<?xml version="1.0" encoding="utf-8"?>
<!DOCTYPE rfc [
  <!ENTITY nbsp "&#160;">
  <!ENTITY zwsp "&#8203;">
  <!ENTITY nbhy "&#8209;">
  <!ENTITY wj "&#8288;">
]>
<rfc xmlns:xi="http://www.w3.org/2001/XInclude" category="exp" docName="draft-ashcroft-approval-record-00" ipr="trust200902" submissionType="IETF" consensus="false" version="3" xml:lang="en" tocInclude="true" sortRefs="true" symRefs="true">
  <front>
    <title abbrev="Approval-Act Record">An Approval-Act Record for Standing Authorization Grants</title>
    <seriesInfo name="Internet-Draft" value="draft-ashcroft-approval-record-00"/>
    <author fullname="Joshua Ashcroft" initials="J." surname="Ashcroft">
      <organization>Kipple Labs</organization>
      <address>
        <email>josh.ashcroft@gmail.com</email>
      </address>
    </author>
    <author fullname="Blake Morrison" initials="B." surname="Morrison">
      <organization>Alter Meridian Pty Ltd</organization>
      <address>
        <email>blake@truealter.com</email>
      </address>
    </author>
    <author fullname="Nick Mathews" initials="N." surname="Mathews">
      <organization>Agentic Verification Architecture</organization>
      <address>
        <email>nick@avalayer.com</email>
      </address>
    </author>
    <date/>
    <abstract>
      <t>A standing authorization grant is an artifact that outlives the event that produced it. Existing authentication-strength vocabulary (acr/amr) describes authentication events and travels on tokens. Nothing carries a record of how a standing grant's approval was performed to the party verifying the grant, possibly days later and with no human present. This document defines an approval-act record: evidence of the approval ceremony, bound to the exact grant issuance approved, and carried in-band with the grant it authorizes. It defines one initial evidence type (WebAuthn assertions over a hash of the canonical grant), states precisely which of the record's claims a verifier can recompute, which rest on vendor attestation, and which remain attested by the operator, and provides a worked example with a failure matrix.</t>
    </abstract>
  </front>
  <middle>
  <section anchor="intro">
    <name>Introduction</name>
    <t>The Monday/Thursday problem, stated once. An operator approves a delegation on Monday behind a hardware authenticator. The agent acts on Thursday with no operator present. A token minted Thursday carries Thursday's acr: a service credential, or nothing. The approval act that actually stood behind the authority is invisible to every verifier that matters, and it cannot be reconstructed after the fact. If it is not captured when the grant is made, it is gone.</t>
    <t>The unattended case is the reason standing grants exist, so this document treats it as the main case.</t>
    <section anchor="reqs">
      <name>Requirements Language</name>
      <t>The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in BCP&nbsp;14 <xref target="RFC2119"/> <xref target="RFC8174"/> when, and only when, they appear in all capitals, as shown here.</t>
    </section>
    <section anchor="terms">
      <name>Terminology</name>
      <ul>
        <li>Standing authorization grant (grant): a signed artifact delegating authority to an agent, valid across many later presentations.</li>
        <li>Operator: the party that runs the agent and issues or obtains the grant.</li>
        <li>Approver: the human whose approval the grant is meant to reflect.</li>
        <li>Approval act: the ceremony in which the approver approves one grant issuance.</li>
        <li>Approval record (record): the evidence of the approval act defined in <xref target="record"/>.</li>
        <li>Verifier: any party that checks the grant and its record, typically a relying party receiving the agent's request.</li>
        <li>Operator anchor: the operator's JWK Set, published at the URI the Signature-Agent field names, under a name the operator controls, as used by the Web Bot Auth protocol <xref target="I-D.ietf-webbotauth-httpsig-protocol"/>.</li>
        <li>Covered issued-at: the issuance time carried inside the grant bytes covered by the record's challenge.</li>
        <li>Approval interval: the verifier-local interval defined in <xref target="time"/> within which the approval act provably occurred.</li>
      </ul>
    </section>
  </section>
  <section anchor="constraints">
    <name>Design Constraints</name>
