<?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" version="3" category="exp" docName="draft-watts-rats-presilicon-evidence-00" ipr="trust200902" submissionType="independent" tocInclude="true" sortRefs="true" symRefs="true">

  <front>
    <title abbrev="Pre-Silicon Evidence Profile">An Entity Attestation Token (EAT) Profile for Signed Pre-Silicon Hardware Security Evidence</title>
    <seriesInfo name="Internet-Draft" value="draft-watts-rats-presilicon-evidence-00"/>
    <author fullname="Deonte Watts" initials="D." surname="Watts">
      <organization>ai-id-engine project</organization>
      <address>
        <email>deonte@goodshyt.fun</email><uri>https://github.com/DeontewattsV1/ai-id-engine</uri>
      </address>
    </author>
    <date year="2026" month="11" day="16"/>
    <area>Security</area>
    <workgroup>Network Working Group</workgroup>
    <keyword>RATS</keyword>
    <keyword>EAT</keyword>
    <keyword>TVLA</keyword>
    <keyword>side-channel</keyword>
    <keyword>pre-silicon</keyword>
    <keyword>provenance</keyword>

    <abstract>
      <t>Hardware security designs are increasingly assessed before
      fabrication: by gate-level leakage simulation, fault-injection
      simulation, formal proofs and equivalence checking. The results of
      these assessments are routinely copied into reports, slide decks and
      certification packages with no binding to the design that was
      assessed, the stimulus that was used, the conditions under which the
      numbers hold, or the platform that produced them. This document
      defines an Entity Attestation Token (EAT) profile for carrying such
      pre-silicon assessment results as signed Evidence. Every claim is
      classified as identifying the assessed artifact, describing a
      measured behavior, or describing the measurement conditions. A
      provenance claim, which states whether the signing key is rooted in
      hardware or software, is mandatory, and a simulated Trusted Platform
      Module is always reported as a software root. The profile is
      explicitly not a certification format: it records what was measured
      and under which envelope, so that a Relying Party can decide what the
      numbers are evidence of.</t>
    </abstract>

    <note removeInRFC="true">
      <name>About This Document</name>
      <t>A reference implementation (schema "aiid-evidence/v1", JSON
      encoding, TPM 2.0 or software-key signing) is in the ai-id-engine
      repository, files hwid/src/evidence.rs and docs/EVIDENCE.md. This
      draft describes that implementation and proposes how it maps onto
      EAT. It is experimental.</t>
    </note>
  </front>

  <middle>
    <section anchor="intro">
      <name>Introduction</name>
      <t>The RATS architecture <xref target="RFC9334"/> describes how an
      Attester produces Evidence about itself, how a Verifier appraises that
      Evidence, and how a Relying Party consumes the resulting Attestation
      Results. The Entity Attestation Token <xref target="RFC9711"/>
      provides a claims-based encoding for Evidence and allows profiles that
      constrain and extend it for a particular use.</t>

      <t>This document addresses a use that existing profiles do not cover:
      Evidence about a hardware design, produced before that design exists
      as silicon. Typical sources are:</t>
      <ul>
        <li>Test Vector Leakage Assessment (TVLA) <xref target="TVLA"/> on
        simulated, gate-level power traces of a cryptographic block;</li>
        <li>fault-injection simulation of error-detecting codes and state
        encodings;</li>
        <li>formal proofs of safety and non-interference properties;</li>
        <li>equivalence checks between register-transfer-level (RTL) source
        and a synthesized netlist;</li>
        <li>reproducible-build digests of an FPGA bitstream used as a
        prototype.</li>
      </ul>

      <t>Such results are only meaningful relative to the exact netlist,
      cell library, power model, stimulus set and statistical parameters
      that produced them. A first-order |t| value of 3.15 says nothing on
      its own. Together with "4000 traces per group, 4000 samples, alpha
      0.05, Bonferroni threshold 4.37, sky130 typical corner, zero-delay
      cell models, noiseless", it becomes a reproducible statement with a
      known minimum detectable effect. This profile binds the numbers to
      those conditions under a signature, and records the trust root of
      that signature honestly.</t>

      <t>This profile does not define how to perform any of these
      assessments, and it does not turn their results into a certification.
      A report that conforms to this profile can be evidence for a
      certification process; it is never itself a certificate.</t>

