Most moderation bots give you a panic button and hope you’re awake to press it.
This one decides for itself.
Each guard emits a weighted signal when it catches something — spam 2,
dangerous file 4, channel deletion 8. Signals accumulate over a 60-second
window. When the sum crosses a threshold, the bot raises its own protection
tier and steps back down after five minutes of quiet. The owner gets a DM
explaining which signals caused it.
Tiers: calm → elevated (6 pts) → dangerous (14) → critical (26)
Guards- Anti-spam — rate, duplicates, mass mentions, emoji walls, caps
- Anti-raid — join bursts, young accounts, automatic lockdown
- Anti-nuke — channel/role deletion, mass bans, webhooks, unauthorised bot adds
- File guard — .exe, double extensions, zip bombs (read without extracting)
- Link guard — 36,000+ scam domains, subdomains, bare IPs, lookalike hostnames
- Verification — code-based human check behind a button
- Backups — periodic snapshots of channels and roles, with restore
Edits are scanned too. Post clean text, wait for the mod to scroll past,
then edit the scam link in. Most bots only check new messages.
Alt clusters by name. Ivan001, Ivan002, Ivan_003 each look fine alone.
The bot compares recent joins by name similarity — three matches raise an
alert, four trigger lockdown.
- Put the WOLF-RX1 role at the top of your role list
/setup/status
/help lists every command.
