Text Moderation
The Text Moderation check filters incoming text for inappropriate or harmful content. It analyzes the text and routes it down one of two paths: "Clean" if the text passes moderation, or "Flagged" if the text contains content that was deemed inappropriate.
This is especially important for user-generated content that will be displayed on stream or read aloud by TTS. Without moderation, viewers could submit offensive messages through cheers, channel point redemptions, or chat commands. The Text Moderation check acts as a safety net, ensuring only appropriate content reaches your overlays and audio.
Inputs
| Name | Type | Description |
|---|---|---|
Text | string | The text to check for inappropriate content. |
Outputs
| Name | Type | Description |
|---|---|---|
Clean | string | The text, if it passed moderation. |
Flagged | string | The text, if it was flagged as inappropriate. |
Example
Use the Text Moderation check between a Channel Point Redeem trigger (for a TTS reward) and a Text to Speech effect. The viewer's message goes through moderation first. If it is clean, it gets read aloud by TTS. If it is flagged, you can either silently drop it (leave the Flagged output disconnected) or send a chat message telling the viewer their message was blocked.