Server AssistantServer Assistant Roadmap
Where Server Assistant is heading. Priorities are decided by the people who run servers: we send feature polls straight to every server’s staff chat and tally the votes across the whole fleet, every staff member in every server gets an equal say. Between polls, /feedback or /support reach the dev directly.
v6.0, confirmed by our community, now fully shipped All three community-voted features have landed: Anti-nuke (v6.0.0), Verification (v6.1), and Leveling & XP (v6.2.3). The next round is open for votes, the winners will appear here.
The v6.0 community-voted set has shipped in full. The next round's winners appear here once the current vote closes, cast yours when the poll lands in your staff chat.
CommittedBuilding regardless of votes
Deeper community-type presets
Today, picking a community type in/setup tunes your AutoMod pack, punishment ladder and anti-raid sensitivity. Next: presets that configure more of the server out of the box, sensible per-community feature defaults (e.g. age-gate verification for NSFW, scam-strict link rules for crypto, gentler enforcement for education), with the wizard automatically appending any setup a chosen feature needs (a verified role, gated channels) so everything a preset turns on is genuinely ready to use, never enabled-but-unconfigured. Still opt-in and confirmed in the wizard: a preset never silently switches on a feature you didn't see.Member Portal & streamlined commands
Opening the web portal to every member, not just staff: a self-service home for your own settings, two-way records & forms between you and the team, and privacy opt-outs (including the threat network), all in one place. Alongside it, a leaner set of bot commands: every ability stays listed for discoverability, but the heavy lifting moves to easy in-browser wizards, with each command deep-linking straight to the exact setting or form in the portal. The bot also delivers its privacy notice automatically on install and on demand.In the running, next vote (coming soon)The next community vote picks a whole feature group to build next, every themed group below is in the running. Voting isn't open yet; the poll lands in your staff chat soon.
Engagement & Community 10 cards · 3 new · 1 upgrade
Reasons to show up that aren't moderation: currency, giveaways, events and highlights that reward real participation. Built to share one progression system with the Leveling/XP groundwork rather than bolting on separate economies.
/ai for your community + AI Access panel
A Premium-only command any member can use in opted-in servers: /ai <your question> and Server Assistant replies under its own name with a Claude-powered answer. Default off. Useful for "what time is it in Berlin?", "explain this Discord setting", "draft a server rule for X". Image generation stays on /imagine separately so the free-tier Pollinations fallback isn't lost. Ships with its owner-facing control panel, /settings → AI Access: enable toggle, allowed channels, reply visibility (public/ephemeral), per-user and per-channel rate limits, content strictness, and a query log.Embed builder
Interactive wizard for crafting rich announcements with fields, images, and buttons.Economy & rewards New
Server currency members earn from real participation, coins for messages, voice time, and daily check-in streaks, spent in an owner-defined shop on roles, perks, or custom rewards. Opt-in per server, opt-out per user, with anti-farming caps so spam doesn't pay. Designed to share one progression system with Leveling / XP and Giveaways instead of bolting on three separate economies.Giveaways
Run server giveaways end to end: timed draws with multiple winners, reusable templates, reroll, pause/resume and early end, plus instant drop giveaways where the first to click wins. Role-gated entry keeps Nitro and event drops fair. One of the most-requested engagement tools we don't yet have.Temporary voice channels
A "join to create" hub channel spins up a personal voice room for whoever enters, then tidies it away when they leave. The room owner gets a control panel, rename, set a user limit, lock, and kick, so members run their own spaces without staff carving out permanent channels.Content & social feeds
Auto-post to a channel when something new lands, YouTube uploads, Twitch go-lives, RSS articles, or free-game giveaways, so announcements don't depend on a human catching them. Per-feed destination and message template.Starboard / community highlights New
A board that auto-surfaces the messages your community loves: once a message clears a reaction threshold you set, it's reposted to a highlights channel with a jump link. Per-emoji and per-channel rules, a self-star guard, an ignore list, and an NSFW filter. Celebrates your best moments without anyone curating by hand.Scheduled events & RSVP New
Run game nights, streams and community calls without a second tool. Post an event with an RSVP card (Going / Maybe / Can't), auto-DM reminders before it starts, and sync with Discord's native Scheduled Events. Role-gated signups and an attendee cap keep limited slots fair.Birthday tracker
Opt-in birthdays with auto-DM and channel post on the day.Welcome cards Upgrade
Upgrades the shipped onboarding / welcome flow (welcome DM + welcome reaction emoji) with an auto-generated PNG welcome card for each new joiner, avatar, name and member number, so the greeting is visual, not just text.Onboarding & Membership Lifecycle 4 cards · 1 new · 1 upgrade
The full arc of a member: a smarter welcome on the way in, fair triage for applicants, and a humane, explainable path for dormant members on the way out.
