Protection with accountability. Room to be yourself.

Sentinel is the safety system we are designing for NehemSpace. It brings together moderation, protection against scams and coordinated abuse, safer recommendations, teen protections, and community safeguards. Its purpose is to protect people without making ordinary conversation feel policed.

Not as a complete system. This page explains our design commitments, not a list of protections that are all operational today. NehemSpace remains in private beta with public registration closed. Sentinel needs implementation, testing, and specialist review before production enforcement.

Our intended approach puts your explicit choices ahead of passive watch time, separates disagreement from abuse, and treats reports as evidence to review—not votes to remove someone. We want safety decisions to be understandable and appealable. These are our design commitments, not a claim that other platforms have no safeguards or that NehemSpace has already outperformed them.

The goal is fewer unwanted recommendations, less reward for coordinated harassment, and a clearer path to correct mistakes. You should be able to disagree, choose your interests, and understand decisions that affect your account. We will need to demonstrate that through testing, feedback, and how well we handle real problems—not just promises.

The planned focus is conduct and context, not popularity or a political viewpoint. Criticism, debate, and disagreement belong here. Threats, exploitation, targeted harassment, scams, and doxxing do not. Showing someone’s wrongdoing does not justify blaming an entire race or other protected group. The Community Standards explain those boundaries.

A pile of reports must not be treated as proof. Sentinel is designed to assess the underlying evidence and context, and look for coordinated reporting abuse. Serious decisions should not be made simply because a person is unpopular.

No. Whether a post may stay, whether it should be recommended widely, and whether an account needs action are separate decisions. Something can be permitted without being suitable for every audience. Your chosen interests, privacy settings, and sensitive-content preferences should guide discovery instead of shock value alone.

Sentinel is not intended to be an unchecked AI judge. The plan starts with defined rules, known-harm checks, rate limits, and review workflows. Any future AI assistance must remain accountable. Ambiguous cases need human review, and severe decisions need reasons, an audit trail, and an appeal path. No automated system catches everything or gets every decision right.

Our design does not routinely inspect ordinary private-message contents or use them for advertising or recommendations. When someone reports a message, the submitted message and necessary surrounding context can be reviewed. Narrow legal and emergency processes need defined safeguards. Community moderators must not gain access to your private inbox.

The planned process gives affected people an explanation and a way to appeal. A successful appeal should correct the decision and related account-standing or distribution penalties. Account Standing is intended to be private to you and authorized safety staff—not a public reputation score.

The plan calls for stronger teen privacy, tighter messaging protections, safer discovery, and stricter mature-content limits. Turning 18 should not silently loosen someone’s settings. Community moderators can manage their own spaces, but that authority must not unlock private messages or private profile information.

Planned safeguards include checks for confirmed prohibited media, dangerous links, and abusive behavior. Crisis-related support should help rather than punish someone for asking for help. These features are not all available yet, can miss harm, and are not emergency services or medical care. If someone is in immediate danger, contact local emergency services.

We plan to test for false positives, review appeals, audit serious actions, and pause or roll back faulty modules. Feedback and evidence should guide improvements. Detailed enforcement procedures still need specialist review before release; this FAQ does not replace our Community Standards or Privacy page.