FluxyChatFeatures
Moderation console
Pending reports queue, mute/ban/delete actions, and moderator history.
Moderation console
Central moderation queue for AI-flagged and user-reported content. Moderators review, act, and audit history.
Console
Moderation (/moderation)
| Section | Purpose |
|---|---|
| Pending queue | AI flags and user reports awaiting review |
| Stats | Queue depth, top categories, period totals |
| Actions | Mute, ban, delete message, dismiss flag |
| History | moderation_events audit trail |
Use Room picker to scope queue to a specific room.
AI moderation
Text and image flags enqueue automatically when enabled:
| Variable | Purpose |
|---|---|
AI_MODERATION_ENABLED | Text/content classification |
VISUAL_MODERATION_ENABLED | Frame capture and image moderation on live/video |
Actions per severity: log only, mute user, delete message. Configured in worker moderation engine.
Visual moderation
Dashboard helper visual-moderation-client.ts captures video/canvas frames for server-side scoring when visual moderation is on.
Human-in-the-loop
Bulk approve/reject for HITL workflows. All actions are written to the audit log.
API
Moderation routes are under /admin/moderation/* and room-scoped moderation actions on message endpoints. See worker post-message-automations.js for enqueue logic.
Related
- Tool approvals: agent-side moderation tools
- Security