[internal] Failing crons and future token maintenance

[guineveresaenger]

Our token rotation for this provider is not automated, causing CI to fail whenever the token expires.

Discovery and fix for failures were complicated by:

• we do not have a record of how we configure our testing environment/where resources were being created
• tests were being skipped if the configuration variables for Actions were unset, masking longterm failures
• test names were too generic, making debugging individual tests difficult

A lot of this has been addressed:

• in Upgrade terraform-provider-github to v5.12.0 #284
• by setting the testing org to use github/moolumi
• by using an org admin's (@guineveresaenger's) token with repo and delete_repo permissions.

This token expires in 90 days.

To close out this issue, there are a few TODOs:

1. Capture CI setup for future maintainers to find and use
2. Long-term solution for the token
   a. Use an unexpirable token
   b. Create automation to rotate a token on the regular
   c. Add token rotation to the ops tasks