Spamming server errors during subagent execution #4226

Describe the bug

There are a lot of transient server errors, especially during subagent tasks. Every time I run anything using fleet or anything that requires a subagent task, I'll get 10-15 or more copies of "Response was interrupted due to a server error. Retrying...".

Nothing seems to be affected, and the /tasks read-only view does not show any issues.

Affected version

GitHub Copilot CLI 1.0.74-0

Steps to reproduce the behavior

No response

Expected behavior

It would be nice if the messages could be collapsed into one, for example, that updates a disruption count in place, or even just a little "Unstable connection, 5 retries" at the bottom somewhere.

Additional context

No response