Verify, Enforce and Prove AI Agent Actions
Pre-action verification • <200ms • Court-admissible
Agents have valid credentials. Can't enforce spending limits. Row caps. Time windows.
Auditors ask: 'Prove this was authorized.' Logs exist. Cryptographic proof doesn't.
Can't add 500ms latency. Can't rewrite entire stack. Need something that works.
Three words. One promise.
Persistent agent identity. W3C DID/VC. Portable across platforms.
Pre-action enforcement. Sub-200ms. Daily caps. Row limits. Currency checks.
Ed25519 signatures. Court-admissible. Immutable audit trail.
Express. FastAPI. NGINX. CloudFront Functions.
import { requirePolicy } from '@aporthq/express-middleware';
// Protect any route with policy enforcement
app.post('/api/payments/refund',
requirePolicy({
policyId: 'finance.payment.refund.v1',
mode: 'simple', // or 'secure' for enterprise
}),
async (req, res) => {
// Your route logic here
// APort has already verified the agent passport
// and enforced policy limits
res.json({ success: true });
}
);Where compliance is mandatory, not optional
Court-admissible attestations. AI contract review. E-discovery. Legal research.
SOX 404 compliance. AI trading. Payments. Fraud detection.
HIPAA compliance. Prior authorization. Claims processing. Patient data access.
First 3-5 companies shape the passport for AI agents.
White-glove onboarding. Grandfathered pricing. Direct founder access.