      <section anchor="terminology">
        <name>Requirements Language</name>
        <t>The key words "<bcp14>MUST</bcp14>", "<bcp14>MUST NOT</bcp14>",
        "<bcp14>REQUIRED</bcp14>", "<bcp14>SHALL</bcp14>",
        "<bcp14>SHALL NOT</bcp14>", "<bcp14>SHOULD</bcp14>",
        "<bcp14>SHOULD NOT</bcp14>", "<bcp14>RECOMMENDED</bcp14>",
        "<bcp14>NOT RECOMMENDED</bcp14>", "<bcp14>MAY</bcp14>", and
        "<bcp14>OPTIONAL</bcp14>" 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="defs">
        <name>Terms</name>
        <dl newline="false">
          <dt>Assessed artifact:</dt>
          <dd>The design under assessment: RTL source, synthesized
          netlist, derived power model, cell models and cell library.</dd>
          <dt>Assessment host:</dt>
          <dd>The machine that ran the assessment and signs the report. In
          RATS terms it is the Attester. Its Target Environment is the
          assessment run, not the assessed artifact.</dd>
          <dt>Measurement envelope:</dt>
          <dd>The set of modelling assumptions under which the behavior
          claims hold, together with an explicit list of effects that were
          not modelled.</dd>
          <dt>Provenance root:</dt>
          <dd>"hardware" if the report signature is produced by a key held
          in a physical TPM; "software" otherwise.</dd>
        </dl>
      </section>
    </section>

    <section anchor="model">
      <name>Evidence Model</name>
      <t>A report is a body of named fields plus an attestation over the
      body. Each field carries exactly one class:</t>
      <dl newline="true">
        <dt>identity</dt>
        <dd>The field identifies what was assessed or who signed: artifact
        digests, the assessment host's identifier, and provenance.</dd>
        <dt>behavior</dt>
        <dd>The field is a measured outcome: leakage statistics, a gate
        verdict, area, latency, fault-detection counts, proof status.</dd>
        <dt>neither</dt>
        <dd>The field describes the conditions of the measurement:
        statistical parameters, the stimulus set, the measurement envelope,
        tool versions.</dd>
      </dl>
      <t>The classification is part of the signed body. A Verifier
      <bcp14>MUST</bcp14> reject a report in which:</t>
      <ul>
        <li>any field lacks a class, or has a class other than these three;</li>
        <li>a required field (<xref target="required"/>) is missing or
        carries the wrong class;</li>
        <li>a field name appears more than once.</li>
      </ul>
      <t>The classes let a Relying Party apply policy mechanically. For
      example, it can accept behavior claims only when the identity claims
      match a netlist digest it has independently obtained, and only when
      the neither claims show a trace count at or above its own minimum.</t>

      <section anchor="required">
        <name>Required Fields</name>
        <table anchor="tab-required">
          <name>Fields every conforming report carries</name>
          <thead>
            <tr><th>Name</th><th>Class</th><th>Content</th></tr>
          </thead>
          <tbody>
            <tr><td>provenance</td><td>identity</td>
                <td>Signing root; see <xref target="provenance"/>.</td></tr>
            <tr><td>hwid</td><td>identity</td>
                <td>Salted identifier of the assessment host.</td></tr>
            <tr><td>model_digest</td><td>identity</td>
                <td>Map from artifact path to SHA-256 <xref target="FIPS180-4"/>
                digest. It covers the RTL, netlists, power-model netlists and
                capacitance files, cell models and the cell library.</td></tr>
            <tr><td>stimulus_set</td><td>neither</td>
                <td>Digests of the stimulus generator, harness and
                configuration, plus seeds, modes and trace counts per
                group.</td></tr>
            <tr><td>tvla_parameters</td><td>neither</td>
                <td>Profile name, assessed design, traces per group, samples
                per trace (M), alpha, the first- and second-order critical
                values, and the number of second-order tests.</td></tr>
            <tr><td>measurement_envelope</td><td>neither</td>
                <td>See <xref target="envelope"/>.</td></tr>
            <tr><td>tvla_result</td><td>behavior</td>
                <td>Null-control and positive-control peaks, first- and
                second-order peaks, minimum detectable effect, verdict and
                gate exit code.</td></tr>
          </tbody>
        </table>
        <t>Additional fields <bcp14>MAY</bcp14> be present (for example
        area, latency, fault-injection counts or equivalence status). Each
        additional field <bcp14>MUST</bcp14> carry a class.</t>
      </section>

