Safety before conversation.

How moderation must work

1. Before publication: new posts and replies enter a private moderation queue, not the public feed.
2. Review: trained human moderators evaluate context, including bullying, threats, grooming, scams, contact requests and disclosure of personal information. Automated flags may help prioritize but cannot replace review.
3. Decision: approve, request edits, reject, or escalate to a safety lead. High-risk items remain hidden.
4. After publication: members can report content, mute/block accounts, and request removal; moderators can remove material and suspend accounts.
5. Appeal: authors can request review of a decision by someone other than the initial reviewer.

Protecting younger people

• Separate age-appropriate spaces with verified access controls before any minor can participate.
• No adult-to-minor direct messaging, contact sharing, private invitations or unmoderated mixed-age interaction.
• Never ask children or teens to publish school, address, phone, precise location or identifying photos.
• Escalation procedures and trained safeguarding reviewers are mandatory before launch.

What is not allowed

• Harassment, hate, threats, manipulation, grooming, sexual exploitation or encouragement of self-harm.
• Sharing another person's private details, identifying photos or confidential stories without permission.
• Requests to move vulnerable people into private channels, payment scams or unsolicited advertising.
• Presenting unverified medical claims as diagnoses or treatment instructions.

Privacy and urgent situations

Moderators should have minimum necessary access. Before launch, we need defined retention periods, deletion procedures, audit logs, account protection, and a secure method for handling reports. No user-generated content should be indexed before review.

If someone may be in immediate danger, peer support is not an emergency response. Find human safety resources.