informed-decision/layer.yaml

215 lines
9.8 KiB
YAML
Raw Permalink Normal View History

Declare the layer per GH-DEC-2026-012; close T02 Gate House ruled all three questions within a day, attributing the speed to the request being filed before the architecture with candidate answers and their costs. R1 PEP-shaped, confirmed as proposed. The ruling settles the shape; the layer stays ours to declare, so layer.yaml is written in this repository's voice rather than transcribed from the reply. R2 yes to a presentation claim, no second catalog row, under three limits now declared in layer.yaml and tested. Limit 2 — the claim must never be an input to the decision it presents for — is load-bearing: our self-dealing argument was accepted because it holds, not despite it. Limit 3 drives architecture, since here the actor being audited and the evidence source are the same component. R3 (b) with the authority rule: binding digest authoritative for what the request is, view_hash only for what was shown, neither substitutable, and a disagreement between them is a finding against the presenting surface rather than a fact about the request. Linkage is co-reference; nesting was refused because it reproduces the GH-DEC-2026-008 hash cycle. Built to v0.8 obligation 3 rather than migrating later: axis enumerated, unknown resolves to fail_closed, absent distinguishable from unknown in the record, and published-equals-shipped asserted by test rather than claimed. Every stance is fail_closed, which is a conclusion not a shortcut — ops-warden can justify fail_open on a continuity argument that does not exist here. GH-DEC-2026-010 inherited as a declared gap in four documents: a decision cannot today be proven to have come from access-engine. The decision path must not be described as validated while FLEX-WP-0024 is open. 46 tests pass. T05 and T07 unblocked. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01V3W1dQG7GFFM9d94jFx7iR Assistant: claude-code Assistant-Model: opus Assistant-Process: 1565372@bnt-lap001 Assistant-Session: 16bb2f25-b34c-49ef-8e94-5fec3567a568
2026-09-09 22:25:35 +02:00
# informed-decision — NetKingdom security layer declaration
#
# Framework: net-kingdom/canon/standards/security-layer-model_v0.7.md
# Companion: net-kingdom/SECURITY-COMPANION.md v0.2
# Voice: INTENT.md (this repository's own, per §11 "who must declare")
# Ruling: GH-DEC-2026-012 (gate-house@0a1d1d9) answered INFD-IN-0001
#
# Reference form: ops-warden's, adopted by audit-core and kings-guard, with
# kings-guard's adaptation for a repository with no Tooling contacts.
#
# GH-DEC-2026-012 R1 confirmed the SHAPE. The layer is declared here, in this
# repository's own voice, because a layer someone else states about you is not
# a declaration.
schema_version: "0.1"
framework: netkingdom-security-layer-model
standard_version: "0.7"
companion_version: "0.2"
repository: informed-decision
layer: surface
role: pep-shaped
declared_by: INTENT.md
declared_at: "2026-09-09"
ruling: GH-DEC-2026-012
# §6.4 — informed-decision is PEP-shaped: it causes a protected side effect on
# the far side of a decision (recording an approver entry against an approval
# object). Companion §5 is owed and §6.4 applies in full.
#
# Built to v0.8 obligation 3, not v0.7, per GH-DEC-2026-011 — see pep-stance.yaml.
pep_stance: pep-stance.yaml
protected_action: "Approver entry recorded against an approval object (POST /v1/approvals/{id}/entries)"
decision_engine: access-engine
# §6 — no repository other than access-engine exposes an authorization decision.
# This surface renders a question and records a human's answer. A disposition is
# evidence of an act, never a verdict.
decision_surfaces_exposed: none
# §3.3 / GH-DEC-2026-012 R2 — YES to a presentation claim, and NO second catalog
# row: PEP and PIP are shapes a repository has; §4 records the layers it
# occupies. The permission carries three limits, and they are the substance of
# it rather than caveats on it.
presentation_claim:
emitted: true
carries: presentation-only
limits:
- id: L1-presentation-only
rule: >-
The claim carries presentation and nothing else. It MUST NOT carry,
restate, summarise or imply the decision, the verdict, or whether the
act was permitted. A consumer learns from it only what was SHOWN, never
what was DECIDED.
