Loading...
Loading...
The DRD Score is a dynamic 0-100 trust metric that quantifies an agent's compliance, behavior, and verification status. Use it to make informed trust decisions across your platform.
The DRD Score is computed in real-time using a weighted algorithm that evaluates three core dimensions. Scores update automatically as new data becomes available, typically within 30 seconds of any state change.
Score = (Compliance x 0.4) + (Behavior x 0.35) + (Verification x 0.25)
Each factor is independently scored 0-100, then combined using the weights above.
Each factor is independently evaluated and contributes to the overall score based on its assigned weight.
Adherence to governance policies, audit results, and regulatory framework alignment.
Sub-factors:
Runtime behavior patterns, anomaly detection results, and historical action logs.
Sub-factors:
Identity verification status, credential validity, and trust badge level.
Sub-factors:
Score Calculation Frequency
The DRD Score is recalculated every 5 minutes using a sliding window of the last 30 days of activity. Scores are eventually consistent; a change in behavior may take up to 10 minutes to reflect in the score.
Interpret the DRD Score using these standardized ranges.
Excellent
Fully compliant, verified, and trusted. Eligible for Government badge.
Good
Minor issues, generally trustworthy. Can participate in federations.
Fair
Some compliance gaps or unverified elements. Enhanced monitoring.
Poor
Significant issues requiring attention. Require-approval enforcement.
Critical
Major violations or unverified agent. Federation membership at risk.
Access DRD Scores programmatically via the REST API. All endpoints require a valid API key in the Authorization header.
// Response
{
"agentId": "agent_abc123",
"score": 87,
"factors": {
"compliance": 92,
"behavior": 85,
"verification": 80
},
"lastUpdated": "2026-02-12T10:30:00Z",
"trend": "stable"
}// Query params: ?period=30d&interval=daily
// Response
{
"agentId": "agent_abc123",
"period": "30d",
"interval": "daily",
"history": [
{ "date": "2026-02-12", "score": 87 },
{ "date": "2026-02-11", "score": 85 },
{ "date": "2026-02-10", "score": 88 }
],
"average": 86.7,
"trend": "improving"
}Use the @drd/sdk to access scores directly in your application code.
import { DRD } from '@drd/sdk';
const drd = new DRD({ apiKey: process.env.DRD_API_KEY });
// Get current score
const score = await drd.getScore('agent_abc123');
console.log(score.score); // 87
console.log(score.factors); // { compliance: 92, behavior: 85, verification: 80 }
// Get score history (last 30 days, daily intervals)
const history = await drd.getScoreHistory('agent_abc123', {
period: '30d',
interval: 'daily',
});
// React to score changes
if (score.score < 50) {
await drd.guard('restrict_access', {
reason: 'Low DRD Score',
agentId: 'agent_abc123',
});
}Use the DRD Score in your application logic to make trust-based decisions about agent permissions and oversight levels.
// Use DRD Score in your application logic
const score = await drd.getScore(agentId);
if (score.score >= 80) {
// High-trust agent: allow autonomous actions
await executeAction(action, { requireApproval: false });
} else if (score.score >= 50) {
// Medium-trust: require approval for sensitive actions
if (action.sensitivity === 'high') {
await requestApproval(action);
} else {
await executeAction(action, { requireApproval: false });
}
} else {
// Low-trust: all actions require approval
await requestApproval(action);
await drd.alerts.create({
agentId,
severity: 'high',
message: `Low trust score (${score.score}) - manual review required`,
});
}Subscribe to webhook events to react to score changes in real-time.
| Event Type | Trigger |
|---|---|
| trust.score.changed | Score changes by more than 5 points |
| trust.score.threshold.crossed | Score crosses a range boundary (e.g., good to fair) |
| trust.score.critical | Score drops below 25 (Critical range) |
| trust.score.recovered | Score returns to Good or Excellent after being in Poor/Critical |
{
"event": "agent.score.changed",
"timestamp": "2026-02-12T10:30:00Z",
"data": {
"agentId": "agent_abc123",
"previousScore": 85,
"currentScore": 72,
"change": -13,
"factors": {
"compliance": { "previous": 90, "current": 75 },
"behavior": { "previous": 82, "current": 70 },
"verification": { "previous": 80, "current": 80 }
},
"triggerEvent": "policy.violation.detected"
}
}