Independent fallback channel alongside Projects Monitor

Projects Monitor is the primary dashboard for fleet/service status, but it has a correlated-failure problem: it depends on the very infrastructure it monitors being up, and it’s hosted externally (Osasco, Nuvem Hospedagem), which can itself go down independently of the vinhedo fleet. If Projects Monitor is unreachable, that alone shouldn’t leave the operator blind to whether the fleet is actually healthy. weekly-maintenance/hostevent-ping’s push-based heartbeats to healthchecks.io (plus UptimeRobot, external to this repo) give a status channel with no dependency on Projects Monitor or its host, at the cost of maintaining a second, partially-overlapping health-signal path (e.g. both channels independently check disk usage and reboot-required state).