Trust Infrastructure

How We Score AI Agents

Three complementary metrics that answer different questions about agent trustworthiness. Continuously updated, on-chain verifiable, portable across protocols.

Risk Score

How risky is this agent RIGHT NOW?

Dynamic, contextual score that changes with every transaction. Combines on-chain history (40%), network behavior (20%), market conditions (25%), and macro context (15%).

0-100

Updated in real-time

Reputation Score

How trustworthy has this agent been OVER TIME?

Cumulative score that grows with clean operation. Clean history (35%), continuity (20%), rule compliance (20%), KYA verification (15%), model quality (10%). Decays with inactivity.

0-100

Portable on-chain

KYA Certificate

WHO is this agent and IS IT VERIFIED?

Know Your Agent protocol. NFT Soulbound on Base Network containing agent identity, creator info, purpose, and audit links. Non-transferable, publicly verifiable.

NFT

Soulbound on Base

Risk Score Scale

Lower is better. Score updates with every evaluated transaction.

0 - 19

Safe. Agent operating within expected parameters. No action needed.

20 - 59

Attention. Some deviations detected. Monitor closely, review rules.

60 - 79

High Risk. Multiple anomalies detected. Consider activating Circuit Breaker.

80 - 100

Critical. Immediate action required. Circuit Breaker auto-trigger candidate.

Reputation Levels

Reputation grows with clean operation and verified identity. Agents with score >= 70 are publicly queryable.

0 - 29

New. Insufficient history. Building track record.

30 - 49

Building. Limited history, showing consistent patterns.

50 - 69

Established. Consistent operation, compliant with rules.

70 - 89

Trusted. Excellent track record. Publicly queryable by other protocols.

90 - 100

Elite. Top-tier agent. Featured status, priority support.

Designed for Every Stakeholder

Different users need different answers from the same trust infrastructure.

Investor

InvestorIvan

"Is it safe to invest with this agent?"

Search agent by address on public Lookup
See Risk Score + Reputation at a glance
View factor breakdown without signing up
Sign up for full ML intelligence + alerts
Export compliance reports for LPs

Developer

DevDan

"Why did my agent's score drop?"

Monitor agents from personal dashboard
Drill into Risk Score factor breakdown
See ML Intelligence: which features deviated
Fix agent configuration based on insights
Watch score improve in next evaluation cycle

Agent Operator

AgentAlex

"I need to prove my agent is trustworthy"

Build Reputation Score through clean operation
Get KYA Certificate (NFT Soulbound)
Share Passport URL with protocols
Reduce onboarding from 4 days to minutes
Gain access to more DeFi integrations

Security Officer

SecSarah

"I need to contain threats in seconds"

Monitor Risk Scores across all agents 24/7
Receive instant alerts on score spikes
Drill into XAI: what exactly happened
One-click Circuit Breaker containment
Immutable on-chain evidence trail