Conversational onboarding Upgrade
Upgrades the shipped auto-onboarding DMs from a static template to a short AI conversation that auto-assigns roles based on the answers and filters bad-faith joiners before their first message.AI application scoring
Extends the staff-reviewed application / join-approval forms with AI triage: every submission gets a quality score against criteria you set, with optional auto-accept / auto-deny thresholds so low-effort applications never reach a human. Add scheduled open/close windows, an accepted-applicant cap, and a cooldown that stops denied users from re-spamming the queue. Builds on the Ticket Panels intake engine already shipped.Fair inactivity cleanup
Prune dormant members without drama. Each member gets a transparent, explainable activity score from signals you choose (messages, reactions, voice, slash commands, threads, event RSVPs), never message content. Set thresholds for a warning, then an inactive role, then optional auto-kick, with exemptions and slower decay for trusted roles. A staff-only ghost mode lets you watch and review before anything member-facing changes, and a/myscore self-check keeps it honest. Same explainable-scoring philosophy as Threat Score.Win-back / re-engagement nudges New
The gentle counterpart to Fair inactivity cleanup. Before anyone dormant is ever pruned, the bot can send a single friendly, opt-out check-in DM, "we've missed you, here's what's new", gated by a quiet period and a hard cap so it never nags. Shares the same explainable activity score, with a staff-only preview of who'd be contacted before a word goes out.Threat & Raid Defense 5 cards · 1 new · 1 upgrade
Proactive security for the attacks the text AutoMod can't see, decoy traps, ghost-pings, and DM-blast raids, sitting on top of the shipped AltGuard and Threat Score.
Bot-on-bot detection
Flag suspicious automation pretending to be human, relevant for mass-marketing raids that evade text filters.Honeypot spam-trap channel
A decoy#honeypot channel that real members are told to ignore, any account that posts in it is removed instantly (kick / softban / ban, your call) and logged, before it can blast the rest of the server. Catches compromised accounts and spam bots that hit every channel on join. A near-zero-false-positive trip-wire (legitimate members never post there) that complements AltGuard and Threat Score.Ghost-ping detection
Flag the "ping then delete" trick, a message that mentions a user or role and vanishes seconds later. The bot logs who pinged whom for staff, closing a common bait-and-harassment tactic that normal AutoMod misses.DM-spam & advertising shield New
Catches the "join, DM-blast every member, leave" attack that text AutoMod never sees. When several members report the same unsolicited DM, or a fresh account mass-mentions identical promo across channels, the bot flags the source, alerts staff, and can auto-quarantine on a confident match, reusing the AltGuard fingerprinting already shipped. Never actions on a single report.ThreatNet in Minecraft Upgrade
Extends the shipped ThreatNet cross-server threat network onto your Minecraft server: known bad actors flagged across the network can be stopped from joining your game server before they cause harm, the same reputation signal that guards your Discord now guarding your world. Builds on the Minecraft ↔ Discord bridge.Adaptive Moderation 6 cards · 2 upgrades
Context-aware, soft-touch enforcement, the right response, not just the harshest. Two cards upgrade the shipped AutoMod engine; the rest are gentler alternatives to deletion.
AI rule explainer
AutoMod DMs the user not just which filter matched, but why this rule exists, in plain language. Cuts the "I have no idea why I got warned" complaint at the root.Persona / role-based moderation Upgrade
Upgrades the shipped AutoMod engine with per-role-tier filter aggressiveness, community norms scale with trust, so verified members get more latitude and unverified get less.Channel-specific tone tuning Upgrade
Upgrades the shipped AutoMod engine with per-channel policy, vent channels soft, family-friendly strict, debate channels allow heat but flag personal attacks.AI thread management
Off-topic conversations get auto-threaded with a soft DM to participants, "moved this fascinating tangent to its own space."Reverse moderation
When a conversation derails, the bot suggests a constructive question or topic shift, soft-touch alternative to message deletion.Misinformation flagging
Fact-checkable claims get a reaction linking to a verified source. Never auto-removes; treats the user as an adult.Moderation Frontiers 3 cards · 3 upgrades
The big, ambitious bets that extend the shipped AutoMod and Threat Score engines into surfaces they don't yet cover, voice, images, and languages beyond English.
