AutoMod Configuration

LAST EDIT Jun 18 2024

AutoMod is a powerful tool that helps moderators manage their channels with ease. This document will guide you through the steps of configuring AutoMod to ensure that your app is running smoothly, with your moderation policies applied. We'll cover topics such as selecting channel types, user permissions, configuring harms, configuring actions, and configuring Semantic Filtering. With the help of this guide, you'll be able to customize Stream AutoMod to your app’s needs.

Channel types


Our channel type overview page shows the configurability for all channel types active in the app. Moderation configuration can be edited per channel type. To create new channel types, please refer to the channel type documentation.

Channel Types in the moderation dashboard

By clicking on the channel type, you will be able to:

  • activate/deactivate Automod as a whole

  • select the corresponding moderation categories that can be activated and configured

Moderation Engines in the Channel Type page

Moderation actions


By clicking on the category, you can select the desired action to perform in chat when a message is deemed harmful.

  1. Block: Blocks the message

  2. Flag: Sends the message and flags it for review

  3. Bounce: bounces the message every time it is sent. The user is prompted to edit the message. Bounced messages are ephemeral, and are not visible to other channel members.

  4. Bounce, then Flag: Bounces the message the first time a user sends it in a channel, then flags it

  5. Bounce, then Block: Bounces the message the first time a user sends it in a channel, then blocks it

For more details, see our guide on Moderation actions.

Configuring Commercial Spam

Semantic filters


To configure semantic filters, go to the Semantic Filters section

Use the “New list” button to create a new phrase list. A phrase list is a group of phrases that serve as examples of the intent and meaning that you want to detect in incoming messages.

Phrases in a Phrase List should be at least 5 words long, and no longer than 120 words. The model works best when the phrases provided have clear meaning and intent, and are legible sentences with meaning. Short phrases and single words will still result in detections, but generally lack enough information to convey specific meaning and intent. For more information, we refer to our guide.

Once configured, a phrase list can be assigned to one or more channel types.

Once submitted, in the semantic filters page it will be possible able to edit, deactivate or delete the phrase list.

From the Channel Types > Channel > Semantic filters configuration page, it will be possible to select phrase lists and action for the specific channel type.

User permissions


Dashboard Moderators and Admins are automatically exempted from all advanced moderation and Blocklist flagging/blocking. Granular User permissions for moderation can be set on the app or channel type level. See our section on user permissions for more details.