Trust Scoring

DRD Score

One number that tells you everything. The DRD Score combines compliance history, behavior patterns, and verification status into a unified 0-100 trust metric for every AI agent.

Start Free Read the Docs

The Formula

Three pillars, one score

The DRD Score is a weighted composite of three independently calculated sub-scores, each capturing a critical dimension of agent trustworthiness.

40%

Compliance History

Tracks policy adherence over time. Every guard call, policy evaluation, and enforcement action contributes to a rolling compliance record that never lies.

35%

Behavior Patterns

ML-driven analysis of agent behavior patterns. Detects anomalies, consistency, and predictability to surface agents that act reliably versus those that drift.

25%

Verification Status

KYC verification, developer identity, capability assessments, and trust badge tier. Verified agents with higher badge tiers earn more trust.

At a Glance

Instant trust assessment

The DRD Score provides an immediate, standardized measure of agent trustworthiness. No ambiguity. No manual review. Just a clear, data-driven number backed by cryptographic audit trails.

90-100

Excellent

Fully compliant, verified, consistent behavior

70-89

Good

Strong compliance with minor deviations

50-69

Fair

Needs improvement in one or more areas

0-49

At Risk

Significant compliance or behavior concerns

Agent: customer-support-v3

Good

Compliance

Behavior

Verification

Capabilities

More than just a number

The DRD Score platform gives you deep visibility into the trust dynamics of every agent in your fleet.

Real-Time Score Updates

Scores recalculate within seconds of any event. Every policy evaluation, enforcement action, or compliance check instantly reflects in the score.

Historical Tracking

Full score history with trend analysis. See how trust has evolved over days, weeks, and months. Identify patterns and predict future behavior.

Industry Benchmarks

Compare agent scores against industry averages. Understand where your agents stand relative to peers in your sector and globally.

Multi-Dimensional Breakdown

Drill into the sub-scores behind every number. Understand exactly which factors are lifting or dragging an agent's overall trust rating.

Score Thresholds

Set minimum score requirements for agent actions. Agents below threshold are automatically restricted until their score improves.

Trend Alerts

Get notified when scores drop below thresholds or show concerning trends. Proactive alerts before problems become incidents.

Getting Started

How the DRD Score works

Register Your Agent

Register your AI agent with DRD and enable scoring. The agent receives a baseline score of 50 and begins accumulating trust data.

Accumulate Trust Data

As your agent operates through DRD guard calls and reports activity, the platform continuously collects compliance data, behavior signals, and verification checkpoints.

Score Evolves in Real Time

The DRD Score recalculates continuously. Good behavior and compliance raise the score. Violations and anomalies lower it. The score becomes a living, trustworthy metric.

Developer Integration

Access scores via the SDK

Retrieve and monitor DRD Scores programmatically with the TypeScript SDK.

score-check.ts

import { DRD } from '@drd/sdk';

const drd = new DRD({ token: 'drd_live_sk_...' });

// Get the current DRD Score for an agent
const score = await drd.score.get('agent_abc123');
console.log(score.overall);    // 87
console.log(score.compliance); // 92
console.log(score.behavior);   // 81
console.log(score.verification); // 88

// Get score history over the last 30 days
const history = await drd.score.history('agent_abc123', {
  period: '30d',
  granularity: 'daily',
});

// Set a minimum score threshold for actions
await drd.policies.create({
  name: 'min-score-requirement',
  rules: [{ condition: 'score < 70', action: 'block' }],
});

Start measuring trust today

Get started free View documentation

Instant trust assessment

The DRD Score provides an immediate, standardized measure of agent trustworthiness. No ambiguity. No manual review. Just a clear, data-driven number backed by cryptographic audit trails.

90-100

Excellent

Fully compliant, verified, consistent behavior

70-89

Good

Strong compliance with minor deviations

50-69

Fair

Needs improvement in one or more areas

0-49

At Risk

Significant compliance or behavior concerns

import { DRD } from '@drd/sdk'; const drd = new DRD({ token: 'drd_live_sk_...' }); // Get the current DRD Score for an agent const score = await drd.score.get('agent_abc123'); console.log(score.overall); // 87 console.log(score.compliance); // 92 console.log(score.behavior); // 81 console.log(score.verification); // 88 // Get score history over the last 30 days const history = await drd.score.history('agent_abc123', { period: '30d', granularity: 'daily', }); // Set a minimum score threshold for actions await drd.policies.create({ name: 'min-score-requirement', rules: [{ condition: 'score < 70', action: 'block' }], });