Ban, kick, timeout, warn and purge with slash commands. Filter words, links, invites and spam automatically. Escalate repeat offenders without lifting a finger, and log every action to a channel. Your bot, your rules, no code.
Mika posted a discord.gg link in #general. Message deleted and member warned.
Rule: Server Invites. Warning 1 of 3.
Mika pinged 14 members in one message. Message deleted and member warned.
Rule: Mass Mentions. Warning 2 of 3.
Jordan: "Spamming after two automod warnings."
Case #482. Warning 3 of 3.
Three warnings reached. Mika has been timed out and sent the reason by DM.
Automatic punishment. Case #483.
The Moderation module ships these slash commands. Switch any of them off, restrict them to the roles you choose, and add your own.
/ban addBans a user from the server
/ban tempTemporarily ban a user
/ban removeRemoves a server ban from a user
/kickKick someone from the server
/mute addTimeout a user
/mute removeRemove a timeout from a user
/warnWarn a user
/case viewView the details of a case
/case removeRemove a case number from a user
/user historyView the moderation history of a user
/user clear-historyAdmin onlyClear the moderation history of a user
/note addAdd a moderator note to the user
/note viewView a user's moderator note
/note removeRemove a moderator note from a user
/purgeDelete multiple messages from a channel
/slowmodeModify the slowmode of a selected channel
/channel lockRemove the send messages permission from the channel
/channel unlockAdd the send messages permission back to the channel
/user nickModerate a user's nickname
/role addAdmin onlyAdd a role to a server member
/role removeAdmin onlyRemove a role from a server member
/role tempAdmin onlyTemporarily give a user a role
/voice deafDeafen and mute someone in voice chat
/voice undeafUndeafen and unmute someone in voice chat
/sayMake your bot say something
/server-infoGet information about the server
/pingCheck the latency of the bot
Six rules, three punishments, and the options that stop it getting in your staff's way. Pick a rule to see what it catches.
Keep a blacklist of words the bot should catch and a whitelist for words that look bad but are fine in your server. Both are edited from the dashboard or with the automod commands.
lol you absolute [blacklisted word], get out of my lobby
Removed by Automod: Bad Words
Automod will delete message and warn every time this rule matches.
Set how many warnings or timeouts a member gets before the bot steps in, and what it does when they get there. Try it.
Fires once a member reaches a set number of warnings or timeouts. Add as many rules as you like.
Stack as many rules as you like: three warnings for a timeout, five for a kick, seven for a ban. Warnings from moderators and from Automod both count.
The module sends a DM when someone is punished and posts an entry to your log channel. Both use the module's variables, so you write the template once.
{action}The punishment issued{moderator}Who issued it{time}When, as a Discord timestamp{case}The case number{reason}The reason givenPunishment Issued
You received a punishment from our server staff for breaking the server rules. Please revise the rules again, and check the details of your punishment below:
Nightfall's Moderation Team
Spam limits, raid handling, reports and appeals differ from server to server, so you build them from blocks. The punishment blocks open cases and log like the built in commands do.
Fires on every message in the channels you choose. Bots and staff roles are skipped with a condition.
A custom variable per member that resets after a short window, so the count only ever covers the last few seconds.
Set the numbers to suit your server. Nothing happens on the other branch.
Purge their recent messages, apply the timeout, and post to the mod log with the reason "Spam". The case shows up in their history like any other.
The API Request block ships with templates for the two moderation models most servers use. Paste your key into Secrets once, drop the block into a message event, and put a condition in front of the punishment.
https://api.openai.com/v1/moderations{
"model": "omni-moderation-latest",
"input": "{event_message_content}"
}{ai_check.response.results[0].flagged}true or false, plus a score for each category if you want finer control.
flagged equals trueOnly the true branch reaches the punishment. Everything else falls through untouched.
Keys are stored in your bot's Secrets and never appear in a message, a log or a variable. The model returns a value, your condition decides, and the bot acts. Nothing is banned by a model on its own.
Discord's built in AutoMod is worth keeping on. A BotGhost bot picks up everything it stops at.
| Feature | Your BotGhost bot | Discord AutoMod | Hosted moderation bots |
|---|---|---|---|
| Runs under your own bot name and avatar | Yes | Built in | No |
| Word, link, invite, caps, emoji and mention filters | Yes | Limited | Yes |
| Warnings with numbered cases and moderator notes | Yes | No | Varies |
| Automatic escalation after a set number of warnings | Yes | No | Varies |
| Temporary bans and temporary roles | Yes | No | Varies |
| Custom punishment DMs and log embeds with variables | Yes | No | Limited |
| Rules you design yourself: spam limits, raids, reports, appeals | Yes | No | No |
| AI harmful content and toxicity checks | Yes | Limited | Varies |
| Cost for everything above | Free | Free | Varies by plan |
Hosted bot features vary by bot and plan. Checked September 2026.
The same module, tuned very differently
Keep lobby chat playable at peak hours without a moderator glued to every channel.
Protect a fast growing audience from the scams and raids that follow every upload.
Consistent, transparent moderation that members can trust, with a full record behind every decision.
A quiet, professional server where the audit trail matters as much as the action.
Moderation runs on infrastructure that has been hosting Discord bots since 2018.
From a fresh bot to warnings, Automod and automatic punishments
Sign in to BotGhost, create a bot, and invite it to your server with the Moderate Members, Ban Members, Kick Members and Manage Messages permissions.
Open Modules in the dashboard and switch on Moderation. All of the commands above go live on your bot straight away.
Choose the roles that can run moderator and admin commands, or leave default permissions on so Manage Messages and Administrator decide. Pick the channel every action is logged to.
Enable the Automod module, pick the rules you want and the punishment for each. Add a few words to the blacklist and whitelist to start.
In the Moderation module, add a rule such as three warnings leads to a one hour timeout. Add another for five warnings leads to a ban.
Edit the punishment DM if you want your own wording, then run /warn on a test account. Check the DM, the log channel and /user history.
A short, honest guide to what a moderation bot should handle, what the module gives you, and where the builder takes over.
Moderation has three jobs. React fast when someone breaks a rule, keep a record so the next moderator knows what happened, and take the repetitive decisions off your staff entirely. Most bots do the first job well and the other two badly, which is why the same members keep getting the same first warning from different moderators.
The Moderation module on BotGhost is built around cases. Every ban, kick, timeout and warning is numbered, attached to the member, visible with one command, and counted towards the automatic punishments you set. The bot remembers, so your staff do not have to.
Writing a moderation bot by hand means a Discord library, permission handling, a database for cases, a scheduler for temporary bans, and a machine to keep it all running. BotGhost removes all of that. The module is a set of switches in the dashboard, and the bot is hosted for you.
The path from a new server to a moderated one looks like this:
The module covers the commands every server needs. The rules that differ from server to server, like how fast is too fast, how new is too new, and what a report should look like, are built in the event builder from the same blocks as everything else on your bot.
Because the punishment actions are blocks too, a rule you build yourself still opens a case, still sends the DM, and still lands in the log. A homemade anti spam rule and the built in warn command leave the same record behind.
AI moderation works best as a second opinion. Send the message to a model, get back a flag or a score, and put a condition in front of the action. Below the threshold nothing happens, above it the bot deletes and warns, and in the grey zone it posts to a staff channel for a human to decide.
The model reply is just another variable in the flow. It cannot ban anyone on its own, and every decision it influences is still a numbered case a moderator can review and remove.
Moderation is one part of the bot. Here is the rest of it.

Create a bot, switch on the Moderation module, and have warnings, Automod and automatic punishments running in your server today. No code, nothing to install.