Settings

Overview

The Security Settings page controls the global behavior of the security monitoring system. Here you configure when visitors get flagged for review, when they're automatically banned, and whether security monitoring is active at all.

Security Monitoring Toggle

The Enable security monitoring toggle controls whether the entire system is active:

• Enabled — Risk rules are evaluated for all visitors, trust scores are updated in real time, and visitors can be flagged or auto-banned
• Disabled — No rules are evaluated, no scores are updated, and no visitors are flagged. Existing scores and review statuses remain unchanged but stop updating

Thresholds

Two threshold values control the automated response system:

Review Queue Threshold

Default: 30

Visitors whose trust score drops below this value are automatically added to the review queue for manual inspection. They remain active on your site but are flagged for your team to check.

• A higher threshold (e.g., 50) flags more visitors, giving you more to review but catching more edge cases
• A lower threshold (e.g., 15) flags fewer visitors, reducing noise but potentially missing moderate risks

Auto-ban Threshold

Default: 15

Visitors whose trust score drops below this value are automatically blocked. This is the nuclear option — use it for visitors who are clearly malicious.

• Must always be lower than the review threshold (the system enforces this)
• Set to 0 to effectively disable auto-banning while keeping the review queue active

How Trust Scoring Works

The settings page includes a reference section explaining the scoring model:

1. All visitors start with a trust score of 50
2. Each triggered risk rule reduces the score by its configured weight
3. Positive rules increase the score (they have a negative weight)
4. Scores are updated in real time as events arrive from the SDK

The score is bounded between 0 and 100. A single extreme rule (weight +70) can drop a new visitor from 50 to nearly 0 in one event.