Documentation

Three steps to get ZT Plush running in your server:

1. Click the invite link and choose your server.
2. Authorize the requested permissions: Kick Members, Ban Members, Moderate Members, Manage Roles, Manage Channels, Manage Messages, View Audit Log, plus the basics for sending (Send Messages, Embed Links, Attach Files, Add Reactions, Use External Emojis, Read Message History, Send Messages in Threads) and for voice (Connect, Speak, Use Voice Activity).
3. In any channel, run /info bot for a full feature overview, or /help for category-organized commands.

/config opens a setup panel for each part of your server: welcome messages, tickets, counting, logs and more. The panels replace the old setup commands. /config needs Manage Server to show up.

How panels work

• Run /config overview to see every setting at a glance and jump to one from a menu, or open a panel directly, like /config welcome.
• A panel is a private message only you can see, with buttons and menus. Everything saves immediately. There is no Save button.
• A panel stops responding after 10 minutes without use. Run the command again for a fresh one.
• Permissions are checked on every press, not only when the panel opens. If your role changes, or the feature is paused, the panel says so instead of failing quietly.
• Each panel checks who may use it, and the table shows what. Most need a staff tier (Senior+ for most). Some also open for Manage Server instead, and Content mode needs Manage Server. A server can move any permission to another tier.

A few things stay on the dashboard because they need more room: welcome banner images and rotating messages, boost rewards by time or count, custom embeds, and the full permission editor.

Content mode decides how relaxed your server is. Every server starts Locked. Only people with Manage Server can change it, with /config content or on the dashboard's Content mode page.

• Locked (default): clean everywhere. Swearing and crude words are blocked, and sensitive media is blocked everywhere, age-restricted channels included.
• Unlocked: swearing and crude words are allowed everywhere. Sensitive media follows the media rule below.

Media rule (Unlocked only)

• Age-restricted only (default): sensitive media only shows in channels Discord has marked age-restricted. Anywhere else the bot says it's flagged and points to an age-restricted channel.
• Strict: sensitive media is blocked everywhere, age-restricted channels too.

There is no “anywhere” option, on purpose.

What counts as sensitive media? Whatever the source flags: X and Reddit mark posts as sensitive, and some videos report an age limit. If the source doesn't flag something, ZT Plush treats it as normal.

What Locked filters

In a Locked server, swearing and crude words are filtered out of the word lists for Hangman and Word-Chain, out of /quote and the Quote menu, out of the names /grab offers for emoji and stickers, and out of what /fun commands say back. Unlocked servers allow them. Media from /download and /linkembed follows the media rule.

Every change is logged: a line in your mod log channel says who changed it and to what. If no log channel is set, the change still saves.

To see the current setting, run /info server.

Before diving into individual features, here's how the recurring patterns work across the bot. Three things show up everywhere: variables, permission tiers, and defaults.

Variables (the curly braces)

Many features let you customize message templates. Anywhere you see {like_this} in a template, the bot replaces it with real values when the message goes out. Welcome and goodbye messages, booster thank-yous, birthday messages, custom command replies and giveaway templates all support them. Scheduled messages and Mass-Tag don't: they are sent exactly as you wrote them.

Feature-specific variables (like {prize} for giveaways or {boostcount} for booster messages) are documented in each feature section. If a variable doesn't apply to the message you're writing, it just gets left as text. The bot won't crash.

Permission tiers

Every command lists its default tier. Everyone means any member. Helper+ means the Helper tier or higher (so Helper, Mod, Senior, and Owner, plus any custom tiers ranked between them). Servers can move any permission to a different tier. See the Permission Tiers section for full details.

Defaults vs. customization

Many features (welcome messages, giveaway embeds, AutoMod responses) work out of the box with sensible defaults. Customizing is optional: leave a field blank in the dashboard and the bot uses the default. So you can try things safely, and half-finished setup won't break anything.

Four built-in tiers come with every server: Helper, Mod, Senior and Owner. You can add up to 15 custom tiers of your own on the dashboard's Permission Tiers page, free. The four built-in tiers don't count toward the 15. Tiers are a ladder: anyone on a tier can do everything the tiers below them can.

Custom tiers and permissions

Add a tier like “Lead Mod” or “Trial Helper”, drag it anywhere between the built-in ones, and map Discord roles to it. Every one of the 46 permissions can be pointed at any tier, so you can let Lead Mods ban without making them Senior. Owner always stays on top, and the built-in tiers keep their order.

