Triage

The triage queue (/triage) is where analysts review findings, record ground-truth labels, and initiate takedowns. It lists findings from the latest scan (and other recent scans), capped at 25 rows per page.

Filters

• Severity — restrict to a severity band.
• Feedback — show only labeled or unlabeled findings.
• AI-flagged false positives — surface findings the auditor believes are false positives.

Labeling

Each finding can be labeled confirmed or benign (optionally with a reason). Labels feed the adaptive learner:

• confirmed — a real impersonation / brand threat.
• benign — a false positive (e.g. an unrelated domain the brand itself owns, or a squatter’s profile with no brand overlap).

Use bulk labeling for batches of findings at once.

From triage to takedown

Confirmed findings are the takedown candidates. Each row offers a Generate takedown action that drafts an abuse-report email (see Takedowns).

Notes

Findings keep a notes field for analyst commentary. Social findings appear in a separate section with their own match signals and cross-links.

Next

• Takedowns
• Adaptive learning