Autopilot Module

The Autopilot module automatically protects your server from harmful messages and suspicious joins. You can configure each protection separately, choose what Discortics should ignore, and decide how repeat offenders are punished.

Most setup is completed through interactive commands in Discord. Autopilot uses Discord AutoMod for supported filters and Discortics checks for the remaining protections.

Features

• Block profanity, sexual content, and slurs.
• Stop mass mentions and @everyone abuse.
• Detect Discord spam, duplicate messages, and rapid message bursts.
• Block Discord invites, phishing links, link shorteners, NSFW links, and custom domains.
• Remove text walls, emoji floods, and heavily repeated content.
• Protect against suspicious joins, young accounts, default avatars, unwanted bots, and coordinated raids.
• Warn offenders and escalate repeated violations to mute, timeout, kick, or ban.
• Exempt trusted users, roles, and channels from individual protections.

Before You Begin

You need the Administrator permission to use any /autopilot setup command.

Discortics needs these permissions to open every setup panel and apply all available actions:

• View Channels, Send Messages, Embed Links, and Read Message History
• Manage Server and Manage Channels
• Manage Messages
• Manage Roles
• Moderate Members
• Kick Members and Ban Members

Keep the Discortics role above members and roles that Autopilot may moderate. Configure the server's muted role before enabling Mute punishments.

Autopilot does not apply message checks to server administrators. Each protection starts disabled, so opening its command does not enable it.

Autopilot Limits

All six protections are available on Public, Prime, and Custom bots.

Not every protection offers every ignored list. Profanity supports ignored roles and channels. Anti-Raid uses ignored users in its Discord panel and also respects ignored roles configured through the Dashboard.

Set Up Your First Protection

1. Choose the problem you want to stop.
2. Run its command, such as /autopilot antispam.
3. Select Enable.
4. Review the default detection settings.
5. Open Punishments and choose how violations should be handled.
6. Add ignored users, roles, or channels only where needed.
7. Send a safe test message from a non-administrator account.

Changes save as soon as you press a button, select an item, or send a requested value. There is no separate save button in Discord.

Command panels stop accepting interactions after five minutes. Run the command again when an old panel no longer responds. When Discortics asks you to type a value, respond within one minute or type cancel to stop.

Configure Punishments

Profanity, Anti-Mention, Anti-Spam, Anti-Link, and Anti-Flood each have their own punishment policy. Changing Anti-Spam punishments does not change Anti-Link punishments. Normal Anti-Raid join checks use the direct actions in the Anti-Raid panel instead of warning thresholds.

Open a protection command, then select Punishments. Available actions are:

• Warn: Creates a warning and moderation case for every detected violation. This is enabled by default.
• Mute: Applies the server's muted role after the configured number of warnings.
• Timeout: Applies a Discord timeout after the configured number of warnings.
• Kick: Removes the member after the configured number of warnings.
• Ban: Bans the member after the configured number of warnings.

For Mute, Timeout, Kick, and Ban, configure:

• Threshold: Number of Autopilot warnings required before the action is applied.
• Window: Time period in which those warnings must occur.
• Duration: Length of a mute, timeout, or temporary ban. Kick has no duration.
• Notify: Whether Discortics should try to DM the member about the action.
• Delete Messages: For bans, switches between deleting no recent messages and deleting messages from the previous 24 hours.

Warnings are counted separately for each protection. For example, three Anti-Spam warnings do not satisfy an Anti-Link threshold.

Important: Keep Warn enabled when using threshold-based punishments. Without a successful warning case, Autopilot cannot count the offense or apply its follow-up escalation.

The default policy warns and notifies on every violation. Mute, Timeout, Kick, and Ban are disabled until you enable them.

Profanity Protection

Run /autopilot profanity to configure blocked word categories.

Options Explained

Presets selects one or more Discord AutoMod wordsets:

• Profanity: Common severe profanity.
• Sexual Content: Explicit sexual terms.
• Slurs: Common slurs and hateful terms.

All three presets are selected by default when you first enable the protection.

Allowlist contains words that Discord AutoMod should not block. Use it to correct false positives. Select Add, then send comma-separated words. Each entry may contain up to 60 characters. Select Remove to remove named entries or Reset to clear the complete allowlist.

Ignored Roles and Ignored Channels bypass this protection. Profanity does not support ignored users.