Voice-channel moderation Upgrade
Extends the shipped AutoMod + Threat Score engines into voice via real-time transcription, the industry's biggest moderation blind spot, where no major bot operates.Cross-modal moderation Upgrade
Unifies the shipped text AutoMod and link blocklist with image-NSFW and voice analysis into one engine, closing the gap where bad actors flip between modalities to evade single-mode filters.Multi-language moderation NLP Upgrade
Extends the shipped AutoMod, intent detection and natural-language commands, today English-only, to other languages (Spanish, French, German, Polish, Hungarian, Czech…). Exploratory; voteable once a clear cluster of non-English servers asks for a specific language by name. The work is real (per-language prompts, evals, edge-case discovery) and proportional to the languages we choose to support.Server Intelligence 3 cards · 2 upgrades
X-ray vision into your community's health and character. Two cards deepen the shipped Pulse analyst; the third fingerprints what makes your server itself.
Drift detection Upgrade
Upgrades the shipped Pulse analyst from day/week health snapshots to weeks-long tone-and-behaviour tracking, alerting the Owner when material cultural drift sets in, before it erodes invisibly.Community DNA fingerprint
Per-server embedding capturing community character. Powers new-member matching, cross-server compatibility, and rule recommendations.Daily AI channel summaries Upgrade
Extends the shipped Pulse digest with a once-a-day, per-channel content summary posted to a destination you choose, complementing the server-wide health report Pulse already sends.Staff & Team Operations 6 cards · 2 upgrades
Tools for the people running the server, not just the members, scheduling, coaching and shared decisions that build on the shipped retrospectives and approval ladder.
Native-action coaching Baseline admin
Logging of native Discord actions (right-click ban/kick/timeout) shipped in v5.1. What's coming next: when a staff member moderates from Discord's own UI, they get a gentle one-line tip on doing the same thing through Server Assistant (so you keep your escalation ladder, approvals, and undo window intact). Read-only; never blocks the action.Staff insights, coverage map + shared tips
Two staff-intelligence features that ship as one bundle. Staff coverage map, time-zone-aware visualisation of when each mod is typically active; flags coverage gaps for raid-prone hours. Community-of-practice sharing, "servers like yours configure X this way" recommendations based on similar community profiles; privacy-first, patterns only, no per-server identifying data.Predictive mod scheduling
Operational intelligence: "Sunday evenings see 3× your average AutoMod hits, adjust on-call rotation?"Mod-team coaching / bias audit Upgrade
Upgrades the shipped Mod-team retrospectives with per-mod analysis: compares each mod's decision patterns against the team average and surfaces outliers for self-reflection, never public shaming.AI-generated mod training
Auto-generated new-mod onboarding doc based on actual cases your team has handled and house style they've established.Staff-vote moderation Upgrade
Upgrades the shipped approval / escalation ladder with optional consensus moderation, put a kick, ban or timeout to a quick staff vote with a configurable threshold and time limit, so no single mod carries a contentious call alone. Slots into the existing ladder rather than replacing it.Support & Help 2 cards · 2 upgrades
Resolve member questions with less staff load, both cards upgrade shipped features: an AI layer over Ticket Panels, and a knowledge base behind the existing FAQ.
AI ticket deflection & triage Upgrade
An AI layer on top of the shipped Ticket Panels. The bot drafts answers to common tickets from your FAQ and knowledge base, resolving the easy ones, escalating the rest to staff, auto-assigns a priority (low / med / high) from content and sentiment, auto-tags status (waiting on staff / waiting on user), and auto-closes stale tickets after a reminder. Plus a support-analytics view: resolution rate, AI-deflection %, and staff responsiveness.Self-serve knowledge base Upgrade
Upgrades the shipped AI Q&A from your FAQ into a full curated knowledge base that both/faq and tickets draw from, articles staff write once, surfaced wherever a member asks. Powers AI ticket deflection and a searchable member-facing /help. Answers come only from your articles, with a clear "ask staff" fallback when nothing matches, never invented.Server Setup & Structure 2 cards · 1 upgrade
Describe the server you want in a sentence and let the bot build and document it, extending the shipped
/setup server-type templates from moderation profile to full structure.AI server builder & restyle Upgrade
Extends the shipped/setup server-type templates from "pick a moderation profile" to "build the rooms": describe the server you want and the bot creates the whole structure, categories, channels and styled roles, then reshapes it in plain language ("add a VIP role", "rename general to main-hangout"). Includes one-base-colour role palettes and deep-cloning another server's layout onto yours.AI-drafted server policy
Bot reads your channels for two weeks, drafts a rules document tailored to your community's actual norms, posts it for owner approval.Platform & Integrations 3 cards · 1 upgrade
Take Server Assistant's data and controls beyond Discord, deeper browser tooling on the shipped web portal, plus a read-only API for your own builds.