      <section anchor="controls">
        <name>Controls Are Part of the Result</name>
        <t>A leakage statistic without controls cannot be interpreted. The
        null control (random versus random inputs on the assessed design)
        bounds the false-positive behavior of the harness. The positive
        control (fixed versus random on an unprotected reference) shows
        that the harness can see a leak. A conforming tvla_result
        <bcp14>MUST</bcp14> include both control results. A Verifier
        <bcp14>MUST NOT</bcp14> treat a report as evidence of absence of
        leakage if the null control exceeded its threshold or the positive
        control did not.</t>
      </section>

      <section anchor="envelope">
        <name>Measurement Envelope</name>
        <t>The measurement_envelope states what the behavior claims are
        evidence of. It <bcp14>MUST</bcp14> contain:</t>
        <ul>
          <li>a kind string that says whether the traces were simulated or
          measured; for simulated traces it <bcp14>MUST</bcp14> contain the
          words "not silicon";</li>
          <li>the cell library and process corner;</li>
          <li>the power model and the wire-load model;</li>
          <li>the measurement chain parameters (for example low-pass time
          constant and additive noise);</li>
          <li>samples per trace;</li>
          <li>a list of effects not modelled (for example glitches,
          coupling, IR drop, electromagnetic emanation, process variation,
          parasitics after placement and routing);</li>
          <li>tool names and versions.</li>
        </ul>
        <t>The list of effects not modelled is <bcp14>REQUIRED</bcp14>
        because it is the part of a pre-silicon report most often lost when
        results are quoted elsewhere.</t>
      </section>
    </section>

    <section anchor="provenance">
      <name>Provenance</name>
      <t>The provenance field records the root of the key that signs the
      report. Its value contains:</t>
      <ul>
        <li>attestation_kind: "tpm2-quote" or "software-fallback";</li>
        <li>source: how the platform information was obtained, for example
        TPM2_GetCapability for TPM_PT_MANUFACTURER and TPM_PT_VENDOR_STRING_1
        <xref target="TPM2"/>, or "none" when no TPM was used;</li>
        <li>tpm_manufacturer, and a simulated_tpm flag;</li>
        <li>root: "hardware" or "software".</li>
      </ul>
      <t>The following rules apply:</t>
      <ul>
        <li>root <bcp14>MUST</bcp14> be "hardware" only when the signature
        is a TPM 2.0 quote from a physical TPM.</li>
        <li>A software TPM (for example swtpm or the Microsoft simulator)
        <bcp14>MUST</bcp14> be reported as root "software", even though it
        speaks the TPM 2.0 protocol.</li>
        <li>A plain software key <bcp14>MUST</bcp14> be reported as root
        "software".</li>
        <li>The Attester obtains the provenance before signing and places it
        in the body, so the signature covers it.</li>
        <li>A Verifier <bcp14>MUST</bcp14> recompute the expected provenance
        from the attestation it verified, and <bcp14>MUST</bcp14> reject the
        report if the two differ. In particular it rejects a body that
        claims "hardware" over a simulated TPM quote.</li>
      </ul>
    </section>

    <section anchor="signing">
      <name>Signing and Binding</name>
      <t>The body is serialized canonically: compact JSON
      <xref target="RFC8259"/> with fields in the order written. The report
      nonce is:</t>
      <artwork>
  nonce = SHA-256( "AIID-EVIDENCE-V1" || canonical(body) )
</artwork>
      <t>The nonce is used as the freshness value of the platform
      attestation.</t>
      <ul>
        <li>For a TPM, the quote's qualifying data is derived from the nonce
        and the host identifier, and the quote signs it with an Attestation
        Key whose binding to the Endorsement Key was checked by credential
        activation.</li>
        <li>For a software key, the key signs the nonce together with the
        host identifier and a timestamp.</li>
      </ul>
      <t>The report is the triple (body, body hash, attestation). A Verifier
      <bcp14>MUST</bcp14>:</t>
      <ol>
        <li>validate the body against <xref target="model"/>;</li>
        <li>recompute the canonical body hash and compare it;</li>
        <li>verify the attestation over the recomputed nonce;</li>
        <li>check provenance agreement as described in
        <xref target="provenance"/>.</li>
      </ol>
    </section>

