FraudCheckResult

Real-time fraud detection analysis result

blockedboolean

Whether the transaction was blocked (risk_score >= 0.8)

Example: false

risk_scorenumber<double>

Cumulative risk score from all detected signals

Possible values: >= 0 and <= 1

Example: 0.25

risk_levelstring

Risk classification based on score thresholds

Possible values: [low, medium, high, critical]

Example: low

signals object[]

Individual fraud signals detected

Array [

typestring

Signal type identifier

Possible values: [velocity_abuse, velocity_suspicious, impossible_travel, suspicious_travel, bulk_fraud, new_customer, manual_entry, off_hours]

Example: new_customer

severitystring

Signal severity level

Possible values: [low, medium, high, critical]

Example: low

descriptionstring

Human-readable explanation of the signal

Example: First verification for this customer

scorenumber<double>

Contribution to total risk score

Example: 0.1

]

recommendationstring

Human-readable action recommendation

Example: Approve transaction

degradedboolean

True when one or more fraud sub-checks (velocity, IP/device bulk, synthetic-identity) could not run — e.g. a transient DB error — so risk_score reflects fewer signals than usual. These sub-checks fail open rather than blocking the verification, so this is what makes that visible instead of a degraded check looking clean.

Example: false

degraded_checksstring[]

Which sub-checks were skipped, when degraded is true.

Possible values: [velocity, ip_velocity, device_velocity, synthetic_identity]

Example: []

FraudCheckResult

{
  "blocked": false,
  "risk_score": 0.25,
  "risk_level": "low",
  "signals": [
    {
      "type": "new_customer",
      "severity": "low",
      "description": "First verification for this customer",
      "score": 0.1
    }
  ],
  "recommendation": "Approve transaction",
  "degraded": false,
  "degraded_checks": []
}