Sentinel handles automated security incidents and obvious malicious behaviour. Your moderators remain responsible for community context and culture.
Designate channels that ordinary members should never post in. Sentinel can record the event, remove the message, timeout the account, or soft-ban and clean up recent activity.
Every intervention records the configured action separately from the actual outcome. Sentinel reports success, partial completion, or failure without retaining message content.
Premium link protection is planned as a layered pipeline: server rules first, Sentinel threat intelligence second, and cached external reputation providers when a domain is unknown.
Focused image and message-spam detection is planned alongside free reporting controls for immediate alerts, weekly aggregates, dashboard-only reporting, or both.
No music, economy, levelling, AI chat, reaction roles, tickets, staff surveillance, automatic cross-server punishments, or unsolicited direct messages.
Security incidents belong to Sentinel. Human context belongs to human moderators.
