Skip to main content
A Venon Security guide

How do you set up AutoMod on Discord?

Detectors, thresholds and responses — in that order. With concrete starting values, a test and a list of traps.

6

Configuration steps

5

Final tests

4

Errors solved

~7 min

Reading time

The answer

The short answer

To set up AutoMod on Discord, first enable the detectors (anti-spam, anti-caps, the mention limit), then set thresholds that match the pace of conversation on your server, and finally choose the response to a violation — from deleting the message to a timeout. In Venon Security you do all of it on one screen: Panel → Security → Content protection, where AutoMod and LinkShield share the same save. The most important rule: start with gentle thresholds and tighten them based on the logs rather than on hunches — an overly aggressive filter punishes ordinary members and teaches them to work around the bot.

The content is checked against the bot's configuration — last verified: 2026-11-06.

Context

What the problem is

AutoMod is meant to take repetitive work off the moderators, but it is very easy to overshoot in either direction. Too gentle and it lets spam and adverts through; too strict and it deletes messages written in caps out of excitement, or blocks people pasting a few lines of code. The side effect of a badly set filter is worse than not having one: members start working around the rules on purpose.

Without an extra bot

What you can do with Discord alone

Before you reach for a bot, get everything out of what Discord gives you out of the box. Some problems are solved here, and the rest of the configuration will run more steadily.
  1. 1

    In Discord's native AutoMod (Server settings → AutoMod) enable mass-mention blocking and the basic spam rules.

  2. 2

    Set slowmode on the channels that regularly get crowded.

  3. 3

    List the channels where conversation is looser — some filters will get in the way there.

  4. 4

    Prepare a list of banned words and a list of exceptions for staff roles and integration bots.

Venon Security

What Venon Security adds

Only the things Discord's native tools do not handle, or handle without history and automation.
  • Venon's AutoMod combines content detectors with actions per detector — one response to spam, another to profanity, another to too many mentions.

  • The word blacklist and the regex blacklist let you target specific patterns, such as server adverts, without blocking whole categories of speech.

  • Every violation can go to the moderation log channel together with the content, so you can see whether a threshold is catching ordinary conversation.

Configuration

Configuration step by step

Every step says exactly where to perform it and which fields to set. The field names are the same as in the panel, so you can copy them one to one.

0 of 6 steps

0%
  1. Open the content protection screen

    Panel → Security → Content protection/venonsec-panel

    In Venon, AutoMod and LinkShield share one screen, because they are almost always configured together. You save the changes in one place, so it is impossible to enable a filter and forget about the action.

    • The AutoMod active switch is the main switch for message automoderation.
    • The Block links switch is tied to the VenonSec core (invite blocking) — bear that in mind if invites are meant to be allowed.
  2. Enable the detectors one at a time

    Panel → Security → Content protection → Detectors

    Do not turn everything on at once. Start one detector, wait a day and check the logs — only then move to the next. That way you know which filter produces false hits.

    • Anti-spam: catches runs of messages in a short time window.
    • Anti-caps: reacts to messages written mostly in capital letters.
    • Mention limit: defends against mass tagging of people and roles.
  3. Set the starting thresholds

    Panel → Security → Content protection → Detectors

    The values below work on servers with a normal pace of conversation. On a very busy chat raise them by a point or two rather than switching the detector off.

    • Spam threshold (messages): 5, Spam window (seconds): 5.
    • Caps threshold (%): 70, Min. message length (caps): 10 characters — so a short OK or YES is not punished.
    • Max. mentions in a message: 4.
  4. Choose the response to a violation

    Panel → Security → Content protection → Response to a violation

    The response should be proportionate. To start with, deleting the message plus a log entry is enough — enable automatic penalties once you see repeated abuse.

    • Delete the message: on.
    • Send a DM to the member: on, so they know what happened and why.
    • Log to the moderation log channel: on — without it you cannot judge the quality of the thresholds.
    • In the Actions per detector section, set a stricter response only where it is justified.
  5. Add the word and pattern blacklists

    Panel → Security → Content protection → Word blacklist / Regex blacklist

    The word blacklist works on exact words, the regex blacklist on patterns. Leave regex for the cases a list cannot describe, such as invites with substituted characters.

