Why this node matters
in-toto/attestation#597 proposes a repository-owned conformance layer for every predicate. If that shape is accepted, later predicate families can inherit its verifier-interoperability contract.
This is a higher-leverage node than proposing another CounterProof-specific packet: the question is how independent verifiers prove they agree on a shared attestation specification.
CounterProof's narrow contribution
CounterProof has already hit the stale-evidence class in production-facing public claims:
historically valid evidence
!=
currently applicable evidence
The analogous conformance failure is:
passed corpus revision A
!=
conforms to current corpus revision B
A portable conformance receipt should bind:
- predicate type;
- specification revision or digest;
- conformance manifest digest;
- vector-set identity;
- comparison surface;
- verifier identity/version;
- per-vector observed verdict;
- evaluation timestamp.
When the corpus moves, the historical receipt remains valid for its pinned bytes but becomes stale as evidence for the new corpus until rerun.
Current upstream state observed 2026-10-07
Artifact
CounterProof PR #147 carries the first implementation of external-claim freshness plus:
docs/interop/in-toto-conformance-lifecycle.md
Next external handoff
When write access to the upstream thread is available, contribute the lifecycle boundary to #597 without proposing a new normative layer:
Passing repository-owned vectors demonstrates interoperability on the pinned corpus revision; it does not by itself establish conformance to later corpus revisions.
Success is upstream discussion/adoption of the boundary, not ownership of the predicate or conformance design.
Why this node matters
in-toto/attestation#597 proposes a repository-owned conformance layer for every predicate. If that shape is accepted, later predicate families can inherit its verifier-interoperability contract.
This is a higher-leverage node than proposing another CounterProof-specific packet: the question is how independent verifiers prove they agree on a shared attestation specification.
CounterProof's narrow contribution
CounterProof has already hit the stale-evidence class in production-facing public claims:
The analogous conformance failure is:
A portable conformance receipt should bind:
When the corpus moves, the historical receipt remains valid for its pinned bytes but becomes stale as evidence for the new corpus until rerun.
Current upstream state observed 2026-10-07
agent-decision/v0.1: open403 Resource not accessible by integration; this is a connector permission boundary, not upstream rejectionArtifact
CounterProof PR #147 carries the first implementation of external-claim freshness plus:
docs/interop/in-toto-conformance-lifecycle.mdNext external handoff
When write access to the upstream thread is available, contribute the lifecycle boundary to #597 without proposing a new normative layer:
Success is upstream discussion/adoption of the boundary, not ownership of the predicate or conformance design.