Moderation

Moderation is handled by members holding the MODERATOR, STAFF, DEVELOPER or ADMIN role. This page describes what those surfaces do — not because most readers need to use them, but because knowing what a moderator can see is part of knowing what your privacy settings mean.

The queues

Surface	Holds
Reports	Everything members have reported, resolved and unresolved
Penalties	Active and expired penalties, with their appeal threads
Chats	The decrypting, audited view of reported conversations
Categories and Tags	Taxonomy maintenance
Sources	The external sites items may point at
Countries	Geography used by the server browser
Redirects	Slug-change redirects
Status	Incidents shown on the public status page
Changelog	What shipped
Log	The moderation audit log
Activity	Site-wide activity

Limits that are built in, not conventions

Reading a reported chat message is audited. The moderation chat view decrypts on read and records the access. A report against a message carries no copy of it, so reading it is a deliberate, logged act rather than a side effect of triage.

A penalty’s appeal thread is visible to the penalised member. It is a conversation, not a note about them.

An automated server flag can only be lifted by the detector that raised it, or by a moderator. A moderator’s own flag is never touched by an automated job. See Penalties.

Staff can see who authored what, regardless of anonymity. A site where a moderator cannot identify the account behind a report is a site with no moderation. Anonymity governs what visitors see.

Privileged flags are rejected by the API for everybody else. isOfficial on an article or a server, primary on a group, the flagged state on a community — none of them can be set through a content API key, whoever it belongs to.

Site status

Incidents and live checks are published at /status. If something looks broken, that page is worth checking before filing a bug — it is placed directly under the bug tracker in the Resources menu for exactly that reason.