We traced the issue to a caching layer related to last week's Supabase downtime. Jobs are now progressing through the system. Please raise a support request if your job still appears stuck.

We're seeing delays across test runners. Investigating root cause.