OpenID4VP primary sources — 2026-05-15

What I reviewed

Key takeaways from primary sources

  • OpenID4VP is not merely a credential-format or wallet-login effort. The core spec defines an OAuth-based protocol for requesting and receiving credential presentations across multiple credential families, including W3C VC, ISO mdoc, and SD-JWT VC.
  • The most important structural contribution is the decomposition of presentation into separate protocol surfaces: verifier queries, wallet invocation, request-object transport, response delivery, encryption, verifier metadata, wallet metadata, verifier attestation, and conformance profiling are all first-class parts of the stack.
  • The spec’s same-device and cross-device flows are analytically important because they show that presentation transport is not incidental UX. Request transport, QR/bootstrap behavior, request_uri hosting, and direct-post response handling are part of the protocol’s real control plane.
  • DCQL is one of the clearest reasons this entry belongs in the corpus. It makes verifier demand explicit by defining credential queries, trusted-authority constraints, claims queries, and credential-set logic instead of collapsing verifier policy into an opaque application-defined request format.
  • The vp_token response type matters because it reorients an OAuth transaction away from API access-token issuance and toward the return of one or more presentations. That keeps presentation as a transport/governance problem distinct from ordinary login or API delegation.
  • Verifier metadata, client identifier prefix handling, and verifier attestation JWT support make verifier legitimacy a first-class protocol surface. That is important because many real deployments otherwise move verifier trust into wallet allowlists, browser trust cues, or app-specific heuristics.
  • The privacy/security sections are part of the mechanism, not just boilerplate. The spec explicitly surfaces replay prevention, response URI protection, selective disclosure, verifier-to-verifier unlinkability, purpose legitimacy, issuer trust, and conformance testing as requirements for safe presentation.
  • The OpenID Foundation materials make it clear that OpenID4VP is already operating inside an implementation-governance environment: wallets and verifiers have conformance tests, HAIP-coupled test plans, and a 2026 self-certification path tied to jurisdiction-specific deployment profiles.

Specific source signals worth retaining

  • The main 1.0 specification page and the raw WG draft show the architectural split clearly: vp_token, same-device versus cross-device flows, request_uri, direct_post, direct_post.jwt, DCQL, wallet metadata, verifier metadata, verifier attestation JWT, and Digital Credentials API integration are all explicit sections rather than secondary notes.
  • The raw 1.0 working-group draft is especially useful because it spells out the cross-device QR / request-object flow and makes the profiling requirement unusually explicit: OpenID4VP is a framework that still expects ecosystems to constrain optional features and parameter values.
  • The OpenID4VC specifications index is the best compact source for situating OpenID4VP inside the broader OpenID4VC family and for seeing how OIDF treats the specs as a maintained package rather than isolated PDFs.
  • The OpenID4VP repository README is the clearest source for implementation-adjacent maturity signals: actively maintained 1.0 and 1.1 WG drafts, public draft builds, and a dedicated conformance-testing path for wallets and verifiers.
  • The conformance-testing guide is a strong signal that interoperability is being operationalized through concrete test-plan expectations for response modes, client ID schemes, credential formats, request methods, and same-device versus cross-device behavior.
  • The February 2026 self-certification announcement is the best primary source for the governance/adoption layer: OIDF positions OpenID4VP 1.0 plus HAIP 1.0 as infrastructure already selected across dozens of jurisdictions, with open-source tests, CI integration, published certification results, and localized profiling paths.

Why this entry belongs in the corpus

  • It gives the corpus a standards-level baseline for the presentation phase beneath credential wallets, verifier SDKs, passport-proof containers, and region-specific digital-ID rollouts.
  • It strengthens the identity cluster by isolating a frequently flattened layer: verifier-side query design, wallet invocation, verifier trust signaling, and response-delivery policy are not the same thing as credential issuance or credential syntax.
  • The durable insight is that presentation ecosystems can centralize power in query languages, verifier-authentication policy, request-object hosting, transport profiles, and conformance packages rather than in the credential object alone. OpenID4VP makes those seams unusually visible.