    <t>C1. In-band. The record arrives inside, or inseparably with, the grant it describes, checkable at request time from retained material, with no token exchange the verifier cannot perform.</t>
    <t>C2. Evidence relates to its issuer. The document states, per claim, who issues the evidence and what a verifier can recompute versus must take on attestation. It never implies the evidence outranks its issuer.</t>
    <t>C3. Reuse existing vocabulary. acr/amr semantics (OIDC <xref target="OpenID.Core"/>), including the existing amr value "hwk" <xref target="RFC8176"/> and the EAP ACR values <xref target="OpenID.EAP"/> for phishing-resistant authentication, are reused, not reinvented.</t>
    <t>C4. Verifier-required, not self-declared. Whether a record is demanded, and at what strength, is the relying party's policy. An asserted value a verifier cannot check is treated as absent.</t>
    <t>C5. Parameters, not classes. Timing and artifact binding are parameters of the record rather than new assurance classes; see <xref target="parameters"/>.</t>
  </section>
  <section anchor="record">
    <name>The Approval Record</name>
    <t>A grant carries at most one approval record. The grant itself MUST carry a unique issuance identifier (a serial, or issuance timestamp plus id) as a covered field. The REQUIRED form is deliberate. An optional identifier would create a third state, bound but content-only, that every verifier handles differently and no vocabulary names. A grant format that cannot carry an identifier is not excluded; it is priced, through captured binding, under <xref target="parameters"/>.</t>
    <t>Fields of the record:</t>
    <ul>
      <li>evidence_type: registry identifier (<xref target="iana"/>). This document defines "webauthn".</li>
      <li>evidence: the raw evidence artifact, carried inline. For webauthn: authenticatorData, clientDataJSON and signature.</li>
      <li>attestation: OPTIONAL. The authenticator's attestation statement for the approver credential. Attestation belongs to the credential and is fixed at enrollment, so a verifier MAY retain it from the first record it sees for that credential, and a record MAY omit it when the verifier is expected to hold it already. A record that omits attestation still carries the credential identifier, so a verifier can find the attestation it retained from an earlier record for that credential. A verifier that holds no attestation for the credential, whether omitted, never sent or unobtainable, reports claim 2's hardware claim as unattested; it MUST NOT conclude invalid or unverifiable on that ground alone.</li>
      <li>credential: the approver credential ID and public key.</li>
      <li>binding: challenge = H(canonical grant bytes), where the grant bytes include the issuance identifier. There is no record-side nonce. The uniqueness that makes an approval unreplayable is a property of the grant, covered by the challenge, or it protects nothing. Canonicalization is per grant format, not global: each grant format registry entry owns its canonical form. The JSON grant format this document registers uses JCS <xref target="RFC8785"/>; a CBOR or binary format states its own.</li>
      <li>approved_at: timestamp of the approval act. Operator-asserted; <xref target="parameters"/> states what that prices.</li>
      <li>parameters: artifact_binding and timing, per <xref target="parameters"/>.</li>
    </ul>
    <t>The record, attestation included when present, MUST NOT exceed 8&nbsp;KiB. A verifier that measures a record above the ceiling MUST conclude invalid: the check completed and the record failed it. Unverifiable is reserved for a check that could not be performed.</t>
    <section anchor="carriage">
      <name>Carriage</name>
      <t>The record travels beside the grant. Its challenge is the hash of the grant bytes, and a grant already signed by its issuer cannot take a new member without changing the bytes its issuer signed. Keeping the record outside leaves the grant exactly as issued, so any grant format can carry a record, and no grant is re-issued to attach an approval.</t>
      <t>In JSON, the grant and its record travel together in a wrapper object with two members, grant and approval_record.</t>