To keep the bot fast and safe for every server, a few limits are built in. They're the same for everyone and can't be changed in settings.

• Commands: each member can run about 6 commands every 10 seconds, and each server about 120 a minute.
• Buttons and menus: about 8 clicks every 10 seconds per member. Big giveaways are fine: there's no server-wide cap on clicks.
• Dashboard: about 60 changes and 300 page loads a minute per server.
• Scheduled messages repeat at most every 10 minutes. Role and @everyone pings need the Mention @everyone permission.
• Custom commands: up to 200 per server. AutoMod allow/block lists: up to 500 entries.

If you hit a limit, the bot asks you to slow down for a few seconds.

Restricted servers

Servers that try to break the bot, spam past these limits, or attack the dashboard can be restricted. A restricted server keeps moderation and Aurora AutoMod, but loses Mass-Tag, giveaways, custom commands and scheduled messages, can't change dashboard settings, and has Plush+ paused. In the most serious cases the bot leaves the server and won't rejoin.

ZT Plush has 19 commands. Related actions are subcommands, so you type /mod warn or /me remind me instead of one command per action. /music and /soundboard add two more in servers where they're switched on.

Type / and pick a command; Discord shows its subcommands and options as you go. In the tables, [brackets] are optional. “Tier” is the default staff tier, which servers can change. “Plan” says whether it's free or needs Plush+.

Plush+ parts: Mass-Tag, Message Logger and /mod snipe, Overseer Logs, Soundboard, saved playlists, audio filters and 24/7 music, Role Rewards, and the bigger limits or extras for translation, reminders, scheduled messages, welcome rotation, stats channels, birthdays and giveaways. Everything else is free, for every server, including the mod log, AutoMod alerts, the giveaway log and ticket transcripts.

Version 3.8.8 folded dozens of commands into a few. If you're looking for an old one, find it here. Sorted by the new command. Subcommand names are the same unless the row says otherwise.

Good to know

• /roles pick is unchanged. Adding, removing and listing opt-in roles is the /config roles panel now.
• /me language with no choice shows your language and the server's; with a choice it sets yours.
• /fun rps with a member starts the two-player game with buttons. Without one you play against me.
• /games leaderboard asks which game: trivia or typingrace.
• /info bot is the full command reference, the job the bare /info had.
• Some panels have no old twin: /config overview, birthday, tickets, grab, moderation, logs and content.

Old prefix names

The prefix versions keep working: zt!warn, zt!ping, zt!purge and the rest still do what they did, and the new names work too (zt!mod warn, zt!games trivia, zt!me afk, zt!info user). The old staff setup names, like zt!setwelcomechannel, zt!ticketsetup, zt!rrcreate and zt!setrole, now just tell you which /config panel to use.

Right-click a message or a member (long-press on mobile), open Apps, and pick one of ten menus. They work in servers only, and each one follows the same feature switches and cooldowns as the command it matches.

On a message

On a member

If a feature is switched off for your server, its menu says so instead of running. Grab, Download, Translate, Quote and Quarantine follow their feature switches.

Severity-based warnings with auto-escalation. Warnings have three severities: low, medium, high. The bot automatically escalates based on a member's total active warnings.

Ban lengths. A length is one number and one unit, from 1 minute to 1 year: 30m, 12h, 7d, 4w (m minutes, h hours, d days, w weeks), or written out in English like 7days, 7 days and 1 week. Seconds, months, years, decimals (1.5h) and two units (1h30m) are refused with the allowed range, and nobody gets banned. No length means a permanent ban. With the prefix, a length counts only right after the member: zt!ban @user 7d spam. Later in the line, 7d is just part of the reason, and so is anything that isn't a length, like 5x faster or 7dd.

Unbanning. Owner can lift any ban. A Senior can lift a ban they set with the bot, but not one set by someone else or in Discord's own ban list. If the bot was out of the server when a temporary ban ended, it lifts the ban as soon as it is back, as long as that is within 30 days.

Staff can't warn, time out, kick, ban or rename anyone ranked above them on the tier ladder, even if their Discord role sits higher.

Auto-escalation thresholds trigger when total active warnings reach: 3 → automatic timeout, 6 → kick, 8 → ban. Low and medium severity warnings clear when a member leaves; high severity warnings stay across rejoins (a rejoin counts at most 6 of them).

Every row also works with the prefix, like zt!mod warn @user low spam. The older short names such as zt!warn still work too. zt!ban @user 7d spam bans for 7 days; zt!permban @user spam is always permanent.

