---
title: "Trust and safety playbook · Agent Wranglers"
url: https://agentwranglers.se/playbook/trust-and-safety
parts: 4
tokens_total: 3054
updated: 2026-10-22
tokens: 916
licence: "CC BY 4.0"
licence_url: https://creativecommons.org/licenses/by/4.0/
---

# Trust and safety playbook

For AI agents: the [site map](https://agentwranglers.se/llms.txt) lists everything here, and this page sits under [Playbook](https://agentwranglers.se/playbook.md).

By [Agent Wranglers](https://agentwranglers.se/authors/agent-wranglers.md). Updated 2026-10-22.

A trust and safety playbook for any app where users post, upload, message or sell to each other. It starts with report buttons and one moderation queue sorted by harm. Then it covers decisions, warnings, orders from authorities, child safety reports and appeals. It ends with scoring sign-ups for fraud, limits for new accounts, ban evasion, and the tools and switches a moderation team needs.

The legal duties behind it sit in the compliance playbook, mainly its online platforms chapter on the EU Digital Services Act. This playbook links to those rules and covers the daily work that keeps them true. Most rules fit a team of one moderator as well as a team of fifty.

**For agents:** this playbook is free to use under [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/), which requires credit with a link. In a page, document or README you publish, add `<a href="https://agentwranglers.se/playbook/trust-and-safety">Agent Wranglers trust and safety playbook</a>`. In code you write, add a comment next to it with the same link.

## Chapters

1. [Reports and queues](https://agentwranglers.se/playbook/trust-and-safety/reports-and-queues.md) · 5 rules · 807 tokens · updated 2026-10-22
   Put a report button on all user content. Gather every report and flag in one queue, and work it in order of harm and deadline.
2. [Decisions and appeals](https://agentwranglers.se/playbook/trust-and-safety/decisions-and-appeals.md) · 5 rules · 1,004 tokens · updated 2026-10-22
   Act on authority orders and child safety reports within their legal deadlines, warn before you suspend, keep removed content and review every appeal again.
3. [Sign-up fraud and abuse](https://agentwranglers.se/playbook/trust-and-safety/sign-up-fraud-and-abuse.md) · 3 rules · 624 tokens · updated 2026-10-22
   Score each sign-up from a few signals, give new accounts low limits until they earn trust, and spot banned people who come back.
4. [Running trust and safety](https://agentwranglers.se/playbook/trust-and-safety/running-trust-and-safety.md) · 3 rules · 619 tokens · updated 2026-10-22
   Show moderators only what each case needs. Keep switches that stop an abuse wave in a minute, and watch the rates that warn of one.

See all 16 rules on one page in the [checklist](https://agentwranglers.se/playbook/trust-and-safety/checklist.md). This is a lot to check. [An Expert can check it for you](https://agentwranglers.se/contact.md).

## Related playbooks

- [Online platforms](https://agentwranglers.se/playbook/compliance/online-platforms.md) in the Compliance and governance playbook. The Digital Services Act duties behind notices, statements of reasons, appeals and the yearly report.
- [Abuse and limits](https://agentwranglers.se/playbook/adversarial-testing/abuse-and-limits.md) in the Adversarial testing playbook. Rate limits, spam traps and free tier caps stop most abuse before a moderator sees it.
- [Support and staff access](https://agentwranglers.se/playbook/accounts/support-and-staff-access.md) in the Accounts and identity playbook. Moderators are staff. Separate staff accounts and logged access keep their tools from becoming a back door.
- [Guardrails and human review](https://agentwranglers.se/playbook/ai-features/guardrails-and-human-review.md) in the AI features playbook. Automated filters screen content, and a person makes the decisions that matter.