Final response flag when multiple agents participate in one invocation

[nagarajuprimefocus]

When multiple agents are involved in a single invocation, the is_final_response flag is being set to true prematurely (before all participating agents have completed their execution). This creates confusion in orchestrations since the orchestrator or root agent assumes the interaction is complete while other agents are still running or yet to provide their outputs.

What is the best way to handle this scenario?