v1.7 documentation
Discord bot modules,
explained.
What each CortexGuard module does, its default settings, and our recommended configuration for most servers. Open /config in Discord to change any of it. Anti-Raid, Anti-Spam, Anti-Mention Spam, Automod and Logs are on by default; every other module is opt-in.
← Back to homeSecurity
Anti-Raid
Watches join speed and reacts once too many accounts join within a short window — catching coordinated raids before they spread across your server. Safe mode (v1.5) can lock the server down instead of punishing, protecting real members during giveaways, events or viral invites. Since v1.7 it can also pause invites and DMs through Discord's incident actions (and, in safe mode, raise the verification level, restored afterwards), every sanctioned wave gets an “Undo this raid” button in the log channel, and the join risk score now reads Discord spam flags and suspicious names. Adjust it under ⚙️ Tuning.
Anti-Alt
Rejects accounts younger than a minimum age — the classic signal of throwaway or raid accounts. The length of the timeout, when that action is chosen, is adjustable since v1.7.
Captcha
Sends an image captcha on join (DM first, channel fallback) before granting server access, and kicks after a timeout if unsolved.
Anti Token Grabber
Turns specific channels into traps — any message posted there (from a token-grabber bot or otherwise) is instantly deleted and punished.
Anti-Nuke
Tracks bursts of destructive actions (mass channel/role deletion, mass bans, webhook spam) by a single account via the audit log, and stops the offender — ban, kick, 24h timeout or role strip — before real damage spreads. With Extended detection on, v1.7 also watches thread floods, scheduled-event floods and vanity URL changes, and strips dangerous roles from the offender when the sanction leaves them in the server (⚙️ Tuning).
Security Scan
On-demand audit of your server's permission structure — dangerous @everyone permissions, excess Administrators, missing 2FA requirement, low verification level — with a graded protection score.
Content Filters
Anti-Spam
Tracks messages per user in a sliding time window and reacts once the rate crosses the threshold. Since v1.7 it also catches the same text repeated in one channel (long messages count after two), attachment floods and sticker floods; repeated bot commands such as !play are ignored. Thresholds, mute length and spam-wave limits live under ⚙️ Tuning; exempt channels and categories are honoured.
Anti-Link
Deletes links from members without permission, with a per-server whitelist and exempt-channel list, optionally warning before escalating to a ban. The anti-evasion filter applies here too, and v1.6 recognises many more domain endings. Links inside forwarded messages are caught since v1.7, and the number of automatic strikes (last 24 hours) before a ban is adjustable.
Anti-Invite
Removes Discord server invite links instantly, stopping advertising and scam-server links. The anti-evasion filter applies here too. Since v1.7, your own server's invites are allowed by default (switch it off under ⚙️ Tuning) and invites hidden in forwarded messages are caught.
Anti-Mention Spam
Stops mass-mention attacks, a common raid and harassment tactic. Mentions are counted cumulatively over a time window (10s by default), so splitting pings across several messages no longer bypasses it. Can run on Discord's native AutoMod.
Automod
A regex-powered bad-word filter, compiled once per config change so it adds zero runtime overhead to every message. Includes a toggleable anti-evasion filter (strips invisible characters and lookalike letters) and an optional native Discord AutoMod boost that blocks messages before they post. Native escalations (kick/ban) spare your staff (v1.6).
Anti-Shout
Flags messages that are excessively uppercase ("SHOUTING") once they pass a minimum length. Since v1.6 it understands accented and non-Latin letters. Exempt channels and categories are honoured (v1.7).
Anti-File
Blocks risky file attachment types — executables, scripts, archives, web files — by extension — including double extensions like file.exe.png and externally linked files. 21 extensions blocked by default.
Anti-Emoji Spam
Flags messages with an excessive number of emojis (Unicode or custom) — a common low-effort spam and flood pattern. Exempt channels and categories are honoured (v1.7).
Anti-Homoglyph
Catches messages that use look-alike Unicode characters (Cyrillic, Greek, full-width) to sneak past your word filter. Exempt channels and categories are honoured (v1.7).
Automation
Auto-Slowmode
Automatically raises and lowers channel slowmode based on real-time message volume, with no manual adjustment needed.
Warn Expiry
Automatically clears warnings older than a set number of days, so an old minor infraction doesn't follow a member forever.
Reports
Message Report
Adds a "Report to Staff" option under right-click → Apps on any message, letting members flag content directly to your mod team with a reason.
Commands
Every command CortexGuard understands, outside of /config. Command names are also translated in each user's Discord client. Staff-only commands require the matching Discord permission (Ban Members, Kick Members, Moderate Members, or Administrator).
/config
Opens the full configuration panel covered by this Wiki — every module, every setting, all in one place. Also where you switch the bot between 10 languages (English, 中文, हिन्दी, Español, Français, العربية, Português, Русский, 日本語, Deutsch). Administrator only. Anti-Spam, Anti-Raid, Anti-Invite and Anti-Nuke have a ⚙️ Tuning button for finer settings; exempt channels sit next to exempt roles.
/mod
Moderation actions, grouped under one command:
/mod ban <member> [reason] [days]— bans a member, optionally deleting their last 0–7 days of messages./mod kick <member> [reason]— kicks a member from the server./mod mute <member> <duration> [reason]— times out a member for up to 7 days./mod unmute <member>— removes an active timeout./mod warn <member> <reason>— logs a warning against a member./mod warnings <member>— lists a member's warning history./mod clearwarnings <member>— wipes a member's warnings./mod unban <user ID> [reason]— unbans a user by their Discord ID./mod purge <amount> [member]— bulk-deletes up to 100 messages, optionally filtered to one member.
/history <member>
Full sanction history for a member — bot-tracked actions, manual actions from Discord's own audit log, and a shared-ban network lookup. Shown privately to the staff member who ran it.
/info [member]
Shows account and server-join information for a member — yourself if none is given. Warnings are only shown to staff and info roles.
/lockdown
Toggles lockdown on the current channel, instantly blocking @everyone from sending messages. Requires Manage Channels.
/raidpanic <on|off>
Emergency button: on locks every text channel at once, off restores only the channels raid panic locked. Requires Manage Server. Anti-Raid safe mode can trigger it automatically. Depending on Anti-Raid ⚙️ Tuning it also pauses invites and DMs and raises the verification level; off puts back only what it changed.
/snapshot save | list | restore [number]
Persistent copies of your server structure: roles (permissions, position, members) and channels (permission overwrites, categories). Taken automatically every 3 hours when enabled in /config → Anti-Nuke (on by default); 4 versions kept. restore recreates what is missing and never touches what exists. Requires Administrator.
/undo-raid [number]
Without a number, lists the latest detected raids. With a raid ID, reverses its sanctions: unbans and removes timeouts. Raid records are kept 7 days. Requires Moderate Members. The log message of each sanctioned wave also has an Undo button (Ban Members or Administrator).
/save
Saves this server's configuration to your personal profile (server owner only).
/restore
Restores your saved profile onto this server (server owner only). Asks for confirmation first (v1.6); restored mod and whitelist roles cannot be hijacked by look-alike roles. Channel and role IDs aren't transferred — those stay server-specific, as do shared-ban links and native AutoMod rules. Profiles unused for 365 days are deleted.
/report <message>
Sends a bug report or issue directly to the CortexGuard developer — limited to one per hour. Not for reporting server members, see below.
Right-click a message → Apps → Report to Staff
Lets any member flag a message to your staff team with a reason. Requires the Message Report module to be configured — see the Reports category above.