      <t>Over HTTP, the record travels in its own header field, Approval-Record, alongside the grant. When the request is signed under the Web Bot Auth protocol (HTTP Message Signatures <xref target="RFC9421"/>), the signature MUST cover the Approval-Record field. A record stripped in transit then breaks the signature, where otherwise it would move the request silently from the record VALID column to record ABSENT (<xref target="example"/>). The 8&nbsp;KiB ceiling keeps the field inside common default per-header limits.</t>
    </section>
  </section>
  <section anchor="claims">
    <name>What the Record Claims, and Does Not</name>
    <t>Three claims, three issuers.</t>
    <t>Claim 1: This exact issuance was approved. The verifier recomputes the hash of the grant it holds, issuance identifier included, and checks it against the signed challenge. No issuer sits in that path; the claim is recomputable. The choice of semantics is deliberate and stated: the approval binds an issuance, not a content. Two re-issues of identical content are two approvals, and an approval of one is evidence for neither of the other. This is what makes an old record unreplayable onto a re-issued grant. A reader who wants content binding is asking for a different and weaker claim, and this document declines to blur them.</t>
    <t>Claim 2: The approval was performed on a hardware authenticator with a user present and verified. The presence and verification flags ride inside the signed authenticator data, checkable against the credential public key. The hardware claim rests on enrollment-time attestation. Its issuer is the authenticator vendor: trust relocated to a party with no stake in any particular grant, not removed. Where the verifier holds no attestation for the credential, the hardware claim is reported as unattested while the presence and verification flags still verify. An authenticator identifier (AAGUID) carried without attestation is self-asserted and proves nothing about the hardware.</t>
    <t>Claim 3: That credential belongs to that operator's approver. The residue. The party being graded states it, nothing recomputes it, and an operator could enroll its own credential and manufacture human-looking approvals. A verifier's confidence in claim 3 is exactly its confidence in the operator's enrollment practice, and this document says so plainly.</t>
    <t>The honest composite: when all three claims verify, some human holding the enrolled credential approved this exact issuance.</t>
    <t>Three partial states are named, so implementations have something to switch on other than a hard fail:</t>
    <ul>
      <li>APPROVED BY AN ENROLLED CREDENTIAL, HOLDER UNATTRIBUTED. Claims 1 and 2 verify; claim 3 cannot be evaluated (for example, the operator anchor is unreachable).</li>
      <li>APPROVED BY AN ENROLLED CREDENTIAL, HOLDER ATTRIBUTION UNRESOLVED AT REVOCATION. Claims 1 and 2 verify, but the approver credential was revoked at a time inside the approval interval (<xref target="parameters"/>), so the verifier cannot tell whether the credential was still held by the approver when the act occurred. The name carries both facts on purpose. A verifier needs to know which half is open, and a bare label would be implemented as a hard fail.</li>
      <li>APPROVED BY AN ENROLLED CREDENTIAL, HOLDER ATTRIBUTION EXPIRED. Claims 1 and 2 verify, but the operator anchor has lapsed (<xref target="example"/>), so claim 3 rests on an expired anchor and the holder attribution expired with it.</li>
    </ul>
  </section>
  <section anchor="parameters">
    <name>Parameters, Not Classes</name>
    <t>Two orthogonal parameters appear on every record.</t>
    <t>artifact_binding = bound | captured</t>
    <ul>
      <li>bound: the evidence's challenge is derived from the grant itself. Bound evidence is safely reusable across presentations of that grant precisely because it can attach to nothing else.</li>
      <li>captured: the evidence was produced at approval time but is not cryptographically bound to this grant. Reuse is unsafe; verifiers SHOULD treat captured evidence as weaker than bound.</li>
    </ul>
    <t>timing = fresh | standing</t>
    <ul>
      <li>fresh: evidence produced within a policy-defined window of the presentation it accompanies.</li>
      <li>standing: evidence produced at grant time; the artifact outlives the ceremony. This is the normal case for standing grants and the reason this document exists.</li>
    </ul>
    <t>A single class value cannot express that bound plus standing is safe while captured plus standing is not. That asymmetry is the argument for parameters.</t>
    <section anchor="time">
