Skip to content

Features

Anti-spam

Multi-layered protection — a message goes through several independent checks:

  • AI classification — modern neural networks evaluate the message content in the chat context. The model and providers may change — we always use the best available.
  • External ban lists — cross-check with CAS and LOLS; known spammers are cut off before their first message.
  • Own ban memory — the bot builds its own database of confirmed spammers.
  • Spam on images — text recognition on pictures (OCR) and visual analysis, to catch ads and adult bots that hide text inside images.
  • Common schemes — dedicated detectors for fake job offers, crypto scams, “work from your phone”, adult content and intrusive ads.

Newcomer guard

A new member’s first message is checked more strictly — this stops “Hello here 🤗” adult bots and mass mailings right after joining.

Entry protection & anti-raid

  • 🛡 Newcomer link-sandbox — for users who joined recently (< N h), deletes messages with links, @mentions or forwards (daysandbox-style). Off by default; enable in /settings.
  • CAPTCHA — new members pass a “✅ I’m not a bot” check; the time and the action on failure (kick / ban / mute) are configurable. Enabled in /settings.
  • Anti-raid — when many accounts join at once, the bot sends a single shared message with personal verification buttons instead of flooding.
  • Flood-join protection — an account that joins/leaves too many times within an hour gets blocked in that chat. Always on, no configuration.
  • Global spammer database — collective safety: a spammer banned in several chats receives a global status and is cut off at entry to all connected chats. Optional (enabled in /settings).

Punishments

By default, confirmed spammers are muted (silent isolation). A ban is kept for join-leave spammers and the global ban list. The behaviour can be changed in settings.

Anti-Rus

A core principle — cannot be disabled

Detection and removal of Russian-language messages is always active and cannot be turned off. This is a deliberate stance of the project — to keep the Ukrainian-language space clean 🇺🇦.

Only the presentation is configurable, not the protection itself:

  • Notification mode about deletion (/settings → warnings system): full (with an explanation), short or silent (delete without a message);
  • Optional translation (/settings → Filtering rules → “Translate Russian messages”): if a message is not spam but simply a useful text with Russian words (mixed language, wrong keyboard layout), the bot deletes the original and publishes a Ukrainian translation on behalf of the author — so a useful thought is not lost.

Detection and removal of Russian itself stays enabled under any settings.

Air-raid alerts

Notifications about an air-raid alert for the selected regions. In forum groups the notifications go to the chosen topics — one, several or all.

  • /air_raid <region> — add a region/topic to the notifications;
  • /air_raid all — all known topics;
  • /air_raid off — disable.
  • 🔇 Silence — /silence turns on a manual lockdown anytime (only admins write); silence can also be enabled for the duration of an alert — via the button under the alert or in /settings.

The same can be configured with checkboxes in /settings. Data comes from a Ukraine-wide alerts service.

Summaries — “poyasny”

The poyasny command (a reply to a message with a link) briefly retells an article or a video — handy when sharing a long piece.

Anti-conflict

Detection of insults and toxicity — the bot removes offensive messages and warns the author without turning the chat into a battlefield.

Translation

Translation of messages into the chat language with a single command (/translate, as a reply).

Community tools

  • Welcomes (/setwelcome) — a message for new members with placeholders ({mention}, {first}, {last}, {username}, {id}, {chatname}, {count}, {date}).
  • Rules (/setrules, /rules) — the chat’s public rules.
  • Notes (/save, /get, /notes) — saved replies, media, templates.
  • Custom filters (/filter) — an automatic reply to a trigger word.

Link buttons (URL Buttons)

In welcomes, notes, rules and filters you can add buttons: [Text](buttonurl://t.me/your_channel). Add :same at the end so the next button goes into the same row.

Log channel (/setlog)

Connect a separate channel and the bot mirrors every moderation action there (bans, mutes, warnings, AI deletions). The main chat stays clean, while the channel keeps evidence for appeals. Enable with /setlog, disable with /setlog off.

The channel can be private: add the bot as a channel admin, forward any of its posts to the group and reply /setlog — no numeric id needed (/setlog -100… remains a fallback). One channel can be linked to several groups — each entry is labelled with the group name, so logs don't get mixed up.

Extra

  • Stop-words (/addstop) — your own blacklist of words.
  • Bio / link exceptions — whitelists so “your own” people are not touched.
  • Night mode — restricting activity at night.
  • Reputation & ranks — see the separate section.

Community services

  • 📋 Marketplace — /listing, /listings, /listing_del: ads (housing / jobs / goods / services), emoji categories, 14-day TTL.
  • 🤲 Fundraising — /fundraise, /fund_add, /fundraisers, /fund_close: transparent fundraisers, progress bar, the bot never touches money.
  • 📅 Events with RSVP — /event, /events: Going / Maybe / No buttons + a reminder one hour before.
  • 🛡 Scam check — /checkscam: AI assesses a suspicious message and explains the risk.
  • 💬 Question of the day — /question (+ auto): a daily AI question to spark the chat.
  • 🪙 Mute buyout — a muted user lifts their own restriction for coins (with gamification; admin sets the price).

Captcha modes

Choose: “✅ I'm not a bot” button, math (5+3=?), emoji (tap 🐈), a text code and an image (distorted code against OCR bots). Set in /settings → CAPTCHA; kick or ban on timeout.