Security

Automod

Automatically moderate your server with filters, keywords, and strike systems.


Note:

Slash commands for this page

Set this up in Discord with /configure feature:Content filter, or on the dashboard under Security › Automod.

Use the slash version of each command below. Prefix commands keep working until March 5, 2027. What is changing

  • ,filter → /configure feature:Content filter
  • ,filter heat → /configure feature:Content filter
  • ,filter ignore → /configure feature:Content filter
  • ,logs add → /configure feature:Server logs
  • ,logs ignore → /configure feature:Server logs
  • ,filter clear → /configure feature:Content filter
  • ,filter whitelist → /configure feature:Content filter

Overview

greed's Automod system allows you to automatically filter out unwanted content and manage user behavior through a strike system. It provides a multi-layered approach to server security including content filters, custom regex matching, and a persistent violation tracking system.

Tip:

Most advanced automod features, including the strike system and regex builder, are part of greed premium.

Heat filter | free

,filter heat on gives every member a heat score that rises with each message and cools down by itself. Messages add more heat when they repeat the last message, mention people or @everyone, carry attachments, links, many emojis or many lines. When the score reaches the threshold (100 by default, set it with ,filter heat on 150) greed applies the filter action. Regular chatting cools down faster than it heats up, so normal members are not affected.

When three accounts that joined in the last day trip the heat filter within a minute, greed starts a 10-minute heat panic: any account that joined in the last day and sends a message during it is timed out. When the heat action is a timeout, it doubles each time the same member trips heat again within an hour: 5 minutes, then 10, 20 and so on, up to 28 days.

Let a channel, role or member skip one filter

,filter ignore (filter) (channel, role or member) lets them skip just that filter, for example ,filter ignore invites #partners or ,filter ignore spam @Bots. Run it again to undo. A channel also covers the channels in its category. In /configure → Filter → Filters, each filter has Ignored channels and Ignored roles. Requires Manage Channels.

Logging Filter Actions

Run ,logs add #channel automod to post an entry every time the filter acts. Each entry shows the member, the channel, every filter the message tripped, the action taken, the member's strike count (and any escalation), and the removed message. Nickname resets by the Nicknames filter are logged too. One message produces one entry, however many filters it trips. Ignored channels and members from ,logs ignore are respected.

Filters

Filters are the primary mechanism for detecting and punishing specific types of content in real-time.

Warning:

Filters must be explicitly enabled with ,filter [name] on. Default thresholds are used if not specified. Staff members and exempt roles will bypass all filters. The filter does nothing until you toggle it on.

Note:

Premium filters (images, repetition, nsfw, malicious, walloftext) require a greed premium subscription. Most basic filters (spam, caps, emoji, links, invites, massmention, spoilers, musicfiles, nicknames, heat) are included with all servers and cost no premium.

Available Filters

Detects messages sent too quickly by a single user.

  • Threshold: Number of messages allowed within a 10-second window.
  • Usage:
    ,filter spam (setting) [threshold]

Detects messages containing excessive capital letters (only for messages 10+ characters).

  • Threshold: Maximum percentage (%) of uppercase characters allowed (default 70%).
  • Usage:
    ,filter caps (setting) [threshold]

Detects messages containing too many emojis.

  • Threshold: Maximum number of emojis allowed per message (default 10).
  • Usage:
    ,filter emoji (setting) [threshold]

Detects messages with a high number of user or role mentions.

  • Threshold: Maximum total mentions allowed per message (default 5).
  • Usage:
    ,filter massmention (setting) [threshold]

Filters out excessive use of spoiler tags or spoiler attachments.

  • Threshold: Maximum number of spoilers allowed per message (default 5).
  • Usage:
    ,filter spoilers (setting) [threshold]

Filters messages that contain too many image attachments.

  • Threshold: Maximum number of images allowed per message.
  • Usage:
    ,filter images (setting) [threshold]
Tip:

This filter is part of greed premium.

Automatically blocks Discord invite links to other servers.

  • Usage:
    ,filter invites (setting)

Filters out messages containing URLs or specific types of links. GIFs sent from Discord's GIF picker (Tenor, Giphy and Klipy links) never count as links, so the filter leaves them alone.

  • Usage:
    ,filter links (setting)

Detects repetitive messages or high similarity between consecutive messages.

