Skip to main content

Scheduled Runs

Every run so far in this guide is triggered by something: you clicking a button, or a deploy pipeline calling the CI/CD trigger endpoint. A schedule is the third option: VeriWasp triggers the flow itself, on a cadence, so you catch issues that show up between deploys, not just at them.

How it works​

On a run's detail page, the Recurring schedule panel (right below Trigger from CI/CD) lets you attach a schedule to that flow: Hourly, Daily, or Weekly. Under the hood a schedule is nothing more than an automated caller of that same run's CI trigger token, fired by VeriWasp's own job scheduler instead of an external pipeline. It goes through the exact same clone-and-trigger path a CI trigger or the Re-run button does, including billing, lineage tracking, and regression detection.

Only one schedule per flow. If you want a different cadence, remove the existing one and add a new one. There's no need to juggle multiple schedules firing the same flow.

Credits​

A scheduled run costs a credit exactly like any other trigger, deducted the moment it fires. If your balance is too low when a schedule comes due, that firing is skipped, not queued, not retried early, and recorded as insufficient_credits. The schedule stays enabled and tries again on its next normal cadence; top up before then and it picks back up automatically.

Turn on notifications​

A scheduled run nobody ever checks the result of isn't worth much. Enable email or webhook notifications on the source run (see Creating a Playtest Run) before scheduling it, so a scheduled failure actually reaches you instead of sitting unread on a report page. VeriWasp doesn't require this, but the panel reminds you when you add a schedule.

Pausing, resuming, and removing​

The schedule panel shows its cadence, whether it's active or paused, when it's next due, and the outcome of its last firing (with a link straight to that run). Pause stops it firing without losing its history or making you recreate it later; Remove deletes it outright.

Precision​

Schedules fire on a best-effort basis, checked once a minute. "Hourly" means "roughly every hour," not "at the top of every hour to the second." For monitoring purposes (which is what this is for) that precision is more than enough; this isn't meant for time-sensitive automation.