Violations
The Violations inbox lists ads that broke at least one of your rules. Your job here is to review them: confirm the real ones and mark the mistakes.
The list
Section titled “The list”Each row shows the Image, the App, the Rules broken, the Network, the Sightings and the Last seen time, plus the Review status. Newest sightings come first. Load more shows the next page.
Filters
Section titled “Filters”| Filter | Effect |
|---|---|
| App | One app. |
| Rule | Only ads that break this rule. |
| Network | Only ads served by this network. |
| From, To | Last seen on or after / on or before this day (UTC). |
| Score | Only ads whose highest rule score is at least this value. |
| Review | Unreviewed (the default), Confirmed, False positive or any. |
If the list is empty but some impressions were not classified this period, the page says “No violations among the ads that were classified. Impressions that were not classified are not included.”
A violation
Section titled “A violation”Open a row to see:
- the ad image (a signed link that expires soon; reload the page to get a fresh one);
- Why it was flagged: the explanation for each rule that was broken;
- Score per rule, and Scored with policy version {v};
- Where and when: network, placement, first and last seen, sightings, the ad response id (AdMob) and the content hash;
- the Source: Reported by your SDK or Uploaded by you.
Review it
Section titled “Review it”Under Your review, add an optional Note (up to 500 characters) and press:
- Confirm violation when the ad really breaks the rule, or
- False positive when it does not.
“Your answer is kept as feedback and is used to check how well the scoring works.” Owners and admins can review; viewers cannot. A reviewed violation leaves the default Unreviewed view.
Evidence pack
Section titled “Evidence pack”Export evidence pack (JSON) downloads the creative’s details, verdict and classification as a JSON file, for your records or for a conversation with an ad network.