Trust Scoring System
Letting your chat trigger visual elements and audio sound effects is a great way to boost viewer engagement, but it opens your stream to potential raid-spam and disruptions. Scraplet resolves this with an **automated Trust Scoring engine** that assigns a reputation grade to every viewer in real time.
1. Viewer Trust Brackets
Every incoming chat message is checked against the stream's integration logs. Viewers are categorized into the following hierarchical brackets:
Broadcaster & Admin
Full unthrottled access. Commands and audio clips execute with absolute priority and zero cooldown restrictions.
Channel Moderators
Trusted moderators can trigger moderation tools (e.g. `!clear`), manage lower-thirds on the fly, and run administrative layout triggers.
Subscribers & VIPs
Paying subscribers and VIPs. This level typically grants access to premium sound effects (`!sfx`) and higher-tier text commands.
Active Viewers (Regulars)
Regular chatters who have historical log activity in your stream. Granted basic non-obstructive interaction commands.
New / Unverified Chatters
Users chatting for the first time. Blocked from audio triggers and subject to strict global command rate limits.
2. Restricting Commands & Sounds
You can assign minimum required trust levels to commands inside the **Bot Commands** panel:
When editing or adding a command in the dashboard, navigate to the Access Permissions dropdown. Select the desired minimum group. If a Level 1 user attempts to trigger a Level 3 command, the bot will silently ignore the request or post a friendly permission warning depending on your settings.
Similarly, for SFX (sound commands), you can configure a Minimum Trust Score. This ensures that only trusted subscribers or long-time community regulars can trigger custom sound effects on your live broadcast.
3. Automated Safeguards
Scraplet includes automated active safeguards to protect your overlay from sudden, malicious chats:
- Cooldown Multipliers: If a user triggers commands repeatedly, the system increases their individual cooldown multiplier exponentially.
- TTS Character Gates: Lower-trust users are constrained by a lower maximum text-to-speech character limit (e.g. `40` characters), while subscribers can type up to `200`.
- Raid Shield Mode: In the event of a bot-raid, toggle **Raid Shield** inside your dashboard to instantly upgrade all command thresholds to Level 3 (Subscribers Only) for `10` minutes.