I would like the ability to specify the max amount of OpenAI tokens the Announcement generation engine will use. This will give me better control on how much of my OpenAI spend is used.