Reduce flakiness in long-running stateless HTTP tool timeout test

[Copilot]

• Review new PR comments and confirm actionable feedback
• Inspect current implementation and recent commits for the affected test
• Update retry logic to use OperationCanceledException as requested
• Run targeted test for LongRunningToolCall_DoesNotTimeout_WhenNoEventStreamStore
• Run build and full test suite to validate no regressions
• Run automated code review and security scan
• Reply on addressed comment threads with the commit hash

---

💬 We'd love your input! Share your thoughts on Copilot coding agent in our 2 minute survey.

[halter73]

This is intended to reduce test failures like the following:

  Failed ModelContextProtocol.AspNetCore.Tests.MapMcpStatelessTests.LongRunningToolCall_DoesNotTimeout_WhenNoEventStreamStore [1 s]
  Error Message:
   System.Threading.Tasks.TaskCanceledException : The request was canceled due to the configured HttpClient.Timeout of 1 seconds elapsing.

https://github.com/modelcontextprotocol/csharp-sdk/actions/runs/22092487205/job/63841819265?pr=1295

We could consider using a timeout of larger than 1 second to reduce the flakiness, but then we'd have to make the LongRunningToolCall longer too. I think this retry logic should reduce the flakiness enough while still keeping the test shortish, but I'm open to ideas for alternative ways to test this that are faster or more reliable.