When enabled, matching messages are blocked before they are posted.

Mention Protection

Run /autopilot antimention to stop mass mentions.

Choose a Mode

Simple uses Discord AutoMod. Set one Mention Limit from 1 to 50. A message is blocked when it exceeds Discord's mention threshold. Mention raid protection is enabled by default.

Advanced uses separate limits for each mention type:

• Total Mention Limit: Combined user, role, and channel mentions allowed in one message.
• User Mention Limit: Unique users allowed in one message.
• Role Mention Limit: Unique roles allowed in one message.
• Channel Mention Limit: Unique channels allowed in one message.
• Everyone Allowed: Whether messages containing @everyone or @here may pass. Members with Discord's Mention Everyone permission are not blocked by this check.

Each mention limit accepts a value from 1 to 50. A message violates Advanced mode when it contains more mentions than a configured limit.

Use Ignored Users, Ignored Roles, and Ignored Channels for trusted exceptions. In Advanced mode, ignored mentioned users and roles are also excluded from their matching mention counts.

Spam Protection

Run /autopilot antispam to stop unwanted message bursts.

Choose a Mode

Simple uses Discord's spam filter. Keep Discord Spam Baseline enabled to block messages identified by Discord AutoMod.

Advanced adds Discortics checks for duplicate messages and rapid repeated messages. The Discord spam baseline remains active unless you turn it off.

Options Explained

Duplicate Messages controls how many identical messages one member may send inside a time window. Enter limit/window, such as 5/10s.

Repeated Messages controls how many total messages one member may send inside a time window, even when their contents differ. Enter limit/window, such as 8/10s.

Include Stickers counts sticker-only messages during Advanced checks.

Include Threads applies Advanced checks inside public, private, and announcement threads. Turn it off to skip thread messages.

Ignored Users, Ignored Roles, and Ignored Channels bypass the selected spam checks.

Link Protection

Run /autopilot antilink to control links posted in your server.

Options Explained

Discord Invites blocks discord.gg, discord.com/invite, and discordapp.com/invite links.

Custom Blacklist blocks specific domains. Select Add, then send comma-separated domains such as example.com. Do not include paths. Discortics removes http://, https://, www., and paths when you enter them.

Custom Allowlist lets listed domains bypass link classification and Discord invite or custom blacklist rules. Use it for trusted domains that cause false positives.

Phishing Links blocks domains found in Discortics' phishing data.

Short Links blocks known URL shorteners.

NSFW Links blocks known adult domains.

Ignored Users, Ignored Roles, and Ignored Channels bypass link protection.

Autopilot deletes locally detected phishing, short, NSFW, and blacklisted links. Discord AutoMod blocks configured invites and custom blacklist keywords before posting where supported.

Flood Protection

Run /autopilot antiflood to stop oversized or repetitive messages.

Options Explained

Character Limit is the maximum message length before Autopilot treats it as a text wall. Enter a value from 250 to 4,000. The default is 1,500.

Emoji Limit is the maximum combined Unicode and custom emoji count. Enter a value from 5 to 2,000. The default is 15.

Duplicate Sensitivity detects heavily repeated content inside one message, such as a character or phrase pasted many times. Enter a value from 1 to 10. Higher values catch more repeated content and may produce more false positives. The default is 3.

Ignored Users, Ignored Roles, and Ignored Channels bypass flood protection.

Detected messages are deleted before the configured punishment policy runs.

Raid Protection

Run /autopilot antiraid to protect new member joins.

Anti-Raid can take direct action as soon as a member joins. Review every action carefully before enabling the module.

Smart Raid Detection

Smart Detection evaluates join bursts using join speed, account age, missing avatars, similar account creation times, and similar usernames.

• Sensitivity: Low, Medium, or High. Low triggers most easily. High requires stronger evidence.
• Min Joins: Minimum joins required before a raid can be detected, from 3 to 100.
• Action: None, Kick, Ban, Mute, or Timeout for members joining during detected raid mode.
• Punish Duration: Length of the Mute or Timeout action.
• Raid Mode Duration: How long raid mode remains active after the latest matching join, from 1 minute to 1 hour.
• Sweep: Also applies the selected action to members from the burst that activated raid mode.

While raid mode is active, every new join receives the selected action unless the account is ignored or temporarily whitelisted. When a fast join burst looks organic instead of malicious, Discortics sends an alert instead of punishing the group.