Bring your own AI media credits
Connect your own Pollinations account to power AI media generation,/imagine images today, plus video, text-to-speech, and voice over time, on your own balance, with no shared limits. A one-time approve-in-browser flow links your account; you set your own budget, restrict allowed models, and revoke access any time. Media runs at your pace, independent of the built-in allowance.Web portal, advanced settings & analytics Upgrade
Extends the shipped web portal with the deeper tools that still live in Discord today: full AutoMod editing in the browser, escalation ladders, server-type templates and live previews; threat-score trend charts; and undo windows surfaced right in the portal.REST / webhook API
Read-only endpoints for your guild's bot data (warnings, audit log, activity stats). For external staff dashboards and tooling, the same guild data the web portal already uses, exposed for your own builds.Cross-Server Network 1 card
Opt-in trust shared between communities. The negative side, cross-server abuse signals, has shipped as ThreatNet (see Shipped below); what remains here is the positive, portable-reputation counterpart, held until there's clear demand.
Cross-server reputation
Opt-in positive reputation that follows trusted users between partner servers, the inverse of ThreatNet's cross-server risk signals. Conceptually strong, operationally complex (cross-tenant trust, abuse vectors). Held here until two communities ask for it by name with a use case.Parked & under review 2 cards
Tracked but not on a feature path, a pricing tier awaiting validation, and one idea we'll likely never build because Discord shipped it natively.
Premium Plus ($14/mo) Under consideration
For high-volume servers that push past the 750K/month Standard Premium cap. Includes: 2M tokens/month, priority Concierge response queue, multi-server bundle (3 servers per subscription), and bonus features as they ship. Pending validation from 50+ Standard Premium subs, highest-voted ideas from that cohort will define the final feature set.Polls
Likely not building this. Discord shipped native polls in 2024, which removes the need for a bot-driven/poll in almost every case. Kept here as a reference point, if a real use case emerges (advanced poll types, longer durations, role-restricted voting) it can be promoted back.Shipped, the full history
Voice time earns XP Shipped v6.7.0
Hanging out in voice now counts toward your level. While you're actively in a voice channel you earn XP into the same balance as chatting, and the busier the channel the more it's worth per minute. It rewards being present, sitting muted to listen on a movie or study night still counts, but not idling: you earn nothing while parked in the AFK channel, deafened, or alone in a channel, so levels can't be farmed overnight. Double-XP events and buffs apply in voice too, and like all XP it's free for every server. Full guide →Emote commands & /meme Shipped v6.6.3
Members post a fitting GIF with /emote, /e, /cry, /hug, /dance and more, autocomplete over hundreds of emotes (cry, dance, laugh… plus aliases like lol, ty). Run a bare emote and the bot AI-picks a theme fitting your server + the moment, then pulls a random matching GIF; or set it yourself (/cry anime, /cry wow) or pick a personal default with /emotestyle. /hug @member aims it at someone, and posts read as a natural sentence ("@User hugs @Target"). Plus /meme <search>, drops a matching meme with an emote-style caption for what it says about you (/meme this is fine → "@User has decided everything is fine"): AI-written and in-context where your AI allowance allows, with a curated witty fallback otherwise. The auto-theme and meme captions run on your server's AI allowance including the free trial (not Premium-only) and fall back gracefully when that's used up. Moderation-safe, safe-search on, plus enable toggles, channel allowlist, rate limit, NSFW option and an off-by-default flavor pack in /settings → Emotes. GIFs & memes by KLIPY. Full guide →
Review rewards Shipped v6.5.1
Run/review for a personal code, add it to a review you write on a bot directory (Top.gg, DiscordForge, Discord Bot List, discord.bots.gg), then submit the link, the bot confirms your code is on the live page (not a forgeable screenshot; only your own review carries your code) and grants 500 XP plus 1 month of Double XP. The Double-XP buff is personal, it follows you across every server you share with SA (capped at 2×, and a second review on another platform banks another month). Once per platform. The committed companion (vote rewards) already shipped with the XP system. Full guide →
ThreatNet, cross-server threat network Shipped v6.4.0
Opt-in cross-server protection against users with a serious, corroborated history of scams, raids or ban-evasion on other protected servers. Every protected server shares minimized, severity-only signals, never the offence type, your staff's reasons, message content, or which server acted, so a risky joiner lights up as a free advisory Threat Network flag on their profile (advisory band is free for every server; it never auto-acts). Scam-image defense fingerprints posted images and removes ones matching a cross-server scam-image blocklist. Premium auto-protect (opt-in, off by default) can ban a user automatically on join when they meet a cross-server risk threshold you choose. A core, on-by-default safety feature; individuals can opt out of profiling in the web portal, subject to a published safety exception. Go-live v5.8; scam-image defense v6.3.0; Premium auto-protect v6.4.0. Full guide →Setup & settings wizard overhaul Shipped v6.3
A top-to-bottom refresh of the setup and/settings wizards: /settings reorganised into four clear sections (Settings · Members · Notifications · Security, v6.3.1), a faster guided /setup that pre-fills your current config rather than starting blank, plain-language permission review, and a full pass over which features are standard out of the box versus opt-in, so a fresh server is well-protected by default without being buried in choices.