/auroramod holds the staff tools that work in the background: AutoMod, the quarantine role, the mod log channel, Message Logger and Mass-Tag. It needs Manage Server to show up, and each subcommand also checks its own staff tier. The sections below explain what each one does.

Mod log channel: AutoMod hits, warnings, timeouts, kicks, bans and unbans, purges, slowmode and channel locks, content mode changes and right-click reports all land here, free. Changing the channel needs Manage Server, whether you use /auroramod modlogs, the /config logs panel or the dashboard.

Detects and removes phishing, scam and malware links automatically, in new messages, edited messages and forwarded messages.

What Aurora catches (v3.5.0)

• Known bad sites: a curated list of Discord-targeted scams plus two threat feeds, URLhaus (malware) and the community Discord phishing list, refreshed every 12 hours. Subdomains count too, so login.bad-site.com is caught when bad-site.com is listed.
• Fake Discord, Steam and Roblox sites: lookalike letters (a Cyrillic “і” in “dіscord”), misspellings (dlscord, steamcommunlty) and bait names like discord-nitro-claim.
• Links without https://: discord-nitro.com/claim written as plain text, and “defanged” links like hxxps://bad[.]com.

Test it: post aurora-test.invalid from an account without a staff role. Aurora removes it and logs it, but nobody is warned for it. That address can never be a real site, so it's safe to type anywhere. Staff (Helper and up) skip link scanning by default, so use a member account.

Default escalation

Per member, within 24 hours of the first offense:

• 1st offense: Delete + high-severity warning + DM offender + ping staff
• 2nd offense: Delete + high warning + 1-hour timeout
• 3rd offense: Delete + 24-hour timeout
• 4th+ offense: Delete + ban

Allowlist & Blocklist (v3.1.0)

Server-specific overrides for the bot's threat detection. Configured from the dashboard's AutoMod page:

• Allowlist — domains the bot will never flag, even if a feed or a lookalike check would. Use for GIF hosts, your own domain, legit short links.
• Blocklist — domains the bot will always flag, regardless of any threat feed. Use for scammers targeting your community that aren't in URLhaus yet.

Both lists support leading-position wildcards: *.example.com matches all subdomains (e.g. sub.example.com) but not the bare example.com itself. Up to 500 entries per server. Your allowlist always wins, and your blocklist now beats the built-in safe list, so you can block even big sites if you need to.

Customizable welcome and goodbye messages with variable substitution. Auto-assign a role on join. Set it up with <code>/config welcome</code> or on the dashboard's Welcome page. Banner images and the rotating message pool are dashboard-only.

Welcome Rotation (v3.1.0)

Tired of every new member getting the same greeting? Set up a rotation pool of welcome messages and the bot randomly picks one on each join.

• Free: 10 messages in the pool
• Plush+ 30 messages in the pool

Each message can be individually paused without losing its slot — handy for seasonal greetings. Configure rotation from the dashboard's Welcome page. If the rotation is enabled but the pool is empty (or all messages are paused), the bot falls back to the single welcome message above.

Let members register their birthday (month and day — no year) with the bot. On the day, the bot posts an announcement in your chosen channel, optionally pinging a celebration role.

Set the announcement channel, ping role and message with /config birthday or on the dashboard's Birthday page. A custom embed from the dashboard needs Plush+.

When someone boosts your server, the bot can post a thank-you message. Set the channel and message from the dashboard's Booster page or with /config booster.

Boost-time rewards

Give a role, plus bonus giveaway entries if you like, for boosting a set number of months in a row: 0 months (as soon as they boost), 1, 3, 6, up to 60. Up to 12 rewards per server, free. The clock starts from their current boost, so if they stop boosting the rewards come off and the count restarts. Each reward is handed out once per boost, so a role you remove by hand stays removed.

Stack or replace: keep every reward reached (1, 3 and 6 months together, bonus entries added up), or keep only the latest one.

Boost-count tiers Plush+

Roles for how many boosts someone has put on the server (“3+ boosts”, “5+ boosts”). Discord doesn't tell bots that number, so everyone starts at 1 and you set higher counts by hand under Tracked boosters. Still in testing, with its own stack or replace setting. Free servers keep any saved tiers, but they don't hand out roles.

