Moderation
EZChat provides some automated moderation features, to change these settings, visit the moderation tab in the dashboard.
Word Filter
Profanity Filter
EZChat can check messages against a list of profanities. They are labelled with a severity level, from 1 to 3, 3 being the most severe. On the dashboard, you can set a minimum severity threshhold so the profanity filter only uses words of that severity or higher.
Keyword Filter
You can also filter against a list of words or phrases. If a message contains any of the words or phrases, it will be flagged. The words are not case-sensitive, but every letter must match exactly.
URL Filter
EZChat will automatically flag any message containing a URL. You can manually turn this off in the moderation dashboard. You can also add URLs to either the block or allow list.
AI Moderation
EZChat can use OpenAI's chat moderation api to flag messages. It can flag messages based on the content of the message.
Moderation Webhook
If enabled, EZChat will send a POST request to the webhook url with the following JSON body:
{
"message": "<Message to be checked>",
}
The response should be a JSON object:
{
"isApproved": boolean
}
If isApproved is not true, the message will be blocked. If it is true, or EZChat does not recieve a response, the message will be broadcast.