This page documents the current public DealScope guide about repeated Salesforce AI summary tests.
The public guide frames this as repeated summary generation under the same workflow.
Its point is not to prove a universal limit from one session. Its point is to show why a broad claim like “unlimited” is too loose without more context.
The guide says some practical tests can look more permissive than teams expect.
But it also says the stronger conclusion is still narrower:
The public guide explicitly warns against assuming all Salesforce AI actions behave the same way.
It says summary generation and email drafting can have different practical constraints even inside the same product surface.