    <section anchor="eat">
      <name>Mapping to EAT</name>
      <t>This section describes how the model maps onto an EAT
      <xref target="RFC9711"/> profile, so that reports can travel in RATS
      conceptual messages.</t>
      <ul>
        <li>The eat_profile claim <bcp14>MUST</bcp14> be set to the
        profile identifier in <xref target="iana-profile"/>.</li>
        <li>The eat_nonce claim carries the nonce defined in
        <xref target="signing"/>.</li>
        <li>Each body field is carried as a claim in a
        presilicon-evidence claims-set. The claim value is a map with keys
        "class" and "value".</li>
        <li>The provenance field maps onto the existing hardware and
        software model claims where those apply. It <bcp14>MUST</bcp14>
        also be carried verbatim, because EAT has no claim for "simulated
        TPM".</li>
        <li>When the attestation is a TPM quote, the token is carried
        together with the quote in the evidence bundle, as for other
        TPM-based Evidence. A software-root report <bcp14>SHOULD</bcp14> be
        signed with COSE or JOSE using the software key, and still declares
        root "software".</li>
      </ul>
    </section>

    <section anchor="appraisal">
      <name>Appraisal Guidance</name>
      <t>A Verifier producing Attestation Results from such Evidence
      <bcp14>SHOULD</bcp14>:</t>
      <ul>
        <li>report the provenance root unchanged;</li>
        <li>report the trace count and the minimum detectable effect beside
        any leakage verdict;</li>
        <li>never translate a passing verdict at a given trace count into a
        statement about leakage at higher trace counts or on silicon.</li>
      </ul>
      <t>A Relying Party <bcp14>SHOULD</bcp14> compare model_digest
      against artifacts it obtained independently, for example from a
      reproducible build whose digests are themselves published.</t>
    </section>

    <section anchor="example">
      <name>Example</name>
      <t>The following abbreviated body comes from the reference
      implementation. It describes a 2-share masked Montgomery multiplier
      at 4000 traces per group, which fails second-order TVLA as expected.
      Digests are truncated.</t>
      <sourcecode type="json">
{"schema":"aiid-evidence/v1","created":"2026-11-16T16:10:00Z",
 "fields":[
  {"name":"provenance","class":"identity","value":{
     "attestation_kind":"software-fallback",
     "source":"none: no TPM, software-fallback key file",
     "tpm_manufacturer":"","simulated_tpm":false,"root":"software"}},
  {"name":"hwid","class":"identity","value":"9f2c...e1"},
  {"name":"model_digest","class":"identity","value":{
     "build/sky130_masked_v2.v":"4b1e...07",
     "build/tvla_masked_v2.caps":"c3a9...5d"}},
  {"name":"stimulus_set","class":"neither","value":{
     "files":{"config/tvla.toml":"77d0...2a"},
     "description":{"fixed_vs_random":{"seed":17,
                    "traces_per_group":4000}}}},
  {"name":"tvla_parameters","class":"neither","value":{
     "profile":"ci","traces_per_group":4000,"samples":4000,
     "alpha":0.05,"crit_first_order":4.37,"crit_second_order":4.46}},
  {"name":"measurement_envelope","class":"neither","value":{
     "kind":"simulated gate-level power traces; not silicon",
     "library":"sky130_fd_sc_hd__tt_025C_1v80",
     "not_modelled":["glitches","coupling and IR drop","EM"]}},
  {"name":"tvla_result","class":"behavior","value":{
     "null_control":{"peak":3.24,"crit":4.37,"passed":true},
     "positive_control_plain":{"peak":82.96,"detected":true},
     "first_order_peak":3.15,"second_order_peak":26.50,
     "mde_sigma":0.0977,"exit":2,
     "verdict":"second-order leak in 2-share design"}}]}
</sourcecode>
    </section>