      <name>Time, and Whose It Is</name>
      <t>approved_at has an issuer, and it is the operator. Nothing the authenticator signs carries a clock. authenticatorData holds the RP ID hash, the flags, a signature counter (ordinal rather than temporal, and zero on a great many authenticators) and extension output. clientDataJSON holds the type, the challenge and the origin. So the record's one timestamp rides outside the signed material, and a verifier applying a freshness policy measured from approved_at is trusting the operator's clock while believing it is checking something.</t>
      <t>The situation is better than pure trust by exactly one interval. The challenge covers the grant bytes, and the grant carries its issued-at inside those bytes. The assertion cannot exist before the grant it covers, so the covered issued-at is a lower bound on the approval act. The verifier's first sight of the record is an upper bound on its own clock. The approval provably lies inside that interval, and approved_at is an operator-asserted point within it.</t>
      <t>Two freshness claims follow, named, so a verifier picks the one it can afford:</t>
      <ul>
        <li>ASSERTED freshness: a window measured from approved_at. Operator grade: exactly as good as the operator's clock and honesty.</li>
        <li>CHECKABLE freshness: a window measured from the covered issued-at, evaluated against the verifier's own clock. Recomputable from covered material, with no issuer in the path beyond the grant's own.</li>
      </ul>
      <t>Three rules keep the checkable claim checkable:</t>
      <ol>
        <li>The upper bound is VERIFIER-LOCAL and MUST NOT be carried as a record field. Two verifiers reading one record compute different intervals, and the narrower one is not more true. Written into the record, the bound becomes one more operator-asserted timestamp under a new name.</li>
        <li>First sight means FIRST USE. The interval tightens only for a verifier that keeps state across uses of a standing grant. A one-shot verifier gets the loosest interval, and one-shot is the common case.</li>
        <li>The interval is only as independent as the grant's own issuer. Where the operator issues the grant, the covered issued-at is the operator's own assertion and the interval COLLAPSES to operator grade. This caveat sits here, beside the interval, deliberately: a reader who takes the interval from one section and its collapse condition from another will implement the first and not the second.</li>
      </ol>
    </section>
    <section anchor="revocation">
      <name>Revocation, Measured Against the Same Interval</name>
      <t>When the approver credential is revoked, the revocation's effective time is compared with the interval above, not with approved_at:</t>
      <ul>
        <li>Effective after the verifier's upper bound: the record stands; the credential was valid when the act occurred.</li>
        <li>Effective before the covered issued-at: invalid; the approver held no valid credential when the act is claimed to have happened.</li>
        <li>Effective inside the interval: genuinely doubtful, and doubtful verifier-locally in the same way the upper bound is. Rules 1 to 3 above govern it, and the result is the named partial state in <xref target="claims"/>, APPROVED BY AN ENROLLED CREDENTIAL, HOLDER ATTRIBUTION UNRESOLVED AT REVOCATION, which is not a hard fail.</li>
      </ul>
    </section>
  </section>
  <section anchor="webauthn">
    <name>Evidence Type: WebAuthn</name>
    <t>Challenge construction. challenge = H(canonical grant bytes), where H is SHA-256 <xref target="RFC6234"/>, fixed by this document. There is no per-record hash field, because a negotiable hash is a downgrade lever every verifier must defend forever. Agility lives in the registry: a future evidence_type, or a versioned entry, carries a different hash when one is needed.</t>
    <t>Verifier procedure.</t>
    <ol>
      <li>Recompute H from the grant it holds.</li>
      <li>Check equality with clientDataJSON.challenge.</li>
      <li>Verify the assertion signature against credential <xref target="WebAuthn"/>.</li>
      <li>Read the UP and UV flags from authenticatorData.</li>
      <li>Evaluate the attestation, from the record or retained from an earlier record for the same credential. With none held, report the hardware claim as unattested.</li>
    </ol>
    <t>UV-required versus UP-sufficient maps to acr values per relying-party policy.</t>