- id: L2-not-an-input
rule: >-
The claim MUST NOT be an input to the decision it presents for. A policy
reading view_hash to decide whether an act is permitted would let the
presenting surface contribute to its own authorization.
note: >-
Load-bearing, not a formality. GH-DEC-2026-012 accepted this
repository's argument that a renderer attesting its own rendering is not
the self-dealing that kept the approval object out of access-engine —
but only because this limit holds. Without it the two collapse into the
same failure.
- id: L3-independent-evidence-path
rule: >-
The evidence copy reaches audit-core INDEPENDENTLY of this repository.
The claim endpoint and the evidence path are different things and
neither substitutes for the other. The copy that is evidence MUST NOT be
reachable only through the party it is evidence about.
note: >-
The limit that matters most here: audit evidence is protected from the
actor being audited, and in this component the actor and the source are
the same. Architecture consequence, tracked in
docs/specs/ArchitectureBlueprint.md.
# §17 — the shared request-claim schema is still unowned. This repository
# publishes at its own boundary and yields to that schema when it exists.
# Position accepted by GH-DEC-2026-012 and matching approval-engine's in
# APPROVAL-IN-0001.
request_claim_schema:
status: unowned-upstream
local_shape: published-at-own-boundary
yields_to: taxonomy-request-claim-schema
# GH-DEC-2026-012 R3 — (b), with the authority rule written down.
binding_digest_relationship:
ruling: GH-DEC-2026-012
view_hash_authoritative_for: what-was-shown
binding_digest_authoritative_for: what-the-request-is
binding_digest_owner: approval-engine
substitutable: false
disagreement_is: >-
A finding against the presenting surface, never a fact about the request.
Activate nesting per GH-DEC-2026-015: view_hash carries binding.digest approval-engine met the condition. Verified here rather than taken on report: their docs/approval-claim.md carries "Presentation exclusion — GH-DEC-2026-015 §4" in normative language, and I ran tests/test_claim_contract.py::test_presentation_changes_cannot_change_the_approved_act myself — 1 passed. That test pins the digest input set from BOTH sides, and the narrowing half is what makes it real: without it a digest over four fields, or over a constant, would pass the widening half perfectly. view_hash now carries binding.digest and the act-scope is no longer independently canonicalized here, so the act has exactly one canonicalization computed by the layer that owns it. approval_binding_digest is validated for shape and refused without its approval id — it is carried, never computed. The three published vectors are unchanged: they do not carry the new key, so pick omits it. Asserted, not assumed. The cycle condition did not disappear, its protection moved — from refusing nesting to approval-engine's normative exclusion. layer.yaml carries it as cycle_condition with a test, so a future widening meets a rule rather than silence. One thing not assumed. Both gate-house and approval-engine said our binding slice canonicalizes principal and target, two of their five fields. target plainly is act material and is now dropped. But their principal is the party ON WHOSE BEHALF the approval was issued, while ours is the person being BOUND — the approver. Different roles, and dropping ours would remove who was shown this from view_hash and gut the promise. Kept it, declared principal_role_overlap open in layer.yaml, tested that changing the approver still moves view_hash, and raised it rather than silently resolving it either way. L0/L2 are unaffected: with no approval there is no digest to defer to, and test_act_scope_still_binds_when_there_is_no_carried_digest pins that. 100 tests pass. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01V3W1dQG7GFFM9d94jFx7iR Assistant: claude-code Assistant-Model: opus Assistant-Process: 1565372@bnt-lap001 Assistant-Session: 16bb2f25-b34c-49ef-8e94-5fec3567a568
2026-09-10 20:58:22 +02:00
linkage: nesting
