Make tests more impervious to changes

[lunakv]

Right now, the tests are breaking (again) because a new subtype was added and it's not included in our test data. Given how often this tends to happen, it would be nice to figure out a way to set up the tests in a way that doesn't break them every set release.

Some ideas on possible approaches to this:

• Generate our own in-repo set of test data and test against that instead of the live APIs.
• Somehow load the test data live from the APIs instead of hard coding them into code
• Try to limit tests to rules/cards that won't change often and just suck it up when they do.

[lunakv]

(An example I stumbled upon just now that illustrates why this is relevant:)

[Fryyyyy]

How about we just make the rule number a regex (\d+\.\d+ or something) and just make sure it work without caring specifically what number it ends up being