  </section>
  <section anchor="related">
    <name>Relationship to Existing Work</name>
    <ul>
      <li>OIDC acr/amr <xref target="OpenID.Core"/> and EAP ACR values <xref target="OpenID.EAP"/>: vocabulary source (C3). An approval record is roughly acr/amr semantics attached to a durable artifact instead of an authentication event.</li>
      <li>IPSIE <xref target="IPSIE"/>: overlapping interest in standardized authentication-strength claims. Flagged, deliberately not settled here.</li>
      <li>draft-ietf-webbotauth-httpsig-protocol-01 <xref target="I-D.ietf-webbotauth-httpsig-protocol"/>: supplies the operator anchor and the request signature that covers the Approval-Record field (<xref target="carriage"/>). The anchor outcome vocabulary in <xref target="example"/> follows its discipline of keeping verified, invalid and unverified distinct.</li>
      <li>draft-brezun-human-continuity-http <xref target="I-D.brezun-human-continuity-http"/>, Section 7.5: the per-request half of this mechanism appears expressible as a presentation-assurance profile of that framework. Its rule that asserted values are unsatisfied unless the profile verifies evidence is the same discipline as C4. Boundary: a profile carrying only this document's assurance semantics does NOT establish that a credential belongs to a unique human, and must say so explicitly.</li>
      <li>draft-morrison-acts-for <xref target="I-D.morrison-acts-for"/>: complementary. Acts-For defines a pass an agent carries when acting for a person and leaves the grant behind it out of scope.</li>
      <li>AAuth <xref target="AAuth"/> (Person/Mission): complementary. The token path answers the interactive case; this document answers the standing case.</li>
      <li>Post-execution record layers: an approval record gives an execution record something to say about the question disputes actually litigate. A record attributes; it never manufactures authority, and it cannot recover approval strength that was not captured at grant time.</li>
    </ul>
  </section>
  <section anchor="example">
    <name>Worked Example and Failure Matrix</name>
    <t>One agent. Operator anchor on one side (a JWK Set at a Signature-Agent URI the operator controls), approval-bound grant on the other. Materials are real WebAuthn assertions over real grant hashes, produced by a working prototype and labeled as such. The prototype mints a JCS grant, runs the passkey ceremony over its hash and verifies per Sections <xref target="claims" format="counter"/> to <xref target="webauthn" format="counter"/>. The failure set is 13 vectors, one fault each, with each basis quoting the section it tests. The current fixture is a 1Password passkey with attestation "none", so the verifier reports claim 2's hardware claim as unattested.</t>
    <t>Anchor outcomes are three: VALID; LAPSED, a conclusive outcome where the anchor expired and the verifier knows it; and UNREACHABLE, which is not an outcome because the check never completed. Anchor changed-hands is deliberately absent from the matrix. It is an ordering question between two events, which is record-layer territory, referenced rather than absorbed.</t>
    <t>Cells report per-claim outcomes, not a single verdict.</t>
    <table anchor="matrix">
      <thead>
        <tr><th></th><th>Record VALID</th><th>Record ABSENT</th><th>Record INVALID</th></tr>
      </thead>
      <tbody>
        <tr><td>Anchor VALID</td><td>Claim 1 ok, Claim 2 ok, Claim 3 ok: accountable operator, human-approved this exact issuance</td><td>Accountable operator; approval strength unknown (status quo)</td><td>Invalid record. Forgery is one possible cause. The one cell treated worse than absent.</td></tr>
        <tr><td>Anchor LAPSED (conclusive)</td><td>Claim 1 ok, Claim 2 ok, Claim 3 rests on an expired anchor: APPROVED BY AN ENROLLED CREDENTIAL, HOLDER ATTRIBUTION EXPIRED. A named partial state, not a hard fail.</td><td>Operator accountability expired; nothing else asserted</td><td>Invalid record, expired anchor: nothing, loudly</td></tr>
        <tr><td>Anchor UNREACHABLE (no outcome)</td><td>Claim 1 ok, Claim 2 ok, Claim 3 unevaluated: APPROVED BY AN ENROLLED CREDENTIAL, HOLDER UNATTRIBUTED. A named partial state, not a hard fail.</td><td>NO VERDICT: the check never completed</td><td>Invalid is invalid regardless of anchor state</td></tr>
      </tbody>
    </table>
