Sensitive Community Mode

Safety-critical settings for a vulnerable community — a mental-health support group, a group for minors, a nonprofit serving an at-risk population — are usually scattered across a dozen individual toggles in a general-purpose product. Scattered settings have an obvious failure mode: someone forgets to turn one on, or a new admin doesn't know the full checklist exists, and the community operates with a gap nobody noticed.

Bundling the relevant settings into one named, auditable mode addresses that directly. Turning the mode on applies the whole set at once — display identity decoupled from real identity, crisis-content detection that routes to trained human moderators rather than an automated reply, mandatory review before sensitive content becomes visible, and tighter rate limits and audit — and any individual override away from the bundle's defaults is itself logged.

The crisis-routing piece deserves particular emphasis: for content indicating someone may be in crisis, the correct response is a trained human, every time — never an automated system attempting to handle it, however capable that system might otherwise be. That's treated as a firm rule inside the bundle, not a configurable option.