Dispatch vs Schedule for uptime.yml

[M1XZG]

Has anyone ditched the built-in cron schedule for uptime.yml in favor of dispatching it from an external host?

My issue is simple: the every-5-minute schedule almost never actually runs every 5 minutes. GitHub Actions cron is "best effort," and in my experience, the best it offers still leaves a lot to be desired, runs get delayed, skipped, or bunched together unpredictably.

Since I'm already running a self-hosted runner, my plan is to take scheduling into my own hands. I'd trigger the workflow externally using gh workflow run from a real Linux cron job on that same machine. That way I get actual, reliable intervals.

The catch is that I can't just edit uptime.yml directly to remove the schedule,it gets regenerated and I'd be right back where I started, fighting with GitHub's cron.

So my current thinking is to set a minimal schedule in .upptimerc.yml, something like once a day at midnight — just to keep the workflow valid, and then let my Linux box handle the real cadence:

uptime: "0 0 * * *"

That way, GitHub's scheduler is basically a fallback rather than the primary driver.

Has anyone taken a similar approach, or is there a cleaner way to disable the built-in schedule that I'm overlooking? Open to any suggestions!