    <section anchor="security">
      <name>Security Considerations</name>
      <t>This profile improves the binding and honesty of pre-silicon
      results. It does not make them stronger than the assessment that
      produced them.</t>
      <dl newline="true">
        <dt>Simulation is not silicon.</dt>
        <dd>Zero-delay gate-level simulation does not model glitches,
        coupling, IR drop or electromagnetic emanation, and second-order
        statistics may be validated only synthetically. A Relying Party
        <bcp14>MUST NOT</bcp14> treat a conforming report as a side-channel
        certification.</dd>
        <dt>Trace count bounds what a pass means.</dt>
        <dd>A pass at N traces per group shows only that no leak larger than
        the stated minimum detectable effect was found at N.</dd>
        <dt>Provenance is host-reported.</dt>
        <dd>The platform information comes from the host's TPM software
        stack. It is covered by the Attestation Key signature, but a
        compromised host can misreport the TPM vendor unless the Verifier
        also validates the Endorsement Key certificate chain, which the
        reference implementation does not yet do.</dd>
        <dt>A software root is a file key.</dt>
        <dd>It proves possession of a key file, not the integrity of the
        host. Reports from continuous-integration systems will typically
        have a software root.</dd>
        <dt>Digests bind; they do not judge.</dt>
        <dd>model_digest ties the results to particular files, but the
        Relying Party must still establish that those files are the design
        it cares about.</dd>
        <dt>Replay.</dt>
        <dd>The nonce is derived from the body, so the same body always
        yields the same nonce. Verifiers that need freshness
        <bcp14>SHOULD</bcp14> also check the attestation timestamp or
        request a report bound to their own challenge.</dd>
      </dl>
    </section>

    <section anchor="privacy">
      <name>Privacy Considerations</name>
      <t>The hwid field is a salted identifier of the assessment host.
      Different salts yield unlinkable identifiers. Implementations
      <bcp14>SHOULD</bcp14> use a per-deployment salt and
      <bcp14>MUST NOT</bcp14> include raw hardware identifiers. Artifact
      paths in model_digest can reveal project structure, so publishers
      <bcp14>MAY</bcp14> replace them with opaque labels if the digests
      remain stable.</t>
    </section>

    <section anchor="iana">
      <name>IANA Considerations</name>
      <section anchor="iana-profile">
        <name>EAT Profile Identifier</name>
        <t>This document uses the profile identifier
        "tag:ai-id-engine,2026:presilicon-evidence/v1". No IANA action is
        requested for the identifier itself.</t>
      </section>
      <section anchor="iana-claims">
        <name>Claims</name>
        <t>If this document progresses, it would request registration of a
        "presilicon-evidence" claim in the CBOR Web Token (CWT) Claims and
        JSON Web Token Claims registries, with the semantics of
        <xref target="eat"/>. Until then, implementations use a private
        claim key.</t>
      </section>
    </section>
  </middle>