Join and Account Checks

Join Rate uses limit/window, such as 10/60s. Smart Detection uses this window and rate during its analysis. If Smart Detection is disabled, reaching the limit directly kicks the joining member.

New Accounts sets the youngest allowed account age and an action: None, Kick, Ban, Mute, or Timeout. Use a duration such as 7d for age. Mute and Timeout also ask for an action duration.

No Avatar applies None, Kick, Ban, Mute, or Timeout to accounts without a custom avatar. Mute and Timeout require a duration.

Bots controls bot joins:

• Allow All: Let every bot join.
• Allow Verified Only: Let Discord-verified bots join and apply the selected action to other bots.
• Block All: Apply the selected action to every bot join.

The bot action can be None, Kick, or Ban.

Fancy Names replaces unsupported decorative characters in a new member's nickname with readable characters when possible.

Global Bans, Panic Mode, and Exceptions

Global Bans checks new members against Discortics' shared global ban list. Choose None, Warn, Mute, Kick, or Ban. None disables this check.

Panic Mode immediately kicks every new member for a temporary period. Set a duration up to 30 minutes, then enable it.

Warning: Panic Mode does not distinguish legitimate members from raiders. Add expected members to the temporary whitelist before enabling it.

Temp Whitelist bypasses Anti-Raid for one user from 1 second to 1 day. Select Add, then send a user mention or ID followed by a duration, such as @member 1h30m. Select Remove and send the user again to end the exception early.

Ignored Users permanently bypass Anti-Raid. Use temporary whitelist entries for short events or planned joins instead.

Reset a Protection

1. Run the protection's /autopilot command.
2. Select Reset.
3. Review the confirmation message.
4. Select Yes, Reset.

Warning: Reset removes every custom setting for that protection and restores its defaults. This includes ignored lists, allowlists, detection values, and punishment settings. It cannot be undone automatically.

Reset protections individually. Resetting Anti-Link does not change Anti-Spam or any other Autopilot protection.

Temporary Anti-Raid whitelist entries are stored separately. Resetting Anti-Raid does not remove them; they remain until you remove them or their duration expires.

Dashboard Settings

Open Dashboard → Moderator → Autopilot to edit the main detection values, ignored lists, and punishment policies in one form. Dashboard changes are not saved until you use the save bar.

The Dashboard also provides the Log Channel setting used for smart raid alerts. If no Autopilot log channel is selected, raid alerts fall back to the moderation log channel when one is configured. It is also the place to turn Mention Raid Protection on or off; the Discord panel displays this setting but does not include a toggle for it.

Discord commands provide additional Anti-Raid controls that are not currently shown on the Dashboard, including Smart Detection, Global Bans, and temporary whitelist management.

Command Reference

These commands do not have slash-command options. Each opens an interactive settings panel.

Frequently Asked Questions

Why does Autopilot not act on my test message?

Autopilot skips server administrators. Test with a non-administrator account. Also confirm the protection is enabled, the test account is not covered by an ignored user or role, and the channel is not ignored.

Why are the buttons no longer working?

The settings panel expires after five minutes. Run the same /autopilot command again to open a current panel.

Why was a message blocked but no warning or punishment appeared?

Confirm Warn is enabled for that protection. Check that Discortics can create moderation cases and has the permissions required for the selected action. Role hierarchy can prevent Mute, Kick, Ban, or nickname changes.

Why can a muted member still send messages?

Check the muted role's channel permissions and move it below the Discortics role. The role must deny sending messages in channels where muted members should be unable to talk. You can use Timeout instead when you prefer Discord's built-in communication restriction.

What is the difference between Anti-Spam and Anti-Flood?

Anti-Spam watches message activity over time, including duplicate messages and rapid bursts. Anti-Flood inspects one message for excessive length, emoji, or repeated content. You can enable both.

Does the link allowlist override the blacklist?

Yes. An allowlisted domain is skipped by local link classification and is also included in the managed Discord AutoMod rules. Keep the allowlist limited to domains you trust.

Why did a fast join burst only create an alert?

With Smart Detection enabled, join speed is one signal. If the full group does not look like a raid, Discortics alerts staff instead of punishing likely legitimate members.

How do I stop Panic Mode early?

Run /autopilot antiraid, open Panic, and select Disable. Do not wait for the configured duration when legitimate members need to join immediately.