Opening a run takes you to its results page, where you can see every asset's outcome, understand why it was decided, and make your own decisions. This article walks through every part of that page.
When to use this
Open a run whenever it finishes (status Processed) - or even while it's still running - to review outcomes, clear the Needs Review queue, and export or save the results.
The run header
At the top of the page you'll find:
Run name - click it to rename.
Status and time - Running (with live progress), Processed, Stopped, or Failed, plus when it completed.
Rules - chips for each rule the run applied. Click a chip to see that rule's setup at a glance: its Decision Behavior (review threshold and rejection reason) and its configuration.
Needs Review counter - how many assets still need a human decision. Click it to jump straight into reviewing them.
Export CSV - download all results as a spreadsheet.
Save Metadata to Assets - write results to your assets in the Media Library (see "Save moderation results as metadata").
Tip: Export the CSV when stakeholders outside the app need the outcome.
Filter and browse the results
Status tabs - switch between All, Needs Review, Rejected, and Approved. Each tab shows its count.
Rule filter - the All Rules dropdown narrows results to a single rule, so you can see everything one check flagged. When a rule is selected you can also sort by score and search assets by name.
View modes - Grid (large cards), Compact (small cards), or Preview (one asset at a time, full detail - the best mode for reviewing).
Each asset card shows a thumbnail, the asset's name, and its result badge. Select a card to see its details in the sidebar; double-click to open it in Preview.
Tip: Use the rule filter to calibrate a rule: seeing all of its decisions together quickly shows whether it's too strict or too lenient.
The asset details sidebar
Selecting an asset opens its details on the right:
Details - the asset's File Information (format, dimensions, file size, created date, plus buttons to copy the Public ID and the asset URL) followed by Rule Results: each rule with its outcome (Pass, Fail, or Review), confidence score, and an explanation of what was found. For rejected assets, the rejection reasons appear here.
Metadata - shown when the asset has metadata: the values saved (or ready to be saved) on this asset.
History - a full audit trail: automatic decisions, manual approvals and rejections, who made them, when, and any reasons given.
Review assets one by one (Preview)
Preview mode is built for working through a queue:
Filter to Needs Review and open the first asset in Preview.
Inspect the asset - images at full size, videos with a player. If a rule's explanation mentions a moment in a video, click the timestamp to jump straight to it.
Decide:
Approve (keyboard: A)
Reject (keyboard: R) - you can select rejection reasons and add a note.
Skip (keyboard: S) - move on without deciding.
Tip: Review with the keyboard (A / R / S) - it's dramatically faster for large queues.
Your manual decision overrides the automatic result and is recorded in the asset's History. Made a mistake? Select Clear Manual Decision to return the asset to its automatic result.
When everything is reviewed, the Needs Review tab shows "Nothing to review."
Bulk decisions
Select multiple assets in Grid or Compact view to approve or reject them together. The sidebar shows a summary of your selection, and bulk rejections can include shared rejection reasons.
What happens next
Decisions take effect immediately in the results and in the History trail.
If automatic metadata saving is on, updated results are written to your assets; otherwise use Save Metadata to Assets when you're done reviewing.
The run's summary on Moderation Runs reflects your final numbers.