Two classic Discord channel games, both built in:

• Counting — members take turns counting up: 1, 2, 3, 4… The bot tracks the current number, the all-time high, and (optionally) prevents the same user from counting twice in a row.
• Word-Chain — each new word must start with the last letter of the previous word. The bot enforces the rule, tracks the chain length, and remembers your server's best streak.

Counting rules (server-wide): allow math, so members can type sums like 2+3 or (4*5)-1 and the bot works out the number (it is never run as code); what a wrong number does, either reset to 0 or just a warning; and whether the same person can count twice in a row, which is set per channel.

Word-chain rules (server-wide): check every word against a built-in dictionary (English, Spanish, Portuguese, French or German; Japanese and Korean servers use the English list); turn the last-letter chain rule off for a free-word channel; no repeated words; allow the same person twice in a row; a minimum length from 1 to 10; and an optional time limit per turn from 5 to 600 seconds.

Channels and rules are set with <code>/config counting</code> and <code>/config wordchain</code> or from the dashboard, and they're only for staff. A wrong entry is deleted, and what happens next follows your rules.

Five free games for any channel, with cooldowns so nobody floods a chat. Trivia questions with slurs or swear words are never asked, and a Locked server also filters the Hangman words (see Content Mode). Each game can be paused by the ZT Plush team if something goes wrong. Trivia questions come in English; the other games follow the server's language.

Scores are kept per server, and the dashboard's Games page shows each game's status and the leaderboards. Members can export or delete their own scores with /privacy.

In a Locked server, the word lists for Hangman and Word-Chain leave out swearing and crude words. See Content Mode.

Trivia questions come from the Open Trivia Database, licensed CC BY-SA 4.0.

Quick, free commands for any channel. Nothing is saved. In a Locked server, swearing and crude words are filtered out of what these commands say back (see Content Mode).

Click-to-assign role panels: members react to a message to get a role and un-react to drop it. Supports unicode and custom emojis.

Opt-in roles

Separate from reaction roles: a picker menu members open with /roles pick to give themselves any role you've added. You choose the roles, their labels and descriptions in /config roles, up to 25 at a time.

Roles on a staff tier, and roles with moderator or admin powers, are never handed out by the picker.

Define your own prefix commands, like zt!rules, with custom responses. Optionally restrict who can use them, and add multiple responses to pick from at random.

Members run them with the bot's prefix. A response can use {user}, {server}, {channel} and {membercount}. For random replies, separate the options with | and turn Random on.

Run prize drawings with role-based bonus entries. Start from the dashboard or via slash command. Each server picks how bonuses combine on the Giveaways page: add them up (+2 and +3 roles give 1 + 2 + 3 = 6 entries) or highest only (1 + 3 = 4 entries). The change applies to entries made after you switch.

Who can enter (v3.4.5)

Each giveaway can have its own entry rules. Members who don't qualify get a private note telling them why.

• Required roles: pick one or more. Choose whether members need any one of them or all of them.
• Blocked roles: members with any of these can't enter.
• Time in server: members must have been in the server for at least N days.
• Plush+ Account age: the member's Discord account must be at least N days old. Good for keeping alt accounts out.

Rule limits: up to 10 roles in each list, and day counts up to 3650.

Editing a running giveaway (v3.4.5)

Press Edit on a live giveaway in the dashboard, or use /giveaway edit, to change the prize, end time, number of winners or entry rules. The giveaway message updates straight away. New rules apply to new entries; people who already entered stay in.

Button, pings and winners (v3.4.5)

• Enter button: set your own text, emoji and color server-wide. Plush+ lets you style it per giveaway too.
• Role ping: ping a role when a giveaway starts, ends, or both.
• Winner DM: write your own message. Variables: {prize} {server} {channel} {hostmention} {claimby}.
• Plush+ Claim window: winners get a Claim prize button (in the channel and in their DMs). If they don't press it in time, a new winner is drawn automatically, up to 10 times per giveaway.

Customization (v3.2.0)

From the dashboard's Giveaways page, you can set server-wide defaults for new giveaway embeds — title template, color, footer, thumbnail, plus a rotating pool of up to 4 image URLs (randomly picked per giveaway). Templates support variables:

“Who entered” view (v3.2.0)

On the dashboard's Giveaways page, each giveaway card has a collapsible Show entrants dropdown. Available to anyone with Manage Server permissions. Shows entry weights and win chances, who has left the server, who won, and whether each winner has claimed. You can search it and export it as CSV.

Queue messages to fire at a specific time, or have them recur on a schedule. Each message can broadcast to multiple channels in one shot.

