How It Works
The actual logic behind the trap and the moderation flow.
The core loop
Spammers (or compromised accounts) typically blast every visible channel. Honeypot works because the trap channel is visible and looks ordinary, so their automation hits it and gets caught.
1
A trap channel is defined
The honeypot channel is a normal text channel chosen in
/honeypot.2
Message is detected
Any message sent in the honeypot channel triggers the flow. Slash command actions are attributed to the user; other bot messages are ignored.
3
Optional DM is sent
The bot attempts to DM the user first (unless “No DM” is enabled).
4
Action runs
Softban = ban then unban to remove recent messages. Ban = permanent ban.
5
Log + counter update
The action is logged and the warning message count is updated if enabled.
Action details
Softban uses a ban/unban sequence to remove the last hour of messages. Ban removes the user permanently and also clears their recent messages.
Softban is default
Softban is the default action because it removes spam while keeping the option to rejoin.
Edge cases
Honeypot is conservative and logs failures clearly:
Server owner
If the server owner triggers the honeypot, the bot cannot action them and will log a warning instead.
Missing permissions
If the bot lacks Ban Members or its role is too low, the action fails and a log warning is posted.