    <t>Non-independence warning. The two axes share a trust root. Claim 3's issuer is the anchor party, so anchor-valid is a precondition of claim 3, not corroborating evidence for it. The anchor column and the claim 3 row are the same trust, spent twice; a reader must not sum them.</t>
  </section>
  <section anchor="impl">
    <name>Implementation Status</name>
    <t>Per <xref target="RFC7942"/>, to be removed before publication as an RFC. One prototype exists: a Cloudflare Worker that mints a JCS grant, performs the WebAuthn ceremony over its hash, verifies per Sections <xref target="claims" format="counter"/> to <xref target="webauthn" format="counter"/>, and generates the 13-vector failure set from a real ceremony. Coverage: one 1Password passkey fixture, attestation "none". The vectors are available from the authors on request.</t>
  </section>
  <section anchor="security">
    <name>Security Considerations</name>
    <ul>
      <li>Replay: an approval binds one issuance because the issuance identifier is inside the covered grant bytes. There is no record-side nonce to manage and nothing to reuse across re-issues.</li>
      <li>Stolen approver credential: bounded by expiry and revocation of the credential enrollment. Whether revocation casts doubt backward is decided by measuring its effective time against the approval interval (<xref target="revocation"/>): after the interval the record stands, before it the record is invalid, inside it the result is the named partial state in <xref target="claims"/>.</li>
      <li>Self-enrollment (claim 3 residue): stated, not solved. Mitigations are possible at enrollment (attestation policy) and are out of scope here.</li>
      <li>Record stripping: a request whose signature does not cover the Approval-Record field can lose its record in transit and fall silently into the record ABSENT column; <xref target="carriage"/> requires coverage for that reason.</li>
      <li>Size ceiling: the 8&nbsp;KiB ceiling keeps a record inside common default per-header limits on HTTP servers, and a record over the ceiling is invalid so that oversized records cannot be pushed into the unverifiable column.</li>
      <li>Grant revocation: revocation of the grant itself is unchanged by this document and must remain cheaper than approval.</li>
      <li>Credential correlation: the raw credential public key travels to every verifier of every grant the approver signs, and it is exactly the correlator the at-rest rule in <xref target="privacy"/> exists for. Pairwise keys stay incompatible with standing grants, so that rule, not the wire, is where correlation is limited.</li>
    </ul>
  </section>
  <section anchor="iana">
    <name>IANA Considerations</name>
    <ul>
      <li>Approval Record Evidence Types registry (new). Policy: Specification Required. Columns: identifier, hash function, description, reference. Initial entry: "webauthn", SHA-256, this document.</li>
      <li>Approval Record Grant Formats registry (new). Policy: Specification Required. Columns: identifier, canonicalization, issuance-identifier field, issued-at field, reference. Initial entry: JSON grant, JCS <xref target="RFC8785"/>, this document.</li>
      <li>HTTP Field Name Registry: register Approval-Record (<xref target="carriage"/>). Status: permanent. Reference: this document.</li>
    </ul>
  </section>
  <section anchor="privacy">
    <name>Privacy Considerations</name>
    <t>The record exposes an approval event and a credential identifier to every verifier of the grant. Three facts follow.</t>
    <t>The credential public key is the sharpest exposure. A verifier needs it to check the assertion, so the record cannot withhold it. Where the approver is the person the agent acts for, every verifier receives the same key and can link all of that person's agents through it. The raw key is exactly the correlator the at-rest rule below exists for.</t>
    <t>Pairwise credentials per relying party are INCOMPATIBLE with standing grants. They require enumerating the verifiers at approval time, and the reason standing grants exist is that you cannot.</t>
    <t>The workable lever is at rest, not on the wire. A verifier SHOULD store H(credential_id, per-verifier salt) rather than the raw identifier. Per-verifier continuity survives, cross-verifier correlation ends, and C1 is untouched because nothing about the record changes. The protocol cannot enforce this, since it is storage policy, which is why it is a verifier SHOULD rather than a wire mechanism.</t>
