Your AI agent just closed a $50k deal. Did your client know it was authorized to? Entruva gives every agent a cryptographically signed credential, a live reputation score, and a public trust card — so enterprises say yes instead of maybe.
Enterprise asks "How do I know your agent won't go rogue?" You have no answer.
Agent completes a task. Client doesn't know if it was authorized to do that.
Agent marketplace needs to vet 500 agents. No standard. No audit trail.
Regulator asks for proof your agent was compliant on a specific date.
Every agent gets a signed credential with its capabilities and authority scope.
Every action can be verified against the credential in real time via API.
Reputation score built from real interaction outcomes — not self-reported.
Full audit trail: who verified the agent, when, and what happened after.
How it works
Create your org's Principal DID, register your agent class with its capability scope, and issue a signed verifiable credential. Takes 5 minutes via dashboard or API.
Every credential gets a public verification page at entruva.ai/verify/{id}. Send the link to any client or embed the trust badge in your product. They see scope, reputation score, and live status.
Log interaction outcomes after each task via the API. Reputation score updates in real time. Set auto-revoke thresholds. If an agent violates scope, kill it everywhere with one call.
The protocol
Built on open standards — W3C DIDs, Verifiable Credentials, and signed JWTs. No vendor lock-in. Any system can verify without an SDK.
DID-based identity for orgs, agent classes, and individual agent instances
Bayesian AgentRep scoring from real interaction outcomes and verification signals
On-chain liability coverage priced to capability scope and live reputation score
EU AI Act, SOC 2, and automated regulatory attestation for enterprise deployment
Who uses Entruva
Developer-first
Any system can verify an Entruva credential before letting an agent act. No SDK required — a single HTTP GET returns scope, reputation, and live status. Every verification is logged, timestamped, and attributed to your org.
Get your free API key →curl https://app.entruva.ai/api/credentials/{id}/status
# Response:
{
"valid": true,
"status": "ACTIVE",
"agentClass": {
"name": "Sales Agent v2",
"capabilities": ["FINANCE", "CRM"],
"jurisdictions": ["US", "EU"]
},
"reputation": {
"composite": 0.87,
"tier": "TRUSTED",
"outcomes": 1423
}
}Pricing
14-day free trial on all paid plans. No credit card required to start.
FAQ
AI agent identity is a cryptographically verifiable proof of who built an agent, what it is authorized to do, and that it has not been tampered with. Without it, enterprises have no way to know if an agent acting on their systems is authorized — creating security, liability, and compliance risk. Entruva issues each agent a W3C-standard DID and a signed verifiable credential that any system can check in real time.
Entruva computes a live AgentRep score from real interaction outcomes logged after each task via the API. The score reflects actual performance — success rate, scope compliance, and outcome quality — not self-reported claims. It updates in real time and is visible on the agent's public Trust Card.
Yes. The EU AI Act requires high-risk AI systems to maintain documentation of capabilities, authorization scope, and audit trails. Entruva automatically generates this evidence: every credential is timestamped and cryptographically signed, every verification is logged, and every outcome is recorded. Pro plan users can export a compliance evidence bundle on demand.
Yes — verification is a single public HTTP GET. No SDK, no account required. The response includes authorization scope, reputation score, and live status. Any marketplace, enterprise system, or third-party application can verify an agent before granting access.
You can revoke any credential instantly from the dashboard or via API. Revocation propagates immediately — any verification call will return invalid. You can also set automatic revocation rules: if reputation drops below a threshold or a scope violation is detected, the credential is killed automatically and a webhook fires to your system.
Free 14-day trial. No credit card required. First credential in 5 minutes.