Run it on a schedule

Drop a Cron Schedule trigger into your project. Set a cron expression. Activate it. The project runs on that schedule until you deactivate it. That is the whole feature.

Good for daily digests, hourly data pulls, nightly reports, periodic health checks, or anything else that needs to happen on a clock without a human pressing a button.

A daily digest example

A project that runs every day at 9am, summarizes overnight AI news, and logs the summary.

The Cron node emits two outputs whenever it fires: scheduledTime (when it was supposed to fire, RFC 3339) and actualTime (when it actually fired). Every tick starts a full execution of the project, so nodes downstream of the trigger run automatically. You do not need to wire the tick into the next node unless you actually want to use the timestamp (for logging, a time-windowed query, or anything else that cares about when the tick happened).

Cron expressions

Standard 5-field cron: minute hour day month weekday.

If you hate reading cron, ask Tangle: "fire this every weekday at 9am". It will write the expression for you.

Activating

A Cron trigger does nothing until you activate it. Open the runner view of your project, find the trigger strip in the ActionBar (above the Run button), and click Start. The scheduler starts firing the project on your expression. The button flips to "Active: listening for events". Click again to stop.

Every fire is a full execution, visible on the Executions page like any other run. If one fails, the next scheduled fire still happens: failures do not stop the schedule.

Missed fires

The scheduler loop asks for the next upcoming fire time and sleeps until then. If the server is down when a scheduled fire should have happened, that fire is missed. On restart, the loop asks for the next upcoming fire after the current time, so missed ticks are not replayed.

For a daily report, missing one day is usually fine. For anything where every tick matters (billing, rate-limit windows, exact hourly sampling), store state in a durable database that tracks what was processed, so a catch-up run can detect and fill gaps.

Cron plus other triggers

A project can have more than one trigger. Cron plus an API endpoint is a common combo: the endpoint lets external services fire the project on demand, the cron keeps it ticking on its own as a fallback. Both triggers produce executions into the same graph.