Anti-Raid Protection
Raid attacks can destroy a server in minutes: hundreds of bots join, spam messages flood channels, roles are deleted, and members are mass-banned. SYNTHET's anti-raid system detects these attacks in real-time and triggers automatic responses to minimize damage. Customize sensitivity and actions to match your server's activity patterns.
What is a Raid?
A raid is a coordinated attack where multiple bot accounts (or compromised real accounts) join your server nearly simultaneously to:
- Spam:Flood channels with messages, links, mentions, or inappropriate content
- Sabotage:Delete channels, remove permissions, create chaos roles
- Harassment:Target specific members with mention spam or harassment
- Disruption:Trigger channel locks and server-wide notifications
How Anti-Raid Detection Works
Detection Methods
Join Velocity (Primary)
Monitors how many unique users join in a time window. Normal join rate might be 1-5 per minute. A raid shows 20+ joins per minute. System detects this pattern instantly.
Account Age Detection
Newly created Discord accounts ({<1 day old) are common raid vectors. System flags accounts matching this pattern and can quarantine them.
Behavioral Patterns
Multiple new accounts joining from same IP, with identical names/profiles, or performing identical actions suggests coordinated attack.
Message Correlation
If multiple new accounts send identical or near-identical messages, system recognizes spam pattern.
Sensitivity Levels
Configure how aggressive raid detection should be. Higher sensitivity catches more raids but risks false positives:
Low Sensitivity
Trigger: 50+ joins in 10 seconds OR 100+ joins in 1 minute
For: Large servers with 1000+ members joining daily. Very high false negative rate - only catches extreme raids.
Medium Sensitivity (Recommended)
Trigger: 20+ joins in 10 seconds OR 40+ joins in 1 minute
For: Most servers. Balances protection with avoiding false alerts during growth spikes or community events.
High Sensitivity
Trigger: 5+ joins in 10 seconds OR 10+ joins in 1 minute
For: Small servers or those that expect minimal joins. Will trigger on community events or server invitations. False positives likely.
Recommendation: Start with Medium. Monitor for one week. If you get false positives from legitimate growth, increase sensitivity. If raids slip through, decrease sensitivity.
Lockdown Mechanism
When a raid is detected, SYNTHET can automatically enter lockdown mode:
Instant Actions on Detection
- ✓ Block new member joins (members cannot join for X minutes)
- ✓ Create temporary "Unverified" role with no permissions
- ✓ Assign all new joiners to Unverified role (quarantine)
- ✓ Alert staff in designated alert channel with summary
- ✓ Log raid event with details (join count, affected accounts)
Lockdown Duration
Lockdown stays active for configured duration (default: 30 minutes). During lockdown:
- • New members cannot send messages or reactions
- • New members cannot connect to voice channels
- • Existing members function normally
- • Staff can manually lift lockdown via /lockdown end
Account Age Filtering
Suspicious accounts (created within hours/days) are common in raids. Configure minimum account age:
Default: 1 Day
Accounts created within the last 24 hours are flagged as new.
Action: Assign them to Unverified role, require CAPTCHA verification, or auto-ban them during lockdown.
Configuration
Set via Dashboard → Protection → Moderation → Anti-Raid → "Min Account Age". Options: 1 hour, 1 day, 7 days, 30 days, or never restrict.
Manual Raid Response
If you suspect a raid is underway, you can manually trigger lockdown:
/lockdown [duration]
Manually trigger lockdown mode. Useful if you see suspicious activity but system hasn't auto-triggered yet.
Activates lockdown for 30 minutes. You'll receive confirmation and can cancel with /lockdown end.
/lockdown end
Immediately end lockdown and restore normal join permissions. Unverified members are still quarantined until manually released or verified.
/antiraid ban [user-id]
Quickly ban suspicious raid accounts identified by system. Works with user IDs (useful since they might not be loadable).
Whitelist & Exemptions
Prevent false positives for legitimate users:
Whitelist Roles
Users with these roles bypass age filtering and lockdown restrictions:
- • Bot roles (for legitimate bots)
- • Staff roles (moderators)
- • VIP/Patron roles
Protected Members
Individual users can be added to a protected list (e.g., community leaders). During lockdowns, they are not quarantined even if newly joined.
After a Raid: Recovery
Immediate Actions
- 1. /lockdown end (restore normal operations)
- 2. Review quarantined members list
- 3. Ban obvious raid accounts
- 4. Release legitimate new members from quarantine
Cleanup
If channels were spammed or deleted:
- • Use /purge to bulk delete raid messages
- • Use backup feature to restore deleted channels (Premium feature)
- • Review roles and permissions for changes
Post-Raid Analysis
Review the raid event log to understand what happened and adjust settings if needed. Did the system catch it quickly? Were there false positives?
Best Practices
- 1.Start with Medium sensitivity and adjust based on false positives/negatives
- 2.Configure alert channel so staff is notified immediately of raids
- 3.Test /lockdown in testing channel to understand the feature before relying on it
- 4.Whitelist bot roles to prevent your own bots from triggering false alerts
- 5.Have a documented raid response plan for your staff team