +60
Failed Tests	—
Skipped Tests	—

All tests are passing successfully.

✅ Patch coverage is 100.00%. Project has 2499 uncovered lines.
✅ Project coverage is 89.99%. Comparing base (base) to head (head).

Coverage diff

@@            Coverage Diff             @@
##          main       #PR       +/-##
==========================================
+ Coverage    89.96%    89.99%    +0.03%
==========================================
  Files          193       193         —
  Lines        24954     24955        +1
  Branches      9004      9004         —
==========================================
+ Hits         22448     22456        +8
- Misses        2506      2499        -7
- Partials      1434      1436        +2

---

Generated by Codecov Action