A bug slipped through into production today that caused happo runs to fail across the board. The bad deploy has been rolled back so things should be fully functional again. We're working on a proper fix that should (most likely) be deployed later in the day.