The issue was identified, connected to failed internal authenticated requests, caused by a faulty deployment. The fix was already applied in a follow-up deployment.
Some jobs may have failed between 23:00 and 23:30 CEST on 7 August 2026. Please rerun them manually if needed.