Declare the layer per GH-DEC-2026-012; close T02 Gate House ruled all three questions within a day, attributing the speed to the request being filed before the architecture with candidate answers and their costs. R1 PEP-shaped, confirmed as proposed. The ruling settles the shape; the layer stays ours to declare, so layer.yaml is written in this repository's voice rather than transcribed from the reply. R2 yes to a presentation claim, no second catalog row, under three limits now declared in layer.yaml and tested. Limit 2 — the claim must never be an input to the decision it presents for — is load-bearing: our self-dealing argument was accepted because it holds, not despite it. Limit 3 drives architecture, since here the actor being audited and the evidence source are the same component. R3 (b) with the authority rule: binding digest authoritative for what the request is, view_hash only for what was shown, neither substitutable, and a disagreement between them is a finding against the presenting surface rather than a fact about the request. Linkage is co-reference; nesting was refused because it reproduces the GH-DEC-2026-008 hash cycle. Built to v0.8 obligation 3 rather than migrating later: axis enumerated, unknown resolves to fail_closed, absent distinguishable from unknown in the record, and published-equals-shipped asserted by test rather than claimed. Every stance is fail_closed, which is a conclusion not a shortcut — ops-warden can justify fail_open on a continuity argument that does not exist here. GH-DEC-2026-010 inherited as a declared gap in four documents: a decision cannot today be proven to have come from access-engine. The decision path must not be described as validated while FLEX-WP-0024 is open. 46 tests pass. T05 and T07 unblocked. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01V3W1dQG7GFFM9d94jFx7iR Assistant: claude-code Assistant-Model: opus Assistant-Process: 1565372@bnt-lap001 Assistant-Session: 16bb2f25-b34c-49ef-8e94-5fec3567a568
2026-09-09 22:25:35 +02:00
linkage_rule: >-
Activate nesting per GH-DEC-2026-015: view_hash carries binding.digest approval-engine met the condition. Verified here rather than taken on report: their docs/approval-claim.md carries "Presentation exclusion — GH-DEC-2026-015 §4" in normative language, and I ran tests/test_claim_contract.py::test_presentation_changes_cannot_change_the_approved_act myself — 1 passed. That test pins the digest input set from BOTH sides, and the narrowing half is what makes it real: without it a digest over four fields, or over a constant, would pass the widening half perfectly. view_hash now carries binding.digest and the act-scope is no longer independently canonicalized here, so the act has exactly one canonicalization computed by the layer that owns it. approval_binding_digest is validated for shape and refused without its approval id — it is carried, never computed. The three published vectors are unchanged: they do not carry the new key, so pick omits it. Asserted, not assumed. The cycle condition did not disappear, its protection moved — from refusing nesting to approval-engine's normative exclusion. layer.yaml carries it as cycle_condition with a test, so a future widening meets a rule rather than silence. One thing not assumed. Both gate-house and approval-engine said our binding slice canonicalizes principal and target, two of their five fields. target plainly is act material and is now dropped. But their principal is the party ON WHOSE BEHALF the approval was issued, while ours is the person being BOUND — the approver. Different roles, and dropping ours would remove who was shown this from view_hash and gut the promise. Kept it, declared principal_role_overlap open in layer.yaml, tested that changing the approver still moves view_hash, and raised it rather than silently resolving it either way. L0/L2 are unaffected: with no approval there is no digest to defer to, and test_act_scope_still_binds_when_there_is_no_carried_digest pins that. 100 tests pass. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01V3W1dQG7GFFM9d94jFx7iR Assistant: claude-code Assistant-Model: opus Assistant-Process: 1565372@bnt-lap001 Assistant-Session: 16bb2f25-b34c-49ef-8e94-5fec3567a568
2026-09-10 20:58:22 +02:00
view_hash CARRIES approval-engine's binding.digest as a field. That digest
is referenced, never recomputed or restated from this repository's own
vocabulary, and the act-scope is no longer independently canonicalized here
— so the act has exactly one canonicalization, computed by the layer that
owns it.