Leveling & XP Shipped v6.2.3
Server XP earned from real participation, with level-up announcements you control (/leveling), Double-XP events, anti-spam and per-user opt-out. XP doubles as a trust signal and is spent in Crestbound, the Crest-collecting game in the web portal. The third of the community-voted v6.0 features. Full guide →
Crestbound, the XP collecting game Shipped v6.2.3
The XP your members earn in Discord becomes a spendable currency in Crestbound, the Crest-collecting game in the web portal (under XP Games). Spend XP on booster packs to collect the Crests of the first Chapter, Embers of the First War, across four rarities; the first player anywhere to pull a brand-new Crest illuminates its artwork for everyone and keeps a permanent discovery credit (the First Light achievement). Burn spare duplicates back into XP. No purchase and no Premium, if you've earned XP, you can play. A Wardeck battle mode is on the way. Full guide →Verification & join-approval Shipped v6.1
Risk-scored joiners are held in a no-access pending state and wave the safe ones straight through; higher-risk joins route to staff as a one-tap Approve / Kick / Ban review with a full report, and the member is DM'd the outcome with a portal appeal form. A one-tap Auto-configure channels helper (v6.3.2) wires the hidden/visible channels for you. The second of the community-voted v6.0 features.Anti-nuke / rogue-admin guard Shipped v6.0.0
A smoke alarm for a server nuke or a compromised admin: it watches for catastrophic bursts (mass channel/role deletes, mass bans/kicks, permission grabs) and responds on a ladder, observe, alert staff, or auto-quarantine the account (owner-immune, one-tap revert). Keeps a snapshot so it can restore your roles & channels exactly and recreate deleted channels, with a trusted-actor whitelist and a two-person rule. Local-only; the first of the community-voted v6.0 features. Full guide →Reaction roles Premium Shipped v5.6.27
Self-service role buttons built from one/rolepanel wizard, pick roles from a dropdown, name the panel, post it. Members click to add/remove a role, and a My roles button opens a private menu showing a ✓ next to the roles they already have. The bot only offers roles it can actually assign.
Custom slash commands Premium Shipped v5.6.27
Owners define their own real/command (text or embed) from the /customcmd wizard. Embed commands have an optional Live toggle, editing the embed updates every copy already posted, turning /rules, /announcement, etc. into living documents.
AI Q&A from your FAQ Premium Shipped v5.6.27
Members run/faq with a question and the bot answers only from your curated entries, when the FAQ doesn't cover it, it says so and points them to staff rather than inventing a policy. Staff curate from the /faq management wizard.
Translation Premium Shipped v5.6.27
Right-click any message → Translate, or use/translate, into your Discord language or one you name. On-demand, so there's no per-message cost or channel noise.
Mediator Premium Shipped v5.6.27
The/mediate wizard has the bot DM two members, collect each side privately, and post a neutral summary to staff, a chance to de-escalate before anyone reaches for a moderation action. The bot never takes sides or recommends punishment.
Server backup & restore Premium Shipped v5.6.27
The/backup wizard snapshots your roles, categories, channels and their exact permission overwrites. Restore is owner-only + confirmation-gated and strictly additive, it recreates anything missing and reapplies overwrites, never deletes, safe insurance after a nuke or a botched reorg.
White-label branding Premium Shipped v5.6.27
Rebrand the bot for your server from the/brand wizard: a custom per-server nickname, your name + icon on the bot's embeds, your colour scheme, and your name in place of "Server Assistant" in footers. (Discord allows one global bot identity, so this rebrands the nickname + the bot's output, not the account itself.)
