Resolved 16:25 UTC: the API and dashboard are available again. Root cause: during a platform-wide incident at our cloud hosting provider, our application stopped completing live requests at about 12:00 UTC while its database and other services stayed healthy. The provider blocked restarts for most of its incident; once a restart went through at 16:24 UTC, service recovered immediately on the same release. No data was lost. We are reviewing how to detect and recover from this faster, and will ...