Trust & Safety Moderation Queue

Trust & Safety is the moderation queue for content that the system flags as potentially unsafe — profanity, harassment, threats, or other policy concerns across messages, daily-sheet notes, and announcements. Flagged items are quarantined automatically; the queue lets a human reviewer see them and decide what to do next.

Where to Find It

Open the sidebar and click Trust & Safety under the Compliance section. Only managers, center admins, and super admins can see this entry.

What's in the Queue

Each row in the queue represents a single flagged event. The row shows when it was flagged, the surface it came from (parent message, staff message, daily-sheet note, announcement), a severity pill, and a short snippet of the content.

• Low — minor signal, mild profanity or borderline phrasing.
• Medium — moderate concern that warrants a human look.
• High — strong signal for harassment, threats, or self-harm. These also auto-escalate.

Filtering the Queue

The filter bar above the list lets you narrow by surface, by severity, or by free-text search. Filters compose — for example, you can show only High-severity parent messages from the past day. Clear all filters with the chip at the right of the bar.

Inspecting an Event

Click any row to open the read-only event drawer. The drawer shows the full content, who sent it, the recipient, the moderation categories that fired with their scores, and the timeline of system actions. Use this view to understand exactly why an item was flagged.

History and Settings

The History tab is reserved for resolved events — once review actions ship, decided events will appear there. The Settings tab will house per-surface threshold tuning. Both tabs currently show a placeholder while those features finish out the rollout.

Privacy and Audit

Every queue view, drawer open, and filter change is recorded in the Activity Log. Reviewers see the full content of flagged items so they can make accurate calls; senders only see the Pending state on their end.