Skip to main content
How It Works

Not a Blacklist. An Intelligence Layer.

Confidence-based scoring instead of binary judgments. Every decision is explainable, every listing is temporary.

Transparent
Auto-Decay
Free Remediation

Scoring Engine

Real-time analysis

Live
1

Signal Ingestion

Honeypots, partners, community

+15
2

Score Calculation

Weight × Diversity × Recency

72
3

Publication Decision

Score ≥ 30 → Published

In Feed
203.0.113.42 72/100
Clean Observed Published High Risk
The Process

Three Simple Steps

From signal collection to automatic recovery — here's how Fraudcache protects without over-blocking.

1

Signals Collected

We aggregate abuse signals from multiple independent sources. Each signal is weighted by source reliability and recency.

Honeypots Partners Community
2

Confidence Scored

Our engine calculates a 0-100 confidence score based on signal weight, source diversity, and recency.

Example Score 72/100
Clean Observed Published High Risk
3

Auto-Decay & Remediation

Scores decay automatically over time. If abuse stops, the IP is delisted. Anyone can submit a dispute.

Recovery is the goal

No permanent punishment

Scoring System

Understanding Confidence Scores

Our 0-100 scoring system provides nuanced risk assessment instead of binary all-or-nothing decisions.

0
0-14

Ignored

Insufficient evidence to act

15
15-29

Observed

Tracked but not published

30
30-59

Published

Included in public feeds

60
60-100

High Risk

Verified threat, high confidence

What Affects the Score?

Four key factors determine an IP's confidence score

Signal Weight

Own observations (honeypots) weigh heavier than external feeds

Source Diversity

Multiple independent sources increase confidence significantly

Recency

Recent signals weigh more heavily than older observations

Time Decay

Scores automatically decrease over time if no new abuse signals appear

Data Sources

Multi-Source Intelligence

We combine multiple data sources with varying trust levels to ensure accuracy and reduce false positives.

Highest Trust 30+

Honeypots

Own infrastructure observing real attacks in real-time

High Trust 20+

Partner Reports

Verified partners submitting abuse data

Standard Trust 8+

Community

User-submitted reports with supporting evidence

Automatic Recovery

Default to Rehabilitation

Unlike traditional blocklists that require manual delisting, Fraudcache automatically reduces scores over time. If abuse stops, rehabilitation happens naturally.

Time-based decay

Scores reduce 5-10% daily without new signals

Dispute process

Submit evidence for faster resolution (24h review)

No pay-to-delist

Remediation is always free, forever

Score Decay Timeline

Example: IP with no new abuse

Day 0
85
Day 3
70
Day 7
50
Day 14
25
Day 21
10

Delisted at score < 30

Ready to See It in Action?

Check any IP address to see how Fraudcache scores it, or download our feeds to protect your infrastructure.