Sailing the seas of Discord
Actively in developmentHarborBot turns scattered reports and gut-feeling flags into organized, evidence-backed cases — so one admin or a thin volunteer team can watch over far more members than they could alone. It sits beside your auto-mod bot; it doesn't replace it.
New case opened
A server with a few hundred to a few thousand members and one or two active mods can't realistically track who's been flagged before, whether three reports are the same incident, or whether the evidence still exists by the time someone reviews it. HarborBot gathers the evidence and routes the attention — the judgment stays human.
The reasons pinned messages and manual notes stop working.
The moment a report fires, HarborBot snapshots the last 20 messages in that channel into the case record — before the offending message can be deleted.
Any member — not just mods — can run /report or right-click a message → Apps → Report Message. Lowers the barrier to flagging problems.
Multiple reports on the same user within a short window merge into a single case instead of creating duplicates, with every reporter attached.
See which mods are actively on-call and get flagged the moment coverage lapses — solo and thin teams finally get visibility into gaps.
Reports post to your mod channel by default. A DM escalation fires to one on-call mod only if a case goes unclaimed — no notification fatigue.
New accounts joining in tight clusters with default avatars get a lightweight, rule-based risk tag — pure math, no AI, surfaced for a human to judge.
Slash commands, right-click reporting, no separate app to learn.
/report @user reason:any member — files a report, captures evidence/flag @user reason:mod-only manual flag/case view @userchronological case log for a user/dock set-alert-channelpoint HarborBot at your mod channelA deliberate trust and cost decision, not a missing feature.
Not an auto-mod bot
No word filters, no auto-ban. Pairs alongside MEE6, Wick, or Carl-bot — doesn't replace them.
Not AI-powered
No sentiment analysis, no LLM judgment calls. Flat, predictable cost, no opaque system deciding who gets flagged.
Not a decision-maker
HarborBot surfaces evidence and routes attention. A human mod always takes the actual action.
HarborBot is free today, and free will stay genuinely useful on its own. Premium is in development for teams that want more — nothing to buy yet, just what's planned.
Choose 7, 14, 21, or 30 days per server instead of the free tier's fixed 14-day default.
Shared context on a case instead of every mod re-investigating from scratch.
Make it explicit who's actually handling a case, right in the dashboard.
See who's active and get flagged the moment coverage lapses.
Server-defined strike thresholds instead of one fixed default.
Flag volume trends over time, and pull a case out for appeals or documentation.