  </section>
  </middle>
  <back>
    <references anchor="refs">
      <name>References</name>
      <references>
        <name>Normative References</name>
        <reference anchor="RFC2119" target="https://www.rfc-editor.org/info/rfc2119">
          <front>
            <title>Key words for use in RFCs to Indicate Requirement Levels</title>
            <author fullname="S. Bradner" initials="S." surname="Bradner"/>
            <date month="March" year="1997"/>
          </front>
          <seriesInfo name="BCP" value="14"/>
          <seriesInfo name="RFC" value="2119"/>
          <seriesInfo name="DOI" value="10.17487/RFC2119"/>
        </reference>
        <reference anchor="RFC8174" target="https://www.rfc-editor.org/info/rfc8174">
          <front>
            <title>Ambiguity of Uppercase vs Lowercase in RFC 2119 Key Words</title>
            <author fullname="B. Leiba" initials="B." surname="Leiba"/>
            <date month="May" year="2017"/>
          </front>
          <seriesInfo name="BCP" value="14"/>
          <seriesInfo name="RFC" value="8174"/>
          <seriesInfo name="DOI" value="10.17487/RFC8174"/>
        </reference>
        <reference anchor="RFC8785" target="https://www.rfc-editor.org/info/rfc8785">
          <front>
            <title>JSON Canonicalization Scheme (JCS)</title>
            <author fullname="A. Rundgren" initials="A." surname="Rundgren"/>
            <author fullname="B. Jordan" initials="B." surname="Jordan"/>
            <author fullname="S. Erdtman" initials="S." surname="Erdtman"/>
            <date month="June" year="2020"/>
          </front>
          <seriesInfo name="RFC" value="8785"/>
          <seriesInfo name="DOI" value="10.17487/RFC8785"/>
        </reference>
        <reference anchor="RFC6234" target="https://www.rfc-editor.org/info/rfc6234">
          <front>
            <title>US Secure Hash Algorithms (SHA and SHA-based HMAC and HKDF)</title>
            <author fullname="D. Eastlake 3rd" initials="D." surname="Eastlake 3rd"/>
            <author fullname="T. Hansen" initials="T." surname="Hansen"/>
            <date month="May" year="2011"/>
          </front>
          <seriesInfo name="RFC" value="6234"/>
          <seriesInfo name="DOI" value="10.17487/RFC6234"/>
        </reference>
        <reference anchor="RFC9421" target="https://www.rfc-editor.org/info/rfc9421">
          <front>
            <title>HTTP Message Signatures</title>
            <author fullname="A. Backman" initials="A." role="editor" surname="Backman"/>
            <author fullname="J. Richer" initials="J." role="editor" surname="Richer"/>
            <author fullname="M. Sporny" initials="M." surname="Sporny"/>
            <date month="February" year="2024"/>
          </front>
          <seriesInfo name="RFC" value="9421"/>
          <seriesInfo name="DOI" value="10.17487/RFC9421"/>
        </reference>
        <reference anchor="WebAuthn" target="https://www.w3.org/TR/webauthn-3/">
          <front>
            <title>Web Authentication: An API for accessing Public Key Credentials Level 3</title>
            <author>
              <organization>W3C</organization>
            </author>
            <date/>
          </front>
          <seriesInfo name="W3C" value="Web Authentication Level 3"/>
        </reference>
        <reference anchor="OpenID.Core" target="https://openid.net/specs/openid-connect-core-1_0.html">
          <front>
            <title>OpenID Connect Core 1.0</title>
            <author>
              <organization>OpenID Foundation</organization>
            </author>
            <date month="November" year="2014"/>
          </front>
        </reference>
        <reference anchor="RFC8176" target="https://www.rfc-editor.org/info/rfc8176">
          <front>
            <title>Authentication Method Reference Values</title>
            <author fullname="M. Jones" initials="M." surname="Jones"/>
            <author fullname="P. Hunt" initials="P." surname="Hunt"/>
            <author fullname="A. Nadalin" initials="A." surname="Nadalin"/>
            <date month="June" year="2017"/>
          </front>
          <seriesInfo name="RFC" value="8176"/>
          <seriesInfo name="DOI" value="10.17487/RFC8176"/>
        </reference>
        <reference anchor="OpenID.EAP" target="https://openid.net/specs/openid-connect-eap-acr-values-1_0.html">
          <front>
