Destaris
Browse the docs

Running & scheduling

How workflows run, on a schedule or on demand, and where to see results.

Once a workflow is saved, it runs on your machine — on the schedule you set, or whenever you run it by hand. This page covers both, plus where to watch what happened.

Scheduled runs

A trigger.cron node defines the schedule. The workflow then runs on your machine at that cadence, as long as the app is running. Each run executes the graph from the trigger forward, step by step.

Pair a schedule with the dedupe cache when you're polling a source, so overlapping or frequent runs only ever act on new items.

When a run is still going

Sometimes a run takes longer than the gap between scheduled fires — the clock comes around again while the last run is still working. Left unchecked, the new run could start over the top of the old one and process the same things twice.

Each workflow has an overlap setting that decides what happens then:

  • Skip (the default for scheduled workflows) — don't start the new run while the last one is still going. The skipped fire is recorded in the run history as "skipped — previous run still going", so you can see it was intentional, not a missed beat.
  • Queue — wait for the running one to finish, then run once. If several fires pile up while it's busy, they collapse into a single follow-up run, never a backlog.
  • Allow — let runs overlap. Only choose this when a workflow is safe to run alongside itself.

Running a workflow by hand always runs — the overlap setting only governs the schedule.

Running by hand

You can run any workflow once, immediately, from the editor's Run button — handy while you're building, or for workflows that start with a trigger.command. Each node lights up with its live status and output as the run progresses, so you can see exactly what happened at each step.

Some workflows need an input to run. When they do, you'll be prompted for it before the run starts.

The run history

Every run is recorded so you can see what ran, what each step produced, and what it cost. The dashboard is where you go to:

  • Confirm a scheduled workflow is firing as expected.
  • Inspect a specific run, node by node.
  • Spot failures and read the error from the step that broke.

Deterministic runs show $0; runs with AI steps show their cost, so the price of a workflow is always visible.