We provide a hosted Discord bot for easy integration. We only require non-intrusive, read + send message permissions on your server(s). Add the bot below:
Add to Server Alternatively, if you'd like to self-host / customize the bot, we've provided the code and installation instructions below.
We scan all submitted messages for hateful content via Toxicity Moderation API, and flag messages above a configured threshold for further review. The following commands are avaliable to manage the bot (users must "ban_member" permissions to access these commands):
Commands:
$cminit [#channel] [threshold]
Configure the channel to send reports to, and confidence level (from 0 to 1) to report at
$status [enabled|disabled]
Enable or disable message reports
$help
Per our Privacy Policy, we do not store or log any submitted content, so rest assured the confidentiality of your conversations are secure.
Required Libraries:
apt-get install python3
pip3 install discord.py requests
Required config.json file:
{
"api_token": "MODERATE_HATESPEECH_API_TOKEN",
"bot_token": "DISCORD_BOT_TOKEN",
"channel_id": "CHANNEL_ID_FOR_REPORTS",
"threshold": CONFIDENCE_THRESHOLD (float)
}
Pass the Channel ID of the channel you want flagged messages to be sent to. Because this setting is set per bot-instance, we do not recommend using a self-hosted bot on more than one server / runtime. Our hosted version runs a slightly modified version to mitigate this issue.
