An agent passport (or agent credential) is a portable, verifiable credential describing an agent’s identity and authorisations — who it is, who issued it, what it may do, and until when — that the agent can present to a service it has no prior relationship with.
Why it matters
Agents increasingly interact with strangers: they browse, book, buy, and call APIs across organisational boundaries. There is no shared directory to look the agent up in. A verifiable credential lets the agent carry its own proof, and lets the relying service check it without a prior trust relationship.
The building blocks are standard:
- W3C Verifiable Credentials Data Model 2.0 defines the issuer/holder/verifier roles and a tamper-evident data model for exactly this kind of claim.
- OpenID for Verifiable Credential Issuance (OID4VCI) defines how such credentials are issued to a wallet or holder.
- eIDAS 2.0 / EUDI wallets (Regulation (EU) 2024/1183) make verifiable credentials part of mainstream EU digital identity, a likely carrier for agent credentials.
- Payment protocols such as Google’s AP2 use signed mandates carried as verifiable credentials so an agent can prove the user’s intent to a merchant.
The “passport” metaphor is informal and not a standard term. What matters is that the credential is verifiable by a third party and scoped to a purpose.
What a useful agent credential expresses
- Issuer (who vouches for the agent).
- The agent’s identity and its principal.
- Authorised actions, limits, and data scope.
- Validity window and status/revocation reference.
How it works in practice
A verifiable credential has three roles: an issuer signs it, a holder (the agent) stores and presents it, and a verifier checks the signature and the claims. This is what lets verification work without a prior relationship, and without the verifier calling the issuer’s database for every request.
For agents, the credential carries the identity and authorisation payload — principal, permitted actions, limits, expiry — so a merchant or API can decide on the spot. Because the format is portable, the same credential model can span services; because it can reference a status method, it can also be revoked. See KYA vs KYC vs KYB for how this differs from customer due diligence, and the verification guide for the check sequence, with code in the examples.
Related terms
Sources
- W3C, Verifiable Credentials Data Model v2.0 (accessed 2026-10-03)
- OpenID Foundation, OpenID for Verifiable Credential Issuance (accessed 2026-10-03)
- European Commission, EUDI Regulation (EU) 2024/1183 (accessed 2026-10-03)
- Google, Agent Payments Protocol (AP2) (accessed 2026-10-03)
Know Your Agent (KYA) explains agent identity, verification and accountability. This is an explainer, not legal or compliance advice — see our Sources & methodology. New to KYA? Start with What is Know Your Agent? and the glossary.