This repository is used as a crude datastore for data produced by content_updates.py.
Each time the wrapper run_and_push.sh runs, the output gets pushed to this repo.
If needed in the future we can walk through each git commit, unwrap the pickled data, and import it into a proper datastore. But using a Github repo for now is free and easy to setup.