Define caching strategy for API calls for user secrets

[Parkreiner]

Blocked by #843

Essentially, we need to figure out exactly how we want to incorporate all the various API calls for users into the frontend. I think there's a few things we need to do:

• If we're not already doing this, figure out a consistent policy on how to handle sensitive data in the frontend, and decide what parts of various API responses we should be caching in our global query cache
• Make sure that all user-based secrets are grouped together predictably and consistently in the query cache, so that there's no guess work as far as how to invalidate data, and there's minimal risk of keeping more stale data in the cache than necessary
• (Ideally) Update any other caching for user-related data, and make sure that they follow the caching patterns defined for the new secrets
• (Potentially) Update the frontend contributing doc to reflect any new major changes in policy or caching strategy

[Parkreiner]

I'm assuming this work will probably take ~3 days, just so we have time to lay the groundwork, and figure things out. I expect that once this is done, the org version of this issue will take far less time