Every community on NbliK sets its own rules, and the Moderation Agent applies those rules the same way every time, day or night, across 30 languages. This guide covers the baseline we expect everywhere, on top of whatever a community adds.
The baseline everywhere
- No harassment, hate speech or threats, including targeted pile-ons.
- No impersonation of another member, a brand or a member of staff.
- No spam, scams, phishing links or coordinated inauthentic behaviour.
- No sexual content involving minors, and no content that sexualises children in any way. This is reported, not just removed.
- No sharing of another person’s private information without their consent.
- No content that promotes self harm or disordered eating.
How moderation works
The Moderation Agent reviews every post as it is made. It flags spam, abuse and impersonation in about three seconds, and it records the reason for every action it takes so your team can review the decision rather than guess at it.
- Agents run in suggest, approve or auto mode. Your team chooses which, per agent.
- Every action is written to an activity log with the rule it matched.
- Anything the agent is not confident about is escalated to a human rather than actioned.
- The agent learns from what your moderators override, so repeated corrections change future behaviour.
- Moderation keeps running even when a community has no agent credits left.
Appeals
If content was removed and you believe that was a mistake, contact the community’s own moderators first, since they set the rules and can restore it. If the issue involves NbliK itself, write to contact@nblik.com with a link to the content.