  • Threshold: Number of similar messages before triggering.
  • Usage:
    ,filter repetition (setting) [threshold]
Tip:

This filter is part of greed premium.

Detects messages that are too long or have too many newlines.

  • Threshold: Maximum length in characters allowed.
  • Usage:
    ,filter walloftext (setting) [threshold]
Tip:

This filter is part of greed premium.

Matches messages against your custom keyword blacklist.

  • Usage:
    ,filter add (word)

Detects messages containing audio attachments.

  • Threshold: Maximum number of music files allowed per message (default 1).
  • Usage:
    ,filter musicfiles (setting) [threshold]
Tip:

This filter is part of greed premium.

Detects members with suspicious or filtered usernames.

  • Usage:
    ,filter nicknames (setting)

Automatically detects and filters NSFW images.

  • Usage:
    ,filter nsfw (setting)
Tip:

This filter is part of greed premium.

Blocks known phishing and malicious links.

  • Usage:
    ,filter malicious (setting)
Tip:

This filter is part of greed premium.

Strike System

Tip:

The strike system is a greed premium feature.

The strike system tracks repeated violations and applies escalating punishments automatically as users accumulate strikes.

You can also turn strikes on, set the most strikes a member can hold, how long each strike lasts, and add or remove levels on the dashboard under Security → Automod.

Strike Management

,filter strikes toggle (setting)
,filter strikes set (strikes) (action)
,filter strikes decay (hours)
,filter strikes cap (maximum)
,filter strikes view (member)
,filter strikes reset (member)
,filter strikes settings

Custom Regex Patterns

Tip:

Regex rules are a greed premium feature.

For advanced users, greed allows you to create custom Regular Expression patterns to catch complex varieties of unwanted content.

Regex Builder Utility

greed includes an interactive Regex Builder to help you create patterns without needing to learn complex regex syntax.

,filter regex

Regex Commands

Once you have a pattern, use these commands to manage your custom filters:

,filter regex add (name) (pattern)
,filter regex remove (name)
,filter regex list
,filter regex test (name) (text)

Filtering and Blacklist

Block specific words or domains from your server.

Blacklist (Domains)

,filter blacklist add (domain)
,filter blacklist remove example.com
,filter blacklist list

Keywords List

,filter add (word)
,filter remove (word)
,filter list
,filter clear

,filter clear removes every filtered keyword, so greed asks you to confirm first.

Exempt specific URLs from the links filter.

,filter links whitelist (url)

Filter Configuration

View All Settings

,filter settings

Set Keyword Punishment

Configure what action is taken when a keyword is detected.

,filter punishment (action)
Note:

Available actions: delete, timeout, warn, kick, ban, jail, strip, stripstaff

Discord AutoMod Migration

Push keyword filters to Discord's native AutoMod (blocks before your bot receives the message).

,filter wordmigrate

To stop using native Discord AutoMod and revert to greed's keyword filter:

,filter unmigrate

Whitelisting

Exempt specific members, roles, from automod checks.

,filter whitelist (word)
,filter exempt (role)

to remove a whitelisted word simply , run the same command ,filter whitelist (word) .

,filter whitelist goodword

Anti-self-react

Stops members reacting to their own messages. This is a premium feature and needs Manage Messages.

,antiselfreact on [action]
,antiselfreact off
,antiselfreact status
,antiselfreact ignore (member or role)
,antiselfreact unignore (member or role)

The action defaults to remove, which simply takes the reaction off and does nothing else. warn, timeout, kick and ban are also available and run through the same hierarchy and immunity checks as the rest of greed's moderation.

Note:

Only reactions on messages greed has seen recently are checked: it does not fetch old messages to find out who wrote them, because that would mean an API call for every reaction in your server.

Using the dashboard assistant

You can change automod by chatting with the assistant in the dashboard. Ask something like "block the word scam and time people out for it" or "turn off the caps filter", and it shows you the exact before and after to approve first.

Turn any filter on or off and change its threshold and action, set the keyword punishment, add or remove keywords, blacklisted terms and regex filters, exempt roles from automod, and show you the full configuration.

You need both Manage Server and Manage Channels. Without them the assistant will not show or change your filters.

Premium filters, blacklisted terms and regex filters need Server Premium, just like the commands. Regex patterns that are invalid or too slow to run are refused. The assistant will not help test, probe or get around automod.