Scraper should run a `git pull` (or fetch+merge) first.

[beporter]

The script should ensure it's working from the most recent copy of its remote available before beginning.

We will potentially run into the same issue that we have with the skeleton's bin/update script, where the update script itself might be updated by the pull during execution of that very update. This is hard to detect and could put us into an endless loop if we tried, but it's something to keep in mind.

(Basically, if we detected that the scraper script was changed during the fetch, we should abort and re-execute the new version, but if that detection is incorrect or unreliable, we could repeat that infinitely and essentially fork-bomb the machine the script was running on.)

[beporter]

Regardless, we should at least perform a fetch+merge before updating the release/ folder and pushing back to the remote. The next scraping run will use any changes made to the scraper script itself from the pull done in the last scraping run.