anti-spam

name	anti-spam
description	Full anti-spam system — new member verification (captcha button), flood control, link blocking, media restriction for new members. Replicates @GroupHelpBot/@MissRose_bot anti-spam. Use when asked for anti-spam, captcha, verification, flood control, or raid protection.
category	moderation
version	1
platforms	["telegram"]
author	adraalabs

Overview

Complete anti-spam suite with 4 layers:

1. New member gate — restrict new members (no media/links) until they click a verify button
2. Flood control — auto-mute users who send too many messages too fast
3. Link blocker — auto-delete messages with links from non-admins
4. Word blacklist — auto-delete messages containing banned words

Trigger: new_member_gate

Event: member_join Condition: { "exclude_admins": true }

Actions:

1. restrict_member user_id:"{user.id}" duration:86400 can_send_media:false can_send_links:false can_send_stickers:false
2. send_message message:"Welcome {user}! Please click the button below to verify you're human." buttons:[{id:"verify_{user.id}", label:"I'm human", url:null}]
3. store_data key:"interaction:verify_{user.id}" value:{"tool":"unmute_member","args":{"user_id":"{user.id}"},"response":"Verified! You now have full access.","unique_per_user":true}
4. store_data key:"_pending_verify:{user.id}" value:"{timestamp}" ttl_seconds:86400

Auto-kick unverified (optional)

Create a cron trigger (every 6 hours) that checks pending verifications:

• list_data prefix:"_pending_verify:" name:"pending"
• For each entry older than 24h, kick_member the user

Trigger: flood_control

Event: message_create Condition: { "exclude_admins": true, "flood_limit": 5, "flood_window": 10 }

Actions:

1. mute_member user_id:"{user.id}" duration:300
2. delete_message
3. send_message message:"{user} has been muted for 5 minutes (flooding)"

Trigger: link_blocker

Event: message_keyword Condition: { "keywords": ["https?://\\S+", "t\\.me/\\S+", "tinyurl\\.com", "bit\\.ly"], "match": "regex", "exclude_admins": true }

Actions:

1. delete_message
2. send_message message:"{user}, links are not allowed in this group."
3. increment_data key:"link_violations:{user.id}" amount:1

Add a check: if violations >= 3, auto-mute for 1 hour: 4. check key:"{incremented_value}" operator:">=" value:"3" else: [{type:"mute_member", user_id:"{user.id}", duration:3600}, {type:"send_message", message:"{user} has been muted for 1 hour (repeated link violations)"}]

Trigger: word_blacklist

Event: message_keyword Condition: { "keywords": [], "match": "contains", "exclude_admins": true }

NOTE: Ask the user what words to blacklist. Add them to the keywords array.

Actions:

1. delete_message
2. send_message message:"Message removed (blacklisted content)."

Commands

/antispam

Tool: send_message Template: "Anti-spam status:\n- New member gate: Active\n- Flood control: 5 msgs/10s\n- Link blocker: Active\n- Blacklist: Active"

Key rules

• Use exclude_admins: true on ALL anti-spam triggers so admins bypass restrictions
• restrict_member is more precise than mute_member — it allows text but blocks media/links
• Flood detection uses a sliding window — flood_limit:5, flood_window:10 means 5+ messages in 10 seconds
• For Telegram: the verify button uses callback_query handled by store_data with interaction: prefix
• Link regex covers http/https, t.me, and common shorteners
• Always tell the user what was blocked and why