GitHub Actions, Microsoft’s CI/CD service, suffered a major outage lasting approximately nine hours beginning August 6 (U.S. Eastern Time), causing cascading failures across Copilot code review, Copilot coding agent, GitHub Pages, and webhook throughput. Engineers detected performance degradation around 11:22 a.m. Eastern Time on August 6, with workflows failing to start or aborting mid-execution, API requests returning errors, and unexpected rate limits. Webhook throughput plummeted to as low as 15% of normal levels, while Enterprise Importer migration jobs halted completely. The root cause was invalid jobs being continuously and incorrectly assigned to execution runners. After a fix, the job success rate recovered to 99%, and GitHub announced at 8:01 p.m. Eastern Time on August 6 that new jobs and webhook-based workflows were processing normally, though Enterprise Importer migration jobs remain suspended. This incident follows a pattern of frequent disruptions, with over 20 incidents recorded per month in recent months. IncidentHub data shows 257 total incidents over the past 12 months, with 48 classified as major, accumulating over 112 hours of downtime. Capacity shortages and deployment issues were the leading root causes. The Zig programming language and the Ghostty terminal emulator have announced migrations away from the platform, citing persistent instability.