Decision explainer + guided appeals Premium Shipped v5.6.26
Two premium upgrades to the free one-reply appeal flow. "Why was I actioned?", warned/muted members get a DM button (or can DM the bot) for a calm, plain-language AI explanation of the staff-supplied reason. Guided multi-round appeals, a banned member's appeal becomes a short AI-facilitated back-and-forth (the bot asks a neutral follow-up or two), then staff get a concise summary + full transcript on the appeal ticket and the portal "Needs input" card. Bounded by a max-rounds setting; the facilitator never decides or promises an outcome, staff make every call. Free servers keep the one-reply flow unchanged.Web portal for server administration Shipped v5.6.25
Your whole staff team manages the server from the browser at serverassistant.wandweb.co (or run
/portal):
- Discord OAuth sign-in, free for every server; owners, admins & moderators each get role-mapped access
- Insights dashboard, members, online & voice counts, moderation & AI-usage trends, a paginated activity log, plus your Daily Pulse and weekly mod recap
- Cases & moderation from the browser, warnings, ban appeals, tickets and a full audit trail
- Act-from-web decisions, resolve ban appeals, AutoMod reviews, owner-approvals and raid lock-downs from a "Needs input" queue, kept in sync with Discord
- Install it as an app on phone or desktop and turn on push notifications
- Same guardrails as in Discord, approvals and full audit logging
Permission health check Shipped v5.6.1
After setup, an admin can edit the bot's role or a channel's permissions and silently break moderation. This periodically verifies Server Assistant still has what it needs, manage roles, post in your log channel, read your staff chat, and DMs the owner (or posts a quiet staff-chat note) the moment something drifts, with exactly what broke and how to fix it. No spam; it only speaks up when something's actually wrong.Server-type templates in /setup Shipped v5.6.0
A picker step in the setup wizard (and the autopilot Apply DM) that asks "what kind of community is this?" and applies a moderation profile tuned to the answer. Categories picked from the actual dominant Discord community types, not generic labels:
Gaming · Creator / Streamer · Anime / Fandom · Crypto / Trading · Education / Study · NSFW / 18+ · Set it up myself
Each template = a tuned combo of AutoMod preset, punishment ladder, anti-raid threshold, Pulse default, and Threat Score mode. Crypto, for example, ships strict anti-scam and link filters by default; Education uses a soft-action ladder (warnings over bans); Creator servers get tighter anti-raid because rival fan bases are a real attack vector; NSFW gets age-verification emphasis. The underlying presets already exist, this surfaces them as a 5-second self-classification instead of leaving owners with a one-size-fits-all default.
/whatsnew command + post-update DM Shipped v5.6.0
A slash command any staff member can run to see what's changed since they last looked, plus a one-time DM to the owner after each update summarising the user-visible changes that affect their server (new commands, default changes, panel additions). Replaces "why did this start happening?" support questions before they get sent.
Smart purge Shipped v5.6.0
purge ai 50 deletes the genuinely problematic messages and archives constructive ones to a thread, instead of nuke-everything.
Mod team retrospectives Shipped v5.6.0
Weekly auto-generated retro embed: case counts, resolution times, notable patterns, items worth discussing. A weekly artifact your mod team will actually want to share with each other and with the owner./image and /img, aliases for /imagine Shipped v5.6.0
Quality-of-life shortcuts. The full /imagine command stays as the canonical name; the aliases are there because muscle memory is real and everyone types one of those two by default.
Co-owner role mapping in /settings Shipped v5.6.0
For servers with multiple owners, business partners, co-founders, or anyone the registered Discord owner wants to grant Owner-tier capabilities to. /settings → Roles → Add a co-owner role picks a Discord role and grants it the same skip-approval-on-dangerous-actions permission the Discord-native owner has.Follow-up to the v5.6 setup-wizard simplification: the Owner role dropdown was removed from
/setup because most servers don't need it (the Discord-native owner is implicit Owner-tier). This is the surface for the minority who do, surfacing it as a deliberate "I want a co-owner" choice rather than an unclear option during initial setup.
Gentler setup reminders Shipped v5.5.14
Servers that add the bot but don't finish/setup now get a short, finite reminder sequence, a friendly nudge around day 2, day 7, and day 14, then it stops for good (the last one says so). Previously these reminders ran weekly with no end. Spaced out, never bunched; owner DM first with a system-channel fallback; and the moment you run /setup, reminders stop. Only touches servers that haven't completed setup, nothing about moderation changes.
