# User Guide

Protect your users from harm with remarkable ease.

[![Overview](../images/overview.png)](../images/overview.png)

Coop is the open source review and moderation tool from [ROOST](https://roost.tools) that provides a comprehensive solution for online safety:

- **[Automated Enforcement](automated-enforcement.md)**: Customizable multi-condition rules that evaluate every submitted item against signals to automatically take action or route reports to a human review queue based on your policies

  | [![Rules](../images/proactive-rules.png)](../images/proactive-rules.png) | [![Scam Rule](../images/proactive-rule-scam.png)](../images/proactive-rule-scam.png) |
  | ------------------------------------------------------------------------ | ------------------------------------------------------------------------------------ |
  | [![Policies](../images/policies.png)](../images/policies.png)            | [![Text bank](../images/text-string-bank.png)](../images/text-string-bank.png)       |

- **[Review Console](review-console.md)**: Configurable human review queues for moderators to quickly make complex policy-based moderation decisions, with added context and wellness features built-in

  | [![Review Console](../images/review-console.png)](../images/review-console.png)                            | [![Job page](../images/job-page.png)](../images/job-page.png)                            |
  | ---------------------------------------------------------------------------------------------------------- | ---------------------------------------------------------------------------------------- |
  | [![Review Console settings](../images/settings-review-console.png)](../images/settings-review-console.png) | [![Wellness settings](../images/settings-wellness.png)](../images/settings-wellness.png) |

- **[Integrations](../integrations/) & [Rich Content Processing](signals.md)**: Support for reviewing and matching against text content and threads, images and video media, account data, and custom content types using built-in signals plus free safety APIs from Google, OpenAI, Zentropi, and NCMEC out of the box

  | [![Integrations](../images/integrations.png)](../images/integrations.png) | [![Items](../images/items.png)](../images/items.png)                             |
  | ------------------------------------------------------------------------- | -------------------------------------------------------------------------------- |
  | [![NCMEC job](../images/ncmec-job.png)](../images/ncmec-job.png)          | [![HMA routing](../images/routing-rule-hma.png)](../images/routing-rule-hma.png) |

- **[Metrics & Reporting](metrics.md)**: Dashboards and detailed audit logs for accountability and insights into moderation effectiveness and trends

  | [![Dashboard](../images/dashboard.png)](../images/dashboard.png) | [![Recent decisions](../images/recent-decisions.png)](../images/recent-decisions.png) |
  | ---------------------------------------------------------------- | ------------------------------------------------------------------------------------- |

- **[API Integration](../api/)**: Simple REST and webhook APIs power seamless bidirectional platform integration for content ingestion, user reports, appeals, performing moderation actions, fetching additional items, and more

  | [![Action](../images/define-action.png)](../images/define-action.png) | [![API keys](../images/api-keys.png)](../images/api-keys.png) |
  | --------------------------------------------------------------------- | ------------------------------------------------------------- |

- **So much more.** Coop supports NCMEC reporting, hash matching, user strikes, appeals, investigation, bulk actioning, powerful user/role management, location-based enforcement, and single sign-on to name a few. And as an open source project with custom integration support, the potential for adaptability is limitless.

## How Coop works

This simplified diagram can help you better understand how data flows between a platform and Coop:

[![Simple Diagram](../images/diagram-simple.svg)](../images/diagram-simple.svg)

For more detail, review the [basic concepts](concepts.md), [technical architecture](../development/architecture.md), and [API reference](../api/).

## Getting started as an admin

We recommend beginning by familiarizing yourself with Coop's [basic concepts](concepts.md). Once you're up to speed:

1. Ensure you have an account and API key for your Coop instance

2. Define your [Item Types](administration.md#item-types); the kinds of content and actors on your platform

3. Input your detailed platform [Policies](administration.md#policies)

4. Define your [Actions](administration.md#actions) and expose [callback endpoints](../api/actions.md) so Coop can trigger enforcement on your platform

Once Coop is configured, your platform can:

1. Begin submitting items to Coop via the [Items API](../api/items.md) so they run through your proactive rules

2. Submit user reports via the [Report API](../api/report.md) to route them into review queues for your moderators

<style>
  .content img {
    margin-block: 0.5em;
    border: 0.5em solid color-mix(var(--bg) 80%, white);
    box-shadow: 0 0.25em 0.5em 0.25em rgba(0 0 0 / 0.1 );
    box-sizing: border-box;
  }

  .content table,
  .content table thead,
  .content table thead tr,
  .content table thead th,
  .content table tbody tr,
  .content table tbody tr td {
    background: none !important;
    border: none !important;
    padding: 0 0.5em;
  }
</style>