Moderation

Every action creates a case, DMs the user if you let it, and logs itself.

#Cases

Bans, kicks, mutes, warns and their undos all create one, with a short ID. /case view <id> pulls it up, /history <user> shows every case on one person, /unwarn <id> removes a warning.

Reply to a message with m!ban and that message is captured as proof - text, author, attachments. /setproof attaches it after the fact.

/reason and /caseupdate edit a case. Both keep the original, so it reads as edited rather than rewritten.

#Punish settings

#Immune roles

Roles moderation cannot touch, globally or per action - so staff can be ban-immune but still muteable. Checked before anything happens, so an immune user is refused rather than half-punished.

#What the user is told

Under Privacy: Show proof and Show author. Turning the author off keeps individual staff out of an argument with a banned member.

The DM is a template. Put your appeal link in it.

#Predefined reasons

An alias that expands. Set raid to mean the full sentence and m!ban @user raid records all of it.

#Warn thresholds

A warn count triggers an action: three warns becomes a mute, five a ban. The escalation creates its own case, so the record shows both.

#Channel locking

/lock denies Send Messages for @everyone here. /lockall does it server-wide.

Ignored roles keep talking through a lock. Channels to lock limits what /lockall touches - empty means every text channel the bot can see.

Unlocking restores what was actually there before, rather than guessing.

#User reports

Members report someone to a channel staff watch. /report-blacklist blocks a member from reporting when the feature is being misused.

#Commands