/setup rebuilt end-to-end + safer defaults Shipped v5.5.13
The whole setup experience, rebuilt from the moment the bot joins to the final Finish. Pre-filled detection: your staff chat, log channel, and Admin/Moderator roles are detected and filled in, confirm a clean detection in a single click. Guided and can't-get-stuck: each step stays locked until it's actually valid, and one button creates any missing channels or roles for you (private staff channels, sensibly-permissioned roles). Pre-flight permission check: if the bot was added without a permission it needs to moderate, setup tells you up front, with a one-click re-invite, instead of letting you finish a setup that silently can't work. Plain-language permission review: if an existing role or channel needs a change to be safe (a Moderator role that can manage the whole server, a staff channel visible to everyone), the bot shows exactly what it wants to change and waits for your approval, it never silently strips or weakens permissions you set on purpose. When the bot can't fix something itself (a staff role sitting above it in the list), it gives exact step-by-step instructions rather than failing silently. Safe re-runs: running /setup again pre-fills your current setup and never wipes settings you've tuned elsewhere. Safer defaults: Dev Updates ship on for new servers (existing servers are unaffected, your choice is kept), and the welcome can never fail silently, if the full wizard can't load instantly, a short text welcome points you at /setup.
UX hotfix sweep, privacy defaults, in-place wizards, menu correctness Shipped v5.5.11
A focused pass on places where the bot's menus didn't match its behaviour. Fixed: Pulse panels showing "Disabled" while Pulse was actually firing (three UI sites now read the same source the digest does);/help → ← Overview button doing nothing on click (now correctly disabled on the Overview screen); /autopilot Apply silently swallowing errors. Defaults moved to opt-in: Pulse and the weekly digest are now off by default, no more surprise daily messages; autopilot's one-tap install no longer silently enables Pulse. Improved: /setup wizard is now a single ephemeral that edits in place (no more viewport-jumping ack messages), with a live "Selected so far" summary; welcome embed for new installs no longer references the closed beta perk. Existing servers that explicitly configured anything keep their choices, only the never-touched defaults change.
One-tap setup on install Shipped v5.5.10
As soon as you add Server Assistant, the bot scans your server, detects sensible defaults (staff chat, log channel, Owner / Admin / Mod roles), and DMs the owner with a one-button "Apply" proposal. Tap once and your server is fully configured. Fallback to your system channel with an owner @-ping if DMs are closed, so nobody misses the setup prompt because of a privacy setting./setup and /autopilot still work for manual flows.
Per-server command usage logging Shipped v5.5.9
Every slash command, right-click context menu, and natural-language text command appears in your server's log channel as a compact entry: user · command · channel · invocation type. Owners and staff see exactly what's being used and by whom, in the same channel they already check. Fire-and-forget, never blocks the command itself. Toggle off in/settings → Behaviour if your server is chatty.
/concierge renamed to /sai + admin read-only Shipped v5.5.7
Same Server Assistant intelligence chat, shorter name. Owners get full access (read insights + propose config changes); admins now get read-only access, they can ask questions and get insights but can't enroll proposed changes. Embed shows the caller's role tag so it's obvious which mode the answer came from.
Multi-tenant address detection + per-feature NL billing Shipped v5.5.6
The bot now recognises its actual per-server nickname when staff write "Pepper, mute @user" or "Watchdog: ban that user", previously it only matched a hardcoded set of names. Also: every AI call is now properly tagged in your token usage breakdown (nl_intent, nl_chat, nl_extract, report_command, self_trained) so you see exactly which features ate which tokens in /premium.
Refreshed staff onboarding + Dev Updates default ON Shipped v5.5.5
The welcome message posted after/setup is now a 4-embed sequence (Text · Slash · Right-click · Decision cards) and @mentions every configured staff role so your team actually sees it. Slash command list grouped by Core moderation / Raid response / Config / Personal with all v5.5.x additions. Dev Updates toggle starts enabled in setup, owners almost always want release notes for the bot powering their server.
Docs reconciliation + Premium /imagine cooldown removed Shipped v5.5.4
Premium servers now get unlimited /imagine, the 30-second per-guild cooldown is gone, matching what the pricing page has always said. Coordinated sweep across docs, pricing, privacy policy, and listing copy to make sure every page reflects what actually shipped in v5.3 to v5.5.
Ban reasons & one-reply appeals Shipped v5.5
Shipped 2026-06-03. Every ban, manual,/tempban, AutoMod, or native, now DMs the user the reason and invites one reply to appeal. That reply opens a staff appeal ticket with Unban / Deny / Research buttons; Research runs Message Report (the renamed right-click AI summary) on the user's last message so you can judge context first. Soft-bans excluded; on by default.
