Execution

​

Mission control

• Landing view lists the most recent runs with status, trigger type, and duration.
• Filter by task, status, or time range to focus on a specific automation.
• Success rate tiles highlight reliability across the selected window.

​

Run details

• Click any entry to open its session view.
• Inspect the node-by-node timeline, inputs, outputs, and error messages.
• Download the recorded session to replay exactly what happened in the sandbox.
• For interrupted runs, use the log stream to spot the step that needs attention.
• node_debug runs launched from the builder show up here too; they stream only the selected slice and carry the same logs/screenshots as full runs.
• Child Task nodes surface their own nested run records directly in the timeline. Use the “child run” link inside a step to jump into the nested execution without leaving the parent view.
• Runs triggered from a Task’s Past runs tab deep-link into this view, so you can jump from task context to full telemetry in one click.

​

Actions from execution

• Cancel an in-flight run.
• Retry failed runs manually after correcting data or credentials.
• Export outputs in bulk when you need to share results with downstream teams.

​

Schedule runs

• Open a task and go to Run settings to create or edit a schedule (hourly, daily, weekly). Set the time zone, cadence, and optional pause/resume state.
• Scheduled runs use the task’s selected environment and credentials; change the runner in the same tab if a self-hosted desktop is required.
• Each scheduled trigger appears in Execution with its own run entry. You can cancel an in-flight scheduled run from Execution like any other run.
• To stop a cadence, pause or delete the schedule in Run settings; new runs won’t enqueue until you resume it.