Community Safety Guidelines

What these guidelines are for

These guidelines decide when Glitchworks puts someone on the BOLO (be on the look-out) list or the Blacklist, used by Glitchworks Guard and our custom bots. They exist so every decision is consistent, fair and explainable. If a report doesn't clearly meet a line below, no action is taken.

Version 1.0, November 2026.

1. The levels

People are never listed for: disagreeing with staff, arguments, being banned from one server for that server's own rules (for example roleplay rule breaks or "fail RP"), opinions, or anything without evidence.

2. The clear lines

Warnings

• Used for behaviour above that is minor, one-off and low-harm, for example a single heated insult, a first spam or advert run, or borderline behaviour reported by one server. Never used for child safety, threats or doxxing; those go straight to their listed outcome.
• A second Warning in the same category within its 3 months becomes a BOLO.
• The DM is calm and factual, never threatening or mocking. It says a report was received and in which category, that the account hasn't been restricted, that further reports may lead to servers being warned, and how to appeal with /appeal.
• If the person's DMs are closed, the Warning is shown to them the next time they use a Glitchworks bot command or log in to the website.

Bullying

• Bullying means repeated targeting, humiliating or ganging up on a person. Reports from more than one person or server about the same user are combined into one case so the pattern is visible.
• A one-off unkind remark gets a Warning. A pattern (2+ incidents) gets a BOLO. Continuing after a BOLO gets a Blacklist.

Doxxing (extra steps)

• The reporting server is asked to delete the post, and Guard offers to delete it if it has permission.
• The person who was doxxed is told how to report it to Discord and, if they feel unsafe, to the police.

2a. Case types, priority and who handles them

Target response times are what we aim for, not a guarantee.

Reviewer roles

• Safety Admin: oversees everything, including the audit and managing reviewers.
• Safeguarding Lead: the only role that sees child-safety cases. Kept to one or two trusted adults.
• Senior Reviewer: can issue Blacklists, and handles doxxing and threats.
• Reviewer: Warnings and BOLOs; can propose a Blacklist for a Senior to confirm.
• Appeals: any reviewer not involved in the original case.

Reviewers only see the case types their role allows, and never cases involving themselves or their own server.

3. Evidence standard

• Evidence must show who (Discord user ID, not just a name), what and when.
• Screenshots should be uncropped enough to show the username or ID and the date where possible. Logs collected by Guard are preferred.
• One reporter's word with no evidence is not enough, except child-safety concerns, which are always escalated.
• Evidence is only used for the case, seen only by Glitchworks reviewers, and deleted when the entry expires or is overturned.

4. How decisions are made

1. A server reports with /report (or on the website). A private case opens.
2. A Glitchworks reviewer checks the evidence against section 2.
3. BOLO: one reviewer can decide. Blacklist: two reviewers must agree (child safety: one reviewer can act immediately, and a second confirms within 24 hours).
4. The outcome is recorded on the case and sent to the reporter, who is asked if they're happy for it to close.
5. Reviewers must not decide cases involving their own server, friends, or themselves.

5. Appeals

• Anyone listed can appeal with /appeal or from their account page. They're told the category and reason (not the reporter's identity).
• A different reviewer handles the appeal within 7 days. Outcomes: upheld, reduced (Blacklist to BOLO), shortened, or removed.
• People can see the data held about them with /mydata and ask for it to be deleted where the law allows.

How to report and appeal, step by step.

6. When it's bigger than Discord

We're not the police. If someone may be in danger or a crime may have happened, we signpost and, where appropriate, report:

• Immediate danger: 999
• Non-emergency police: 101
• Child sexual abuse or grooming online: CEOP (ceop.police.uk)
• Fraud and scams: Action Fraud (actionfraud.police.uk)
• Discord Trust & Safety: dis.gd/report (for anything breaking Discord's rules)

We don't share case details with other organisations or communities unless the law requires it or there's a formal agreement and a lawful basis to do so.

7. Privacy (UK GDPR)

We rely on legitimate interests (keeping communities safe). We keep the minimum needed, restrict access to reviewers, apply the expiry periods above, and give people access, correction, appeal and deletion rights. Our privacy policy explains this in plain English.