Policy reference
The vocabulary the API accepts, generated from the shared enums so it cannot drift. The Community Guidelines themselves are a published document pending approval — this is not a substitute for them.
Severity and response targets
P0On-call immediately — credible imminent harm, CSAM.
P1Within 4 hours — threats, exploitation.
STANDARDWithin 24–48 hours.
Actions
| No action | Report reviewed and closed. Still recorded — a dismissal is a decision. |
| Hide pending review | Reversible hold. The only action automation may take on its own. |
| Remove content | Content removed. Replies to a removed comment survive as a tombstone. |
| Label content | Content stays up with an advisory label. |
| Warn user | Notice to the account. No capability change. |
| Restrict capability | Removes one capability — posting, commenting, messaging, media or following. |
| Suspend temporary | Time-boxed suspension. Requires a duration, and a job lifts it. |
| Suspend permanent | Trust & Safety admin only. Support moderators cannot take this. |
| Restrict giving | |
| Suspend giving | |
| Lift giving penalty | |
| Suspend organization | |
| Reinstate organization | |
| Escalate | Hands the case up without deciding it. |
Report reasons
HarassmentHateSexual exploitationScam or fraudDoxxingImpersonationViolent threatCsamSpamDangerous misinformationSelf harm concernOther
The reporter’s chosen reason is what drives triage severity, so it is recorded as given rather than reinterpreted. Every decision needs a reason code naming the clause it rests on — an enforcement with no stated clause is one an appeal cannot fairly review.