Updated by fix-consistency on 2026-09-06:
- workplan status: ready → active
Assistant: claude-code
Assistant-Model: opus
Assistant-Process: 713962@bnt-lap001
Assistant-Session: 2718d99d-d3ff-478f-83a2-3a30f01a02fc
glas-harness reports approval-engine production startup now requires audit
delivery, putting sender registration on the deployment critical path. The
task's note read the dependency backwards: audit-core is not waiting for a
producer to emit, the producer is waiting for a sender identity.
T03 is promoted with it. §9.4 makes approval evidence load-bearing and there
is nowhere to record that until evidence_kind exists; registering ahead of T03
would admit a load-bearing source into a system that would silently treat it
as attributive, which is the defect §9.6 names.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_0185wifnLzCxjEY2MT1XbK7L
Assistant: claude-code
Assistant-Model: opus
Assistant-Process: 713962@bnt-lap001
Assistant-Session: 2718d99d-d3ff-478f-83a2-3a30f01a02fc
§9.4 forbids audit-core answering whether an approval is still valid: a
consumer branching on such an answer would route an authorization decision
through the audit fabric. The prohibition was honoured by absence, which is
not the estate's idiom — §6.4 obligation 3 requires a published stance to
equal shipped behaviour asserted by test.
Written against shapes rather than today's route list, so adding a validity
surface later fails here rather than passing quietly. Plausible verdict paths
must 404 rather than 403: a distinguishable forbidden would imply a surface
exists behind auth. No backend class, Postgres included, may carry a
verdict-shaped method name. The declaration must stay in layer.yaml,
INTENT.md and SCOPE.md.
The last test guards the opposite error — an approval-shaped event class is
still ingestible, because the prohibition bounds the verdict, not the record.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_0185wifnLzCxjEY2MT1XbK7L
Assistant: claude-code
Assistant-Model: opus
Assistant-Process: 713962@bnt-lap001
Assistant-Session: 2718d99d-d3ff-478f-83a2-3a30f01a02fc
The Postgres backend returned tamper_evidence=True as a constant while
docs/integrity.md permits the claim only when a live external chain-head
attestation exists. The one attestation on record is 2026-08-16 and no job
renews it, so audit-core was telling every sender it had a property whose
precondition was unverified — the §9.6 defect it twice corrected in
gate-house's doctrine, turned inward.
evaluate_tamper_evidence() derives the flag from the chain report and the
mounted attestation, distinguishing seven states. Absence, staleness,
mismatch, an undated or unreadable attestation, a chain break, and an
unwalkable chain all degrade the claim rather than leave it standing.
Unreadable is treated as absent on purpose: a malformed file must not hold
up a claim a missing file would drop.
The freshness window is 168h against an intended daily cadence — seven
cadences, so a handful of missed runs degrade the claim rather than a single
one flapping it. Window and cadence are one contract in docs/integrity.md.
Production /readyz will now report tamper_evidence: false until
AUDIT-WP-0009-T02 schedules attestation. The claim was already false; it now
says so.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_0185wifnLzCxjEY2MT1XbK7L
Assistant: claude-code
Assistant-Model: opus
Assistant-Process: 713962@bnt-lap001
Assistant-Session: 2718d99d-d3ff-478f-83a2-3a30f01a02fc
The standard is accepted at v0.7 and all three of audit-core's v0.6
findings landed in it (§9.6 threat decomposition, cadence MUST for
load-bearing sources with reconciliation/heartbeat for low-volume
classes, §3.3's Evidence row restated as an estate trade).
INTENT.md: layer/role declared in frontmatter as §11 and companion §2
require — layer.yaml alone did not discharge it. Layer section rewritten
for the Evidence role and its obligations. New Evidence Bound section
carrying the §9.6 sound/unsound forms and the three-row threat table,
including the residual nothing in the model prevents.
SCOPE.md: replaced the statehub register stub, which carried no boundary
at all. Statute-fixed prohibitions now live here, separated from the
merely-not-yet — §16 ruled the stronger-custody gap closed, so WORM and
data.archive are not ours rather than not yet.
Assessment found nine gaps. Headline: postgres_backend returns
tamper_evidence=True unconditionally while docs/integrity.md permits it
only against a live external attestation, and the one on record is
2026-08-16 with no job renewing it — audit-core overclaiming its own
bound, the §9.6 defect turned inward. Also: no cadence, heartbeat,
reconciliation, or load-bearing classification exists, so the obligation
audit-core argued up from SHOULD to MUST is not yet dischargeable
against audit-core.
AUDIT-WP-0009 raised, ten tasks.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01WpeL68AWHqtqPQZEXY5kFe
Assistant: claude-code
Assistant-Model: opus
Assistant-Process: 4040362@bnt-lap001
Assistant-Session: 4fd0fd24-2ee8-4413-bd67-43bd79ca73f1