            <title>OpenID Connect Extended Authentication Profile (EAP) ACR Values 1.0</title>
            <author fullname="Michael B. Jones" initials="M.B." surname="Jones">
              <organization>Self-Issued Consulting</organization>
            </author>
            <date month="June" year="2025"/>
          </front>
        </reference>
      </references>
      <references>
        <name>Informative References</name>
        <reference anchor="RFC7942" target="https://www.rfc-editor.org/info/rfc7942">
          <front>
            <title>Improving Awareness of Running Code: The Implementation Status Section</title>
            <author fullname="Y. Sheffer" initials="Y." surname="Sheffer"/>
            <author fullname="A. Farrel" initials="A." surname="Farrel"/>
            <date month="July" year="2016"/>
          </front>
          <seriesInfo name="BCP" value="205"/>
          <seriesInfo name="RFC" value="7942"/>
          <seriesInfo name="DOI" value="10.17487/RFC7942"/>
        </reference>
        <reference anchor="I-D.ietf-webbotauth-httpsig-protocol" target="https://datatracker.ietf.org/doc/html/draft-ietf-webbotauth-httpsig-protocol-01">
          <front>
            <title>HTTP Message Signatures for automated traffic</title>
            <author fullname="Thibault Meunier" initials="T." surname="Meunier">
              <organization>Cloudflare</organization>
            </author>
            <author fullname="Sandor Major" initials="S." surname="Major">
              <organization>Google</organization>
            </author>
            <date day="29" month="October" year="2026"/>
          </front>
          <seriesInfo name="Internet-Draft" value="draft-ietf-webbotauth-httpsig-protocol-01"/>
        </reference>
        <reference anchor="I-D.brezun-human-continuity-http" target="https://datatracker.ietf.org/doc/html/draft-brezun-human-continuity-http-00">
          <front>
            <title>Human Continuity for HTTP</title>
            <author fullname="Aurelien Brezun" initials="A." surname="Brezun">
              <organization>Tools for Humanity</organization>
            </author>
            <date day="6" month="August" year="2026"/>
          </front>
          <seriesInfo name="Internet-Draft" value="draft-brezun-human-continuity-http-00"/>
        </reference>
        <reference anchor="I-D.morrison-acts-for" target="https://datatracker.ietf.org/doc/html/draft-morrison-acts-for-01">
          <front>
            <title>Acts-For: Signing In to a Website with a ~handle, for a Person or an Agent Acting for One</title>
            <author fullname="Blake Morrison" initials="B." surname="Morrison">
              <organization>Alter Meridian Pty Ltd</organization>
            </author>
            <date day="3" month="November" year="2026"/>
          </front>
          <seriesInfo name="Internet-Draft" value="draft-morrison-acts-for-01"/>
        </reference>
        <reference anchor="AAuth" target="https://datatracker.ietf.org/doc/html/draft-hardt-oauth-aauth-protocol-11">
          <front>
            <title>AAuth Protocol</title>
            <author fullname="Dick Hardt" initials="D." surname="Hardt">
              <organization>Hellō</organization>
            </author>
            <date day="25" month="September" year="2026"/>
          </front>
          <seriesInfo name="Internet-Draft" value="draft-hardt-oauth-aauth-protocol-11"/>
        </reference>
        <reference anchor="IPSIE" target="https://datatracker.ietf.org/wg/ipsie/documents/">
          <front>
            <title>IPSIE working group documents</title>
            <author>
              <organization>IETF</organization>
            </author>
            <date/>
          </front>
        </reference>
      </references>
    </references>
    <section anchor="open-issues">
      <name>Open Issues</name>
      <ol>
        <li>Intended status: Standards Track or Experimental.</li>
        <li>Whether the two new registries should merge into one.</li>
        <li>A second, attested fixture (a security key with packed attestation) for the worked example.</li>
        <li>Final names for the approval_record member and the Approval-Record field.</li>
      </ol>
    </section>
    <section anchor="acks" numbered="false">
      <name>Acknowledgments</name>
      <t>Drew, Alter Meridian Pty Ltd. Aurelien Brezun: parameters, not classes; the unique-human boundary.</t>
    </section>
  </back>
</rfc>
