Smart Moderation

The Complete Moderation Toolkit

Moderation is the backbone of any healthy Telegram community. Without it, spam overwhelms genuine conversation, bad actors disrupt discussions, and valuable members leave for quieter spaces. BotWave addresses this with a comprehensive moderation suite that covers every action an admin might need: permanent bans, temporary bans, kicks, mutes, warnings, promotions, demotions, message purging, and pinning. Each action can be performed silently to avoid drawing attention to the moderation event, or publicly to set an example for the group.

What sets BotWave apart from other moderation bots is the depth of configuration available. Warnings can escalate automatically — after a configurable number of strikes, the bot automatically mutes, kicks, or bans the offender without any admin intervention. The warning limit, the escalation action, and whether warnings persist over time are all adjustable per group. This means a small community can set lenient limits with gentle actions, while a large public group can enforce strict zero-tolerance policies with immediate bans after a single warning.

AI-Powered Content Moderation

The traditional moderation approach — maintaining a list of banned words and hoping members do not find creative ways around them — is fundamentally reactive. By the time you add a new word to the blacklist, the damage is already done. BotWave's AI moderation module, powered by Google's Gemini model, takes a proactive approach. When enabled with /ai_mod on, the bot analyzes every message in real time and evaluates it across four categories: toxicity, scam detection, phishing attempts, and NSFW content.

Each category can be toggled independently, so you can enable scam and phishing detection without necessarily filtering toxicity, or vice versa. When the AI flags a message, it is immediately deleted and the offending user receives a warning via an ephemeral message — a message visible only to them, not to the rest of the group. This private warning approach is less confrontational than a public callout and gives the user an opportunity to correct their behavior without losing face in front of the community. The admin team can review all AI moderation actions through the mini app dashboard, where they can see what was flagged, why, and take further action if needed.

Anti-Raid and Federation Defense

Coordinated raid attacks — where a large number of malicious users join a group simultaneously — can overwhelm even the most attentive admin team. BotWave's anti-raid module detects these patterns automatically. When an unusual surge of joins is detected, the bot can activate raid mode, which restricts new members from posting for a configurable duration. Admins can also manually trigger raid mode with /antiraid on if they receive advance warning of an incoming attack.

For communities that operate across multiple groups, the federation system — called the Shield Network — propagates bans across all linked groups. When you ban a user in one federated group, that ban automatically applies to every other group in the federation. This means a known troublemaker banned in one community cannot simply move to your other groups and start fresh. Federation admins can be promoted and demoted, ban reasons can be standardized, and statistics are available to track the network's effectiveness over time.

Locks, Filters, and Blocklists

Sometimes you need to restrict specific types of content without banning the users who post them. The locks system lets you block entire message categories: links, stickers, GIFs, photos, videos, forwarded messages, voice messages, bot messages, and polls. Each lock can be toggled independently, and lock violations can optionally generate warnings that feed into the escalation system.

For more targeted filtering, the blocklist system lets you define specific words whose appearance triggers an action. You can set the action to delete, warn, mute, kick, or ban, and you can assign custom reasons to individual blocklisted words so that the warning message explains why the content was removed. The smart filters module goes beyond simple keyword matching by detecting obfuscation techniques — like inserting spaces or special characters between letters — that members might use to evade a basic word filter.

Approvals and Trust Scores

Not every member needs to be treated the same way. The approvals system lets admins designate certain users as trusted, exempting them from locks, flood limits, and content filters. Approved members can post links in a no-links group, share media in a no-media group, or send rapid messages without triggering anti-flood. This is useful for verified contributors, long-time members, or partner accounts that need to share content freely.

The trust score system takes this concept further by automatically evaluating every member based on their behavior, warning history, karma, and activity level. Each user gets a numeric trust score that rises with positive engagement and falls with rule violations. Admins can use trust scores to identify potentially problematic members before they cause issues, and the join request manager can automatically approve or decline new join requests based on whether the applicant's trust score meets a configurable threshold.

Available Commands