I am testing a small app idea that calls the API a lot during development, and I want to keep costs under control while I experiment. I have set a usage limit and I am logging how many calls each feature makes. What do you do to keep testing costs predictable…