• Free: 10 scheduled messages per server, 1 channel each
• Plush+ 40 scheduled messages per server, up to 5 channels each (broadcast)

Schedule types:

• once — fires at a specific date/time, then disables itself
• hourly, daily, weekly — repeats at that interval starting from when you save
• custom — fires every N minutes (10 minutes to 1 year)

Each scheduled message can be plain text or a custom embed (title, description, color, image, thumbnail, footer). The text goes out exactly as written: variables like <code>{user}</code> aren't filled in. Configure everything from the dashboard's Scheduled Messages page.

Plush+ Advanced logging for your server, set up on the dashboard. Every entry says who did it: a staff member, ZT Plush (AutoMod or Aurora), another bot, or the member themselves.

What gets logged: bans, unbans, kicks, timeouts and warns; deleted, purged and edited messages, with image previews and a transcript file for purges; voice joins, leaves and moves; member joins, leaves, nickname and role changes; channels and roles created, deleted or edited.

Set it up: Dashboard → Overseer Logs. Switch each event on or off, give each group its own channel (or use your mod log), and ignore channels or roles.

Members who used /privacy optout message_log never have their message text logged.

Plush+ Keeps recently deleted messages so staff can bring them back with /mod snipe, and can post each deletion to your mod log.

What's kept: the text, author and attachment links of deleted messages, never bot messages or DMs. For full server logging, see Overseer Logs.

Who changes it: only the Server Owner (not a tier, not Manage Server), because it decides what is kept about members' deleted messages. Switching it on also needs Plush+; switching it off always works.

Plush+ When a member gains a specific role, the bot automatically gives them another role. Use it to set up “promotion chains” — getting promoted to Junior Mod also grants the all-staff color role, for example.

Each rule has a trigger role (what they gain) and a reward role (what gets added). Optionally configure the reward to be removed when the trigger is removed.

Configure from the dashboard's Role Rewards page. Each server can have multiple rules; they all evaluate independently.

Plush+ Send a DM to every member who holds a specific role. Useful for staff announcements, ticket follow-ups, or coordinating events.

Safeguards: up to 5,000 recipients per send, one send per server per day, and a pause of about three quarters of a second between DMs. A message can be up to 1,500 characters. Members can opt out per server with /me masstag optout, and a running send can be cancelled.

Only sending a Mass-Tag needs Plush+. Looking at the history, cancelling a send and the config view work on any server.

A verification gate, free for everyone. New members press a button, solve a colored-button challenge and get the verified role. It helps keep raid bots out.

Pick the channel and the verified role, add rules if you like, then post the message. Change something after posting and the panel offers to post it again.

Auto-updating voice channels that display live server stats. Pick a stat type, pick a channel, customize the name template — the bot keeps the name in sync.

Available stat types: members, online, bots, humans, roles, channels, boosts, bot status.

• Free: 2 stats channels per server
• Plush+ 10 stats channels per server

Configure with /config stats or on the dashboard's Stats Channels page.

11 ticket categories (such as General, Appeal and Report Member) with auto-transcripts, all free. Set up the panel once; members open tickets from a dropdown.

A server soundboard with 35 clip slots, Plush+ only. Clips play in your voice channel; if music is playing it pauses for the clip and picks back up after. Each clip is capped at 30 seconds. Add and manage clips from the dashboard's Soundboard page.

Pick one role that can lock members out of chosen features without touching their Discord permissions. The feature it can block today is Soundboard; the dashboard's Quarantine page lists what is available. Give someone the role to restrict them; take it away and everything comes back.

Staff can also quarantine or release someone from the Quarantine right-click menu on a member.

Plays music in your voice channel. Searches and links start on SoundCloud, and a text search opens a picker with up to 5 results. A YouTube link can fail on some tracks; then the bot looks the same song up on SoundCloud. Playback is free; saved playlists, audio filters and 24/7 mode need Plush+.

* Anyone can stop, leave or clear when every song in the queue is theirs, or when they're the only person listening. Otherwise it needs the permission (Senior+ by default).

ZT Plush speaks English, Español, Português (Brasil), Français, Deutsch, 日本語 and 한국어. Free for every server.

• Server language: used for everything members see (public replies, giveaways, mod logs, welcome defaults). Staff set it with /config language or on the dashboard's server overview.
• Your language: used for replies only you see, in every server. Set it with /me language or the language menu in the top right of the site.
• Slash command descriptions follow your Discord app's language automatically.
• Anything your team writes (welcome messages, custom commands, embeds) is never translated.