  <back>
    <references>
      <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 initials="S." surname="Bradner" fullname="S. Bradner"><address><email>deonte@goodshyt.fun</email></address></author>
            <date year="1997" month="March"/>
          </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 initials="B." surname="Leiba" fullname="B. Leiba"><address><email>deonte@goodshyt.fun</email></address></author>
            <date year="2017" month="May"/>
          </front>
          <seriesInfo name="BCP" value="14"/>
          <seriesInfo name="RFC" value="8174"/>
          <seriesInfo name="DOI" value="10.17487/RFC8174"/>
        </reference>
        <reference anchor="RFC9334" target="https://www.rfc-editor.org/info/rfc9334">
          <front>
            <title>Remote ATtestation procedureS (RATS) Architecture</title>
            <author initials="H." surname="Birkholz" fullname="H. Birkholz"><address><email>deonte@goodshyt.fun</email></address></author>
            <author initials="D." surname="Thaler" fullname="D. Thaler"><address><email>deonte@goodshyt.fun</email></address></author>
            <author initials="M." surname="Richardson" fullname="M. Richardson"><address><email>deonte@goodshyt.fun</email></address></author>
            <author initials="N." surname="Smith" fullname="N. Smith"><address><email>deonte@goodshyt.fun</email></address></author>
            <author initials="W." surname="Pan" fullname="W. Pan"><address><email>deonte@goodshyt.fun</email></address></author>
            <date year="2023" month="January"/>
          </front>
          <seriesInfo name="RFC" value="9334"/>
          <seriesInfo name="DOI" value="10.17487/RFC9334"/>
        </reference>
        <reference anchor="RFC9711" target="https://www.rfc-editor.org/info/rfc9711">
          <front>
            <title>The Entity Attestation Token (EAT)</title>
            <author initials="L." surname="Lundblade" fullname="L. Lundblade"><address><email>deonte@goodshyt.fun</email></address></author>
            <author initials="G." surname="Mandyam" fullname="G. Mandyam"><address><email>deonte@goodshyt.fun</email></address></author>
            <author initials="J." surname="O'Donoghue" fullname="J. O'Donoghue"><address><email>deonte@goodshyt.fun</email></address></author>
            <author initials="C." surname="Wallace" fullname="C. Wallace"><address><email>deonte@goodshyt.fun</email></address></author>
            <date year="2025" month="April"/>
          </front>
          <seriesInfo name="RFC" value="9711"/>
          <seriesInfo name="DOI" value="10.17487/RFC9711"/>
        </reference>
        <reference anchor="RFC8259" target="https://www.rfc-editor.org/info/rfc8259">
          <front>
            <title>The JavaScript Object Notation (JSON) Data Interchange Format</title>
            <author initials="T." surname="Bray" fullname="T. Bray" role="editor"><address><email>deonte@goodshyt.fun</email></address></author>
            <date year="2017" month="December"/>
          </front>
          <seriesInfo name="STD" value="90"/>
          <seriesInfo name="RFC" value="8259"/>
          <seriesInfo name="DOI" value="10.17487/RFC8259"/>
        </reference>
        <reference anchor="FIPS180-4" target="https://doi.org/10.6028/NIST.FIPS.180-4">
          <front>
            <title>Secure Hash Standard (SHS)</title>
            <author><organization>National Institute of Standards and Technology</organization><address><email>deonte@goodshyt.fun</email></address></author>
            <date year="2015" month="August"/>
          </front>
          <seriesInfo name="FIPS PUB" value="180-4"/>
        </reference>
      </references>
      <references>
        <name>Informative References</name>
        <reference anchor="TPM2" target="https://trustedcomputinggroup.org/resource/tpm-library-specification/">
          <front>
            <title>Trusted Platform Module Library Specification, Family "2.0"</title>
            <author><organization>Trusted Computing Group</organization><address><email>deonte@goodshyt.fun</email></address></author>
            <date year="2019"/>
          </front>
        </reference>
        <reference anchor="TVLA">
          <front>
            <title>A testing methodology for side-channel resistance validation</title>
            <author initials="G." surname="Goodwill" fullname="G. Goodwill"><address><email>deonte@goodshyt.fun</email></address></author>
            <author initials="B." surname="Jun" fullname="B. Jun"><address><email>deonte@goodshyt.fun</email></address></author>
            <author initials="J." surname="Jaffe" fullname="J. Jaffe"><address><email>deonte@goodshyt.fun</email></address></author>
            <author initials="P." surname="Rohatgi" fullname="P. Rohatgi"><address><email>deonte@goodshyt.fun</email></address></author>
            <date year="2011" month="September"/>
          </front>
          <refcontent>NIST Non-Invasive Attack Testing Workshop</refcontent>
        </reference>
      </references>
    </references>

    <section anchor="impl" numbered="false">
      <name>Implementation Status</name>
      <t>[RFC Editor: remove before publication.]</t>
      <t>The ai-id-engine project implements this profile in JSON form:</t>
      <ul>
        <li>"aiid evidence create" and "aiid evidence verify" produce and
        check reports;</li>
        <li>TPM 2.0 signing goes through tpm2-tools, with a software-key
        fallback;</li>
        <li>every TVLA run in continuous integration produces a report with
        a software root;</li>
        <li>unit tests cover tampering with behavior claims, upgrading the
        provenance after signing, missing provenance, wrong classes and
        duplicate fields.</li>
      </ul>
      <t>The EAT and COSE encoding of <xref target="eat"/> is not yet
      implemented.</t>
    </section>
  </back>
</rfc>