AutoMod AI second-opinion + repeat-offender detection Shipped v5.4
Shipped 2026-06-03. Two safety upgrades. AutoMod AI second-opinion (Premium, opt-in via/automod → AI Review) scores borderline filter hits and routes low-confidence catches to a one-tap staff review card, Delete & warn / Allow, instead of auto-deleting, cutting false positives. Repeat-offender detection (free, opt-in via /altguard) fingerprints banned/kicked users and, on a confident match, auto-bans likely ban-evading alts on join while alerting staff.
Baseline admin: /role · /lockdown · /tempban Baseline adminShipped v5.3
Shipped 2026-06-03. Three plug-and-play admin essentials, all permission-gated and written to your audit trail. /role add|remove changes a member's roles in one command, with guards so staff can't assign a role at or above their own. /lockdown start|end locks every channel at once for raid response and, on end, restores each channel's exact prior permissions, not a best guess. /tempban @user 7d bans for a set time and un-bans automatically, with a timer that survives restarts.
Ticket Panels Shipped v5.2
Shipped 2026-06-03. In-server private support tickets, no DMs, no messy threads. Admins run/tickets setup to post a panel with a Create Ticket button; members open a private ticket-NNNN channel via a short intake form, and a full .txt transcript is saved on close. Free: one-question intake, private channels, transcripts. Premium: up to 4 custom questions, custom welcome/close messages, and adding users mid-ticket.
Threat Score v1 (advisory) shipped v3.8
Risk-based moderation that replaces fixed punishment ladders. Actions fit the actual threat, context-aware, opt-in, with extensive safeguards against punishing legitimate users. v1 (advisory) shipped 2026-05-29, free for every server, each user gets an explainable, time-decayed risk score shown to staff in their profile. It never auto-actions; it's pure visibility. The active, score-driven phase (Premium, with the safeguards below) comes only after the advisory version proves accurate in real servers.Safeguards against false positives
- Sentiment + content paired with velocity (eager new members score down, not up)
- First-week grace period for genuinely new users
- Moderate scores trigger a friendly check-in DM, not an action
- Score decay over sustained good behaviour
- No silent kicks or bans, high-score events ping staff for review
- Full transparency: every action shows its score breakdown to user + staff
- Server-defined thresholds; fallback to fixed-ladder mode supported
Bot health insurance Shipped v3.7
Bot watches its own impact in each server and proactively suggests tuning when it's becoming too noisy. Stops Server Assistant from quietly becoming the bot every server eventually mutes. Shipped 2026-05-29, weekly self-check on AutoMod footprint, gentle tuning suggestion to staff chat, read-only, ≤once/14 days.Self-trained AutoMod Shipped v3.6
Bot watches your staff's manual decisions (deletes + reported messages), clusters the patterns, then proposes server-specific AutoMod rules for owner approval. v1 shipped 2026-05-28, opt-in observation, on-demand AI rule suggestions, accept/skip review. Nothing is ever auto-enforced.Pulse Shipped v3.5
Always-on AI server-health analyst. Daily digests, leading-indicator flags, channel-tone shifts. Never auto-actions, just gives staff x-ray vision. v1 shipped 2026-05-28, 24-hour summary, week-over-week deltas, two leading-indicator flags (AutoMod-rate spike, sub-raid join spike). Channel-tone analysis lands in a later iteration.Previous votes, closed poll results
2026-06-20, the v6.0 feature round
“Which features should Server Assistant build first? Vote for all you'd use!”
Verification gate + application forms13% (10)
Leveling / XP13% (10)
Anti-nuke / rogue-admin guard11% (9)
/ai for your community + AI Access panel10% (8)
Embed builder10% (8)
AI rule explainer9% (7)
Native-action coaching9% (7)
Staff insights, coverage map + shared tips9% (7)
Bot-on-bot detection9% (7)
Bring your own AI media credits8% (6)
How to influence the roadmap
The fastest way to move something up the list:
- Vote in the feature polls that arrive in your server’s staff chat, every staff member in every server gets an equal vote. Cards physically climb into the gold band as votes land, and when the poll closes, the top 3 together become the next release.
- Send
/feedbackor open a/supportticket describing the use case, not just “add feature X” but what you’d do with it. Strong cases pull ideas out of the backlog packs and into a vote round.
What ships is what gets requested most clearly. Vague “add more features” feedback is unactionable; “I run a 2,000-member RP server and we need X because Y” gets prioritised.