Right now, we are consistently hitting GitHub API rate limits in our tests, which makes it really hard to know if test failures are a problem or not. I suggest that we re-work our test suite so that we make as few API calls as possible (if any) and instead define a few expected responses that we hard-code into our tests, and use the functions downstream of the GraphQL connector to run the actual tests.
Right now, we are consistently hitting GitHub API rate limits in our tests, which makes it really hard to know if test failures are a problem or not. I suggest that we re-work our test suite so that we make as few API calls as possible (if any) and instead define a few expected responses that we hard-code into our tests, and use the functions downstream of the GraphQL connector to run the actual tests.