Jacob mentioned in a call that storing the mapping from asset type/identifier to "person to nag" can be done in a YAML file. Is it OK for this file to be public along w/ the rest of this reposiory, or would doing this imply that the repository go private?

Ended up deciding that it's probably best to store the mapping info in a model(s) rather than a YAML file, so I'll do that.

So I'm running into some issues w/ structured scopes... I thought Alex said in our last meeting that they were immutable--or at least their asset type and identifier were--but from the admin UI it seems only their asset type is immutable.

When I change the asset identifier in the admin UI, this doesn't trigger any changes in the last_activity_at timestamps of the reports associated with the scope/asset, so the h1sync command, which filters based on last_activity_at doesn't actually capture changes to the scope's identifier. (The same thing goes for eligible_for_bounty, the other scope-related field we're interested in.)

I might be able to work around this if there were some endpoint I could ping to just get a list of all structured scopes for a program, but there doesn't seem to be--the /programs/ endpoint doesn't seem to provide the structured_scope relationship, and the only other documented endpoint, /me/ doesn't either.

One solution might be to simply sync all reports every time we sync, rather than filtering by last_activity_at. This might not be a big deal if the number of reports stays fairly low.

Another solution might be to still sync with last_activity_at, but expect admins to manually re-sync all reports whenever a change occurs to a pre-existing scope. Or we could sync all reports at a lower frequency, e.g. daily, so that it doesn't bog things down.

Ok, I've just merged everything in this PR so far except for the most recent commit (c223f0a) into master, just because it's all tested and I don't want to have to deal with massive merge conflicts down the line as I put this WIP PR on hold for a bit.

This PR is essentially being blocked by #31 (which is just #21 (comment) turned into an issue).