Rules

View, customize, and tune the built-in risk detection rules that determine visitor trust scores.

Overview

Yaplet ships with 23 built-in security rules organized into five categories. Each rule detects a specific risk signal and carries a configurable weight that determines how much it affects a visitor's trust score. The Rules page lets you tune these weights to match your site's risk tolerance.

Rule Categories

Account Takeover

Rules that detect signs of session hijacking or credential misuse:

RuleDefault WeightWhat It Detects
Country changed mid-sessionHigh (+20)Visitor's country changed within 30 minutes
Device changed mid-sessionMedium (+10)Browser/device changed within 30 minutes
IP address changed mid-sessionMedium (+10)IP address changed within 30 minutes

Behavior

Rules that analyze visitor activity patterns:

RuleDefault WeightWhat It Detects
Multiple countriesMedium (+10)Accessed from 3+ different countries
Numerous IP addressesMedium (+10)Used 10+ distinct IP addresses
Multiple devicesMedium (+10)Used 3+ distinct device/browser combinations
High rage clicksMedium (+10)5+ rage clicks detected
High console errorsMedium (+10)10+ console errors triggered
High network errorsMedium (+10)10+ network request errors
Night-time activityMedium (+10)Activity between midnight and 5 AM UTC
Very short sessionsMedium (+10)3+ sessions with only a single event each
Dormant visitor (30 days)Medium (+10)Inactive for 30+ days, now active
Dormant visitor (90 days)High (+20)Inactive for 90+ days, now active
Brand new visitorPositive (-20)Account created today — reduces risk

Bot Detection

Rules that identify automated or scripted access:

RuleDefault WeightWhat It Detects
Suspicious user agentExtreme (+70)Known bot, crawler, or automated tool signature in the user agent string
Suspicious URL patternHigh (+20)Visits to common attack paths (e.g., /wp-admin, .env, phpinfo)

IP Analysis

Rules based on IP reputation:

RuleDefault WeightWhat It Detects
Known bad IP rangeExtreme (+70)IP belongs to a known malicious range
TOR exit nodeExtreme (+70)IP is detected as a TOR exit node
VPN or proxy networkHigh (+20)IP is flagged as a VPN or proxy
Hosting/datacenter ASNMedium (+10)IP belongs to hosting or datacenter infrastructure

Blacklist

Rules for organizational bans:

RuleDefault WeightWhat It Detects
Blacklisted visitorExtreme (+70)Visitor was manually blacklisted by your team

Customizing Rule Weights

Each rule's weight determines how much it contributes to lowering a visitor's trust score when triggered. You can adjust any rule to one of these severity levels:

LevelWeightEffect
Disabled0Rule is ignored — never triggers
Positive-20Increases trust score (used for good signals)
Medium+10Moderate risk reduction
High+20Significant risk reduction
Extreme+70Severe risk reduction — a single trigger can push a visitor into the danger zone

Saving Changes

When you modify a rule weight, a save bar appears at the top of the page showing you have unsaved changes. Click Save changes to apply, or Reset to discard.

Weight changes apply going forward. Existing visitor scores are not retroactively recalculated when you change a rule's weight.

Practical Tips

  • Disable "Night-time activity" if your audience is global — UTC-based night hours will flag legitimate users in other time zones
  • Reduce "Multiple countries" weight if your users frequently travel or use VPNs
  • Keep bot detection rules at Extreme — legitimate visitors almost never trigger these
  • Leave "Brand new visitor" at Positive — new visitors should get a trust boost, not a penalty