Read the measurements
Check each component’s status and measurement timestamp. Components cover public operations such as API access, machine creation, commands, snapshots, application forwarding, desktop access, dashboard access, MCP, and cleanup when measured. A successful sample describes the measured operation at that time. It does not guarantee that your next request will have the same latency or that every workload is available.Missing observations
Missing measurements remain unknown. Gray history does not count as successful uptime. A monitoring-maintenance notice means probes are paused and health has not been remeasured; read the existing timestamps. The published history preserves failed observations. Latency summaries use successful samples, so read status alongside latency.Machine-readable status
status.json includes checkedAt, components, and announcements. Components carry their own checkedAt, status, latencyMs, and message. history.json carries observed history and metrics.