bug(tui): allows an invalid combined instance update and reboot request

[williampnvidia]

Problem

The instance update TUI allows a user to change an instance attribute, such as its operating system, and request an immediate reboot in the same operation.

The resulting request is rejected by the API:

Error: updating instance: API error 400: Failed to validate Instance update request data
Details: {"triggerReboot":"reboot cannot be triggered if Instance attributes are being updated in the same request"}

Expected behavior

The TUI should prevent this invalid combination.

If the user changes the name, description, operating system, or SSH key groups, the TUI should apply those changes without offering reboot options and tell the user to run instance reboot afterward.

Reboot-only requests should continue to work normally.

Testing

Add coverage confirming that the TUI cannot combine attribute changes and reboot options in one request.

[williampnvidia]

Another possible solution is to continue allowing the user to request a reboot, but perform two separate REST calls under the covers:

1. Apply the instance attribute changes.
2. If that succeeds, send a separate reboot request.

The TUI should make the two-step behavior clear. If the update fails, it must not request the reboot; if only the reboot fails, it should report that the attribute changes succeeded but the reboot did not.

[behroozrafii]

I chose to keep update and reboot as separate commands. This means each command sends one valid request, so we don’t need to handle cases where the update succeeds but the reboot fails. After updating an instance, the user can run instance reboot when ready.

Implemented in #5107.