Bursts of GitHub webhook deliveries for runner jobs are now handled asynchronously, with account and pool lookups cached, so the service is less likely to return 502s and miss events. A new recovery worker also automatically asks GitHub to redeliver failed deliveries, reducing the chance that jobs sit stuck in GitHub’s queue.
Hive
GitHub webhook handling for runners is more resilient
Published
May 27, 2026 · 14:40 UTC
Repository
tuist/tuist