Configure a video room's chat content-safety policy
Sets the room’s PROACTIVE chat content-safety policy — a pre-broadcast screen the send handler (POST /chat) applies BEFORE a message is fanned out or persisted. Under mode: block a matched message is rejected and never reaches another participant or the replay history; flag delivers it but marks it for moderator review; log delivers it unchanged and only records the match. Disabling the policy (enabled: false) preserves the default deliver-first behaviour. Host-only — requires the video:write role (owner | admin | developer) and is audited. Complements the reactive moderation endpoints (/chat/delete, /chat/clear).
Authorizations
Dashboard JWT token from Clerk
Path Parameters
Scheduled room id.
Body
Master switch. When false the send handler behaves exactly as before (deliver-first).
What to do with a matched message. block rejects it pre-broadcast; flag delivers-and-marks it for review; log delivers it unchanged and records the match.
block, flag, log Built-in content-safety categories to screen for (deduped server-side).
3profanity, pii, toxicity Operator-curated exact / word-boundary terms, applied whenever the policy is enabled regardless of categories.
2001 - 100