Download or delete what the bot keeps about you, in every server, with /privacy or on the dashboard's Your data page. Exports and deletes can be done once every 10 minutes.

Delete keeps what belongs to a server's safety and history: warnings, bans the bot placed (and who set them), AutoMod hits, tickets you opened, Mass-Tags you sent, giveaways you hosted and things you created for a server. Your opt-outs are kept too, so logging and DMs stay off.

When the bot is removed from a server, that server's data is deleted 30 days later unless the bot is added back. Full details are in the Privacy Policy.

Now and then the developer posts news about new features and updates to every server. Staff can turn these off or choose the channel they go to with /config announcements or on the dashboard's server overview.

DeepL-powered translation into every language DeepL supports (30+, searchable). Caps apply per-user, per-server, per day — meaning your usage in one server doesn't eat into your usage in another.

• Free servers: 5 translations per user per day
• Plush+ 20 translations per user per day (4× the cap)

Cached translations don't count — same text + same target language = free repeat. Resets at midnight UTC.

Turn social links into clean embeds or downloaded videos. To save emoji and stickers, see Grab Emoji & Stickers.

Works with: X (Twitter), TikTok, Reddit, Instagram, Bluesky, Threads, Pixiv, Tumblr, Bilibili, Twitch clips and Streamable. YouTube isn't supported.

Limits: 3 downloads per person per minute, one at a time per server. If a video can't be downloaded (private, login needed, or bigger than the server's upload limit), you get the embed link instead. The bot never logs in to any site.

You can also right-click a message and choose Apps → Download media.

/grab turns a custom emoji or sticker into a file you can save. Animated ones come back as GIFs, the rest as PNGs. Staff can then add the file to your server with one button. Free, for everyone.

What you can give it

• One or more custom emoji, picked from the emoji menu.
• An emoji ID, or a link to an emoji on Discord's CDN.
• A message ID (from the channel you're in) or a message link (from this server): you get every custom emoji, sticker and image in that message. Images come back as files only.

You can also right-click a message and choose Apps → Grab emoji/sticker.

What you get back

• Animated emoji as GIF, static emoji and stickers as PNG. Animated stickers are turned into GIFs when possible; if not, you get the PNG with the animation still inside.
• Limits: up to 10 files per use and one use every 3 seconds. Only messages from this server, in channels both you and the bot can read.
• Lottie stickers can't be converted. The same goes for Discord's own built-in stickers. You get a note instead of a file.

Add to server

• Under the files you'll see Add to server (a menu when there are several, plus Add all).
• Only people with Manage Expressions or Create Expressions can use it, and ZT Plush needs Manage Expressions too.
• The server's switch for that kind must be on. Both are on by default; staff change them in /config grab or on the dashboard's overview.
• You confirm the name first. Letters, numbers and underscores only: 2 to 32 characters for emoji, 2 to 30 for stickers, and stickers need a related emoji (🙂 by default). A name that's already taken gets a number.
• Files over Discord's limit (256 KB for emoji, 512 KB for stickers) are shrunk to fit. If the server's emoji or sticker slots are full, you're told how many it has for its boost level.

Names are checked against Content Mode: slurs are never allowed, and swearing only in Unlocked servers.

/me holds the commands that are about you, not the server: your AFK status, reminders, birthday, language and Mass-Tag opt-out. All free, for everyone.

Look things up: the bot, a member, this server, latency and avatars. All free, for everyone.

Slurs are never allowed in an embed. A link to an X or Reddit post in the image, thumbnail or title link follows the media rule in Content Mode: only a flag from the platform makes it sensitive, and no flag, or a failed lookup, counts as normal.

Discord caches command lists per-server. After updates, new commands may not appear immediately. This is normal — here's how to fix it:

1. Refresh your Discord client first: Press Ctrl+R on desktop, force-close and reopen on mobile.
2. Still seeing old commands? Give it up to an hour. Discord spreads command updates out.
3. Server owners: if commands still look doubled or stale, run zt!sync refresh in your server. It clears leftover server-specific copies and tells you how many commands the server now uses. Only the Server Owner can use it, it works once every 5 minutes per server, and it isn't listed in /help. zt!sync reset removes server-specific overrides and goes back to the global commands.

Why this happens: ZT Plush registers commands globally. Discord's CDN propagates updates within about an hour, and Ctrl+R usually gets you there sooner.