> ## Documentation Index > Fetch the complete documentation index at: https://social-b97141fb-auto-generate-llmstxt.mintlify.site/llms.txt > Use this file to discover all available pages before exploring further. # Message Management > Guide for moderators and administrators to broadcast, monitor, filter, action, and govern chat messages in the Console This page explains how Console moderators and administrators manage real‑time chat: broadcasting, monitoring live streams, filtering, enforcing policy, handling mentions, and applying message-level actions. Send official announcements in Broadcast channels Track real‑time message flow & pause stream Surface flagged or threaded parent messages Delete, clear flags, mute, ban, blocklist Govern @user and @all usage Refine quality & reduce noise ## Broadcast Messages Broadcast channels allow only admin / authorized staff to post while all other members are read‑only—ideal for high-signal announcements, policy updates, campaigns. Create channel → select type Broadcast (see Channel Management for creation basics). Select the broadcast channel from list. Click Broadcast Message → choose message type → write content. @ to mention up to 30 members (strategic, not spammy). Click Broadcast and verify delivery. Overusing broadcast channels for routine chatter reduces member trust. Reserve for material value updates. ## Live Message Monitoring Admins can enter any eligible channel to observe real‑time messages, apply filters, enforce policies, and send official interventions. Crisis escalation, abuse response, event moderation, launch day traffic, sentiment sampling. View live flow, pause stream locally, filter flagged / threaded parents, inspect mentions, action messages. Pausing stream does not halt user posting; backlog is applied once unpaused. ### Pausing Data Stream If messages scroll too quickly for review, pause to stabilize viewport: Identify rapid influx / potential violation cluster. Click Pause (stream stops updating locally). Perform deletions, flag clearing, member actions. Unpause to apply buffered messages. Pause is a local console viewing control—users continue real‑time conversation unaffected. ## Message Types The system supports multiple structured types: | Type | Description | | ------ | -------------------------------------------------- | | Text | Plain textual content | | Image | Message containing an image asset | | File | Message containing a downloadable file | | Audio | Message containing an audio clip | | Custom | Developer-defined payload / structured custom data | Custom messages should be clearly explainable in moderation notes to ensure consistent enforcement. ## Filtering Displays only messages flagged by users/admins for priority moderation. Shows only parent messages that have reply threads (contextual summarization). Pause → Filter Flagged → Action → Resume for high-risk sweeps. ## Mentions Governance Tag specific member for directed clarification or escalation. Use sparingly to avoid dogpiling. Notifies entire channel—restrict to high-importance updates. Can be disabled in Feature Settings. Open channel in management view → Mentioned Users column lists resolved user displayName(UserId) pairs. Settings → Feature Settings → Uncheck "Allow to mention everyone in the channel". @user stays enabled. Excessive @all usage increases notification fatigue; audit usage frequency monthly. ## Message Actions Removes all flags post-review to restore normal status signals. Removes message content (shell may remain as sent by user with blank body). Use for policy violations. Extract problematic term(s) to preempt future violations. Temporarily prevent user from sending in channel; reversible. Force remove user + purge their messages from channel. Use for severe / repeated breaches. Ban before removing if long-term exclusion is intended; removing first may clear enforcement state. ## Enforcement Workflow Example Filter Flagged or observe pattern manually. Pause stream (optional) for heavy influx. Open context thread; inspect history & mentions. Delete / Ban / Blocklist / Mute as appropriate. Log rationale in internal system (if required). Clear flags where resolved → Resume stream. ## Best Practices * Use Pause only when necessary (avoid missing fast-moving signals). * Batch similar actions to reduce context switching. * Revisit blocklist quarterly for false positives. * Delete only when content violates policy; otherwise educate via reply. * Prefer Mute for temporary disruption vs immediate ban. * Escalate self-harm / threat content per protocol before deletion. * Consolidate multiple small announcements into a single structured broadcast. * Avoid marketing repetition; track engagement drop-off metrics. * Encourage targeted @user use for clarity. * Limit @all to critical updates (target threshold e.g., \<2 per week per channel). ## Troubleshooting | Issue | Likely Cause | Resolution | | ------------------------------ | ------------------------- | -------------------------------------------- | | Stream not resuming | Pause still active | Click Resume; verify UI state | | Flag not clearing | Permission or stale state | Refresh; confirm role; retry | | @all still works after disable | Cached settings | Refresh channel view; re-save setting | | Blocklist term not applied | Formatting mismatch | Add normalized (lowercase / trimmed) variant | | Broadcast failed | Network or missing rights | Retry; verify broadcast privilege | | User keeps reoffending | Mute instead of ban | Apply ban + document rationale | | Deleted message visible shell | Expected placeholder | No action—design for audit clarity | ## Quick Reference 30 users / message Flagged • Threaded Parent Clear • Delete • Blocklist • Mute • Ban ## Next Steps Configure channels & members Enhance enforcement workflows Review chat performance Manage user roles & restrictions