We experienced a partial infrastructure outage overnight that led to failing builds (500 errors from API) for some users for up to four (4) hours.
This issue is now fixed. To all those affected, we apologize. Please get in touch if CI builds for your repo are still experiencing any behavior like this.