    • Keep the list short — long lists produce false hits inside innocent sentences.
    • Test every regex pattern on a test channel before leaving it in production.
    • After every larger change, check the logs to see whether it is catching ordinary conversation.
  6. Connect the penalty escalation thresholds

    Panel → Moderation → Sanctions and warnings → Sanction settings

    AutoMod deletes messages, but it is escalation that turns repeated violations into a real consequence. Without it the same member will spam every day.

    • Mute after (warnings), Kick after (warnings), Ban after (warnings): set a rising ladder, for example 3 / 5 / 7.
    • Warning expiry (days): 30–90 days, so a one-off mistake does not weigh forever.

The steps you tick are saved in your browser — you can come back and finish later.

Verification

How to check the configuration worked

A green switch in the panel is not proof. Go through the tests below — ideally on a second, ordinary account with no roles.
  1. 1

    On a test channel send 6 short messages in 5 seconds — they should disappear, and the moderation logs should carry an entry from the spam detector.

  2. 2

    Write a sentence in capitals longer than 10 characters — anti-caps should react; a short OK should pass with no action.

  3. 3

    Mention five people in one message with the limit set to 4 — the message should be removed.

  4. 4

    Check that the test account received a DM about the violation.

  5. 5

    After 24 hours go through the moderation log channel: if more than one hit in ten concerns ordinary conversation, the thresholds are too strict.

Diagnostics

When something does not work

The symptom, the cause and a specific fix. Start from the symptom you see — the cause is usually simpler than it seems.

AutoMod deletes messages from moderators and integration bots.

Cause
The detectors apply to everyone until you point out the exceptions.
Fix
In the Actions per detector section, limit the response for staff roles, and give integration bots a separate channel where the filters are off.

The filter catches ordinary conversation on a busy channel.

Cause
The spam window is too long for the pace of the chat, or the threshold is too low.
Fix
Raise the Spam threshold (messages) by 2 and shorten the Spam window (seconds) to 3–5. Judge the change by the next day's logs, not by a single report.

Messages disappear, but leave no trace in the logs.

Cause
The Log to the moderation log channel option is off, or the Moderation category has no channel.
Fix
Enable logging in the Response to a violation section and check in Panel → Moderation → Moderation log whether the Moderation category has a channel assigned.

A regex from the blacklist does not work, or blocks everything.

Cause
The pattern is too general, or contains special characters without escaping.
Fix
Start with the simplest pattern matching one specific case and extend it only after testing on a test channel.

Honestly

What this configuration does not solve

The boundaries matter as much as the capabilities. If you are counting on something from this list, you need a different solution — or simply a person.
  • AutoMod works on text. It does not analyse images, attachments or voice conversations.

  • A word filter does not understand context — a quote, a joke and a genuine insult look the same to it. Disputed cases still need a human.

  • A determined spammer will get around a word list by substituting characters. A regex blacklist raises the cost, but does not close the subject.

  • AutoMod will not replace raid protection — it reacts to message content, and a raid starts with a wave of joins.

  • Venon does not moderate private messages between members, including those sent after making contact on your server.

FAQ

FAQ – common questions

Which AutoMod thresholds should you start from?+

From 5 messages in a 5-second window for spam, 70 per cent capital letters with a minimum of 10 characters for caps, and 4 mentions per message. Watch the logs for a few days and tighten only the detectors that genuinely let abuse through.

Does Venon's AutoMod replace Discord's native AutoMod?+

It does not have to — they can run together. Discord's native filter blocks a message before it appears, while Venon's AutoMod adds actions per detector, regex blacklists, log entries and a connection to the penalty system.

How do you make an exception for an administrative channel?+

Do not enable the detectors you care least about there, and limit the responses for staff roles in the Actions per detector section. For channels carrying code or bot logs it is best to leave the filters off.

Ready?

Set this up in your own Venon Security panel

Every screen mentioned in the guide is in the web panel. You save the configuration by hand and see its state before anything takes effect on the server.