Server Assistant🌟 Main features
Server Assistant is a full moderation suite with optional AI on top. This is the quick tour of the headline features — what each one does, how to switch it on, and a link to the full guide. Everything below is free unless it's tagged Premium.
🛡️ Moderation & AutoMod
Full moderation toolkit Free
Warnings and private staff notes, timeouts, kicks and bans (incl. temp-ban and soft-ban), channel locks, slowmode, role changes and bulk message clean-up — all usable in any channel.
How to use: just run the command, e.g. /warn, /mute, /ban, /purge.
🤔 AutoMod that asks before it over-reacts Free Premium AI review
Clear violations are actioned instantly on every plan; on Premium, borderline messages get an AI confidence check first (the classic "I love visiting Scunthorpe" problem) and are left up for staff with one-tap Delete & warn or Allow — so the innocent aren't punished by a blunt keyword filter.
How to use: /automod → AI review. The setup wizard seeds sensible defaults from your community type.
🧠 AutoMod that learns your team Free
Self-training watches the calls your mods make and proposes filter and rule refinements tuned to your server — you approve or dismiss each one. It never changes anything on its own.
How to use: /automod → Self-train, then review proposals.
🛡️ A tamper-proof audit trail Free
A clean, reliable log of moderation activity that can't be quietly wiped — so you always have the record of who did what.
How to use: pick a log channel during /setup (or change it later in /settings).
🚪 Join-time protection
Four independent layers that keep known-bad actors out at the door. They're separate systems with different data sources — worth understanding so you know exactly what each one checks.
Smart join verification Free
Scores each new member and only gets in the way of the risky ones. Intelligent mode (the default) waves regulars straight through and holds only suspicious joins — brand-new accounts, no avatar, cross-server or alt signals; On verifies everyone. Lower-risk holds self-serve on a quick web page (Discord sign-in → confirm → access); higher-risk ones come to staff as a one-tap Approve / Kick / Ban card. You set the cut-off. Local to your server.
How to use: /settings → Verification (mode, sensitivity, self-serve ceiling).
Anti-raid & lockdown Free
Detects a sudden flood of joins on your server and lets you slam every channel shut in one move while you deal with it.
How to use: anti-raid sensitivity comes from your community-type preset (tune it in /settings); hit /lockdown to lock or unlock everything at once.
🕵️ Alt / ban-evasion detection (altguard) Free
Catches someone you banned coming back under a new account. On join it fingerprints against your own server's ban/kick history (reused avatar, similar name, fresh account) and either flags the account for review or auto-bans a strong match. This is local to your server — it never consults the cross-server network.
How to use: /altguard to enable and set flag-only or auto-ban.
🛡️ Anti-nuke / rogue-admin guard Free
Stops a server nuke or a compromised admin in seconds. When one account suddenly mass-deletes channels/roles, mass-bans, or grabs dangerous permissions, it scores how likely it's an attack and responds on a ladder — observe, alert staff, or auto-quarantine the account (stripping its power, with one-tap revert). It keeps a snapshot of your roles & channels so it can restore them exactly — and recreate deleted channels — afterwards. The owner is always immune, trusted staff can be whitelisted, and it's local to your server.
How to use: on by default — tune it under /settings → Anti-nuke (auto-quarantine, two-person rule, whitelist, sensitivity).
🌐 ThreatNet Free Premium auto-ban
The opposite of altguard: it flags accounts carrying a serious, corroborated safety record on other servers, before they've done anything on yours. It also catches scam images — the bot fingerprints posted pictures and removes ones matching a cross-server scam-image blocklist (on by default, all plans), so a fake "withdrawal received" or giveaway screenshot a text filter would miss gets caught. It shares only severity signals and opaque image fingerprints across the opt-in network — never offence details, the picture itself, or which server acted — and members can opt out in the web portal. Premium can auto-ban the worst matches on join — that auto-ban is silent and opens no in-server appeal ticket, but the decision is disclosed in our Privacy Policy and stays contestable via the web portal.
How to use: /threatnet status to check your own status; /threatnet autoban (Premium) to enable join-time auto-ban. How the data and opt-out work lives on the Privacy page.
Altguard vs ThreatNet at a glance: altguard = “the person I banned is back” (your server’s records, free). ThreatNet = “this person is trouble elsewhere” (cross-server network, opt-out, Premium auto-ban). A ThreatNet alert may also note when an account matches a local known-offender fingerprint, but the two systems stay separate.
📣 Member experience & team ops
📩 Fair ban appeals Free
A ban isn't a silent door-slam. The member is DM'd exactly why, and one reply opens a staff appeal ticket with one-tap Unban, Deny and Research (an AI report on their recent messages — no tokens spent unless you press it). One exception: a ThreatNet auto-ban is silent and opens no appeal ticket — that automated decision is contested via the web portal instead.
How to use: on by default once a log/staff channel is set; tune it in /settings.
🩺 Pulse — whole-server health insight Free
Instead of waiting for a rule to break, Pulse posts a health report to your staff channel — daily, weekly or monthly — surfacing slow-burn trends humans miss, like a trickle of joins sitting just under your raid threshold. Read-only: it never acts on its own.
How to use: /settings → 🔔 Notifications / Pulse to set the cadence (or turn it off).
🎫 Support tickets Free
A button panel members use to open a private ticket channel with your staff; close it and a transcript is saved to your log. Premium adds custom intake questions and adding users to a ticket.
How to use: /tickets setup to post the panel.
Welcome & onboarding Free
Greet new members with a configurable welcome DM and guide them in — part of the flows new members see.
How to use: /onboarding to set the welcome message.
🤖 AI features Premium
The AI suite Premium
SAi — a concierge that reads your server and proposes specific improvements — plus image generation, translation and AI-facilitated mediation between members. A free 150k-token trial lets you try it, or bring your own provider key.
How to use: /sai, /imagine, /translate, /mediate. Set AI up in /setup or /ai-config.
⚙️ Setup & control
Guided setup, Settings hub & autopilot Free
A one-minute guided wizard gets you configured (the bot detects sensible defaults and you just confirm); the Settings hub gives per-feature control afterwards; and autopilot can detect and apply a full config for you.
How to use: /setup to start, /settings to fine-tune, /autopilot to auto-configure.
Three ways to use every command
You’re never forced into one input style:
- ⌨️ Slash commands — anywhere. Type
/in any channel and pick a command. - 💬 Plain text — in staff-chat. Talk to the bot in your staff channel in natural language.
- 🖱️ Right-click menus. Right-click a message or user → Apps for quick actions like Translate, View info, View warnings and Message report.
Where next
Browse any area in the sidebar, or jump to the full command reference. New to the bot? Start with the setup wizard.