Reading scan results

Finding scans

The Scans list supports a table or grid view and a rich filter bar: URL search, status, group, campaign, run, geo, tags, labels, and date range. Results are paginated.

Repeats and attempts

A campaign with repeats produces several scans for the same URL, geo, and device, which is why a run can hold many more rows than it has URLs. Each repeat is an ordinary scan with its own tags and evidence; the list marks its place in the group (#2/5). Read them as samples — a creative that behaves in four repeats and redirects to a scam in the fifth is exactly what repeats exist to surface. Repeats that shared one browser also shared its cookies, so a difference between them can be the visit count rather than the creative.

A retry is not a new scan. When a crawl fails for a technical reason and the campaign allows retries, the same scan is crawled again and the list shows which attempt it is on (Retry 1/3). A scan waiting for another attempt goes back to pending rather than through failed, so run counters and alerts only ever see the final outcome. The re-crawl runs on its own rather than inside the group's shared browser, so a Retry row starts from an empty cookie jar even when its siblings did not — the one row where the visit-count reading above does not hold.

The scan detail page

Open any scan to see the full picture on a single scrollable page:

• Info — the URL, linked campaign, geo, device/emulator, proxy, labels, and timing.
• Tags — the findings, each with a category and severity; for ad-tag scans they are split into creative vs landing so you can see where each one occurred.
• Classification — detected brand, AI content category, IAB v3 mapping, and any custom taxonomy assignments.
• Screenshots — the rendered creative plus one tab per landing page.
• Ad Detection — for an ad-discovery scan, one card per ad block found on the page, in the order the blocks sit on the page. Same cards as the public report below, with every tag your organization may see rather than only the public ones.
• Redirect chain — the full hop-by-hop tree from creative to final URL, each hop annotated with its own tags.
• Raw JSON — the complete machine payload, collapsible.

Actions

You can cancel a pending scan and copy a public report link — a read-only, unauthenticated view of the scan you can share with a partner who does not have a Kaminari Ad account.

For an ad-discovery scan the public report carries the same Ad Detection section as the scan detail page. Each block shows its creative preview and size, the ad network behind it (its name when we recognise the ad server, its domain always), the target URL it leads to, its tags, and a red Violation or green Clean verdict — red when one of the policy sets attached to the campaign fired on that block. A Creative / Landing switch above the grid flips every preview between the ad creative and the landing page it opens, so you can scan the destinations without opening a single block. Every card links to that block's own report — the public one on a shared page, so a partner can follow a single bad placement down to its redirect chain without an account.

The public grid differs in one way only: it lists the tags cleared for public reports. Internal and hidden tags never leave the platform, and the rule behind a violation is not named — so a block can be marked as a violation while its tag list stays empty, when the detection behind it is one we do not publish.