Runs & Logs

​

Runs Page

The Runs page shows the execution history for your automations. Each run appears as a card with key details at a glance.

​

Run Cards

​

Viewing Run Details

Expand a run card to see the step-by-step execution log. If the composer spans multiple automations, the log entries are grouped by automation along a vertical timeline. Each automation group is represented by a node on the timeline showing the group’s overall status: Below each automation group header, the individual steps are listed. Each step shows:

• Status — whether the step succeeded, failed, was rejected, or is still running
• Step Input — the data passed into the step (collapsible)
• Step Output — the data the step produced (collapsible)
• Error — the error payload if the step failed (collapsible)
• Iteration Data — the current item and index when a step runs inside a Loop (collapsible)

​

Step Statuses

Each step in a log carries a status icon and color:

​

Collapsible Data Sections

Each data section (Step Input, Step Output, Error, Iteration Data) has a chevron next to its label. Click the label or chevron to expand or collapse that section. When a step has only one data section to show, it expands automatically. When multiple sections are present, they all start collapsed so you can focus on what you need.

​

UI and Raw Views

Every data section includes a UI / Raw toggle in the top-right corner:

• UI — renders the data as a formatted, human-readable display (default)
• Raw — shows the raw JSON tree you can inspect and copy

If you click UI or Raw while a section is collapsed, the section expands automatically.

​

HTTP Step Error Responses

For API Request steps that return a 4xx or 5xx HTTP status code, the response appears in an Error section (rather than Step Output) that includes the status code, response headers, and response body. JSON response bodies are automatically parsed and displayed in structured form.

​

Opening a Step in the Editor

When viewing a step’s log details, an Edit button appears in the top-right corner of the log panel. Clicking it closes the log panel and navigates you directly to that step on the workflow canvas — useful when you spot an issue in a run and want to fix the configuration immediately.

​

Copying Step Details

Summary rows in a step log (such as HTTP method, URL, status code, database table, notification recipient, and similar fields) each have a copy icon on the right. Click it to copy that value to your clipboard.

​

Loop and Pagination Labels

Steps that run multiple times display context badges next to the step name so you can identify exactly which execution you are looking at:

• Loop steps — each child step shows an iteration N badge indicating which loop iteration it belongs to. A divider line separates entries from different iterations.
• Paginated HTTP steps — each page request shows a page N badge indicating which page of results was fetched.
• Nested steps — the run card view shows a breadcrumb lineage (e.g. Parent step › Child step) so you can trace which parent step triggered the entry.

​

Errors and Rejected Filters

When viewing logs in the inline canvas Logs panel or in a composer’s expanded log view, two toggles appear in the header: Errors and Rejected. Use them to narrow the list down to the entries you care about:

• Errors keeps only steps that failed with an error.
• Rejected keeps only steps that were rejected during a Review step.

You can turn on either toggle on its own, or both together to see errored and rejected steps side by side. When at least one toggle is on:

• Runs that contain no matching entries are hidden entirely. The run count in the header updates to show how many runs remain visible.
• Within each remaining run, only the matching entries are shown — plus their ancestors (such as the parent Loop iteration or paginated API Request page they belong to), so each entry still has enough context to interpret it.
• If filtering hides every run, the panel shows a “No matching runs.” message instead of an empty list.

Turning both toggles off restores the complete log view.

​

Large Outputs

Step outputs are loaded on demand when you expand a log entry. If a step’s output is very large, only the first portion is shown initially — click Load more at the bottom of the output block to fetch the next chunk. The “Save Output to Library” button appears only after the full output has loaded. If an output contains a large array, the first 20 items are shown with a Load more button to reveal additional items in batches.

​

Saving Step Output to a Library Test

When a step has produced output, a Save Output to Library button appears at the bottom of that step’s log entry. Use this to capture real run output as a test variable in a library prompt’s test case — useful for building realistic test data without having to copy and paste manually.

​

Inline Canvas Logs

While working on the canvas, you can toggle the execution logs panel using the Logs button in the canvas header. This provides real-time updates during execution so you can watch your workflow progress step by step without leaving the canvas.

​

Real-Time Updates

Logs update in real time during execution. You can watch the current status of each step as the workflow progresses — no need to refresh.

​

Related Documentation