Apply GH-DEC-2026-015/016 and the audit-core registration; both blockers cleared Origin and evidence path both landed today. T07 origin: railiance-apps deployed decisions.coulomb.social and corrected the hostname in this repo — not the decide.coulomb.social this workplan proposed. Verified here rather than taken on report: both paths 200, TLS verify 0, Let's Encrypt cert valid to 2026-12-09. T08: audit-core registered the source with every field as proposed and landed the detection half. INFD-IN-0003 closed. Their refinements booked — reconciliation on the high-volume class too, since rate detects a stream stopping but never a stream missing the particular renders that mattered, which is exactly our threat model; and PR-12, the custody locator must be a stable non-secret identifier because redact scans data and an existence declaration arriving without its pointer looks complete while being useless. INFD-IN-0004 ruled as GH-DEC-2026-015: gate-house reversed itself and nesting is permitted for this pair. The decisive ground was not the cycle argument we led with — our binding slice canonicalizes principal and target, two of the five digest fields, so co-reference left us performing a partial recomputation of one act in a second vocabulary, closer to the translation R3 forbade than nesting is. Our ordering objection was withdrawn as mistaken. The permission is conditioned and NOT ACTIVE until approval-engine states its presentation exclusion as normative and tested. layer.yaml is deliberately unchanged and carries nesting_permission_active false — we do not activate on our own initiative. GH-DEC-2026-016 ruled NC-03. Its §5 is live rather than hypothetical and is booked as PR-11: principal_type: human is a property of the client registration, structurally the same shape as the gap-route tenant, so a human-in-the-loop control must not be discharged on it as verified humanity. T07 stays progress: the submission to key-cape is written but unsent, blocked by the local permission classifier rather than by any repository. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01V3W1dQG7GFFM9d94jFx7iR Assistant: claude-code Assistant-Model: opus Assistant-Process: 1565372@bnt-lap001 Assistant-Session: 16bb2f25-b34c-49ef-8e94-5fec3567a568
2026-09-10 19:19:10 +02:00
# GH-DEC-2026-015 (INFD-IN-0004) re-ruled: nesting is PERMITTED for this pair,
Activate nesting per GH-DEC-2026-015: view_hash carries binding.digest approval-engine met the condition. Verified here rather than taken on report: their docs/approval-claim.md carries "Presentation exclusion — GH-DEC-2026-015 §4" in normative language, and I ran tests/test_claim_contract.py::test_presentation_changes_cannot_change_the_approved_act myself — 1 passed. That test pins the digest input set from BOTH sides, and the narrowing half is what makes it real: without it a digest over four fields, or over a constant, would pass the widening half perfectly. view_hash now carries binding.digest and the act-scope is no longer independently canonicalized here, so the act has exactly one canonicalization computed by the layer that owns it. approval_binding_digest is validated for shape and refused without its approval id — it is carried, never computed. The three published vectors are unchanged: they do not carry the new key, so pick omits it. Asserted, not assumed. The cycle condition did not disappear, its protection moved — from refusing nesting to approval-engine's normative exclusion. layer.yaml carries it as cycle_condition with a test, so a future widening meets a rule rather than silence. One thing not assumed. Both gate-house and approval-engine said our binding slice canonicalizes principal and target, two of their five fields. target plainly is act material and is now dropped. But their principal is the party ON WHOSE BEHALF the approval was issued, while ours is the person being BOUND — the approver. Different roles, and dropping ours would remove who was shown this from view_hash and gut the promise. Kept it, declared principal_role_overlap open in layer.yaml, tested that changing the approver still moves view_hash, and raised it rather than silently resolving it either way. L0/L2 are unaffected: with no approval there is no digest to defer to, and test_act_scope_still_binds_when_there_is_no_carried_digest pins that. 100 tests pass. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01V3W1dQG7GFFM9d94jFx7iR Assistant: claude-code Assistant-Model: opus Assistant-Process: 1565372@bnt-lap001 Assistant-Session: 16bb2f25-b34c-49ef-8e94-5fec3567a568
2026-09-10 20:58:22 +02:00
# conditioned on approval-engine stating its presentation exclusion as
# NORMATIVE and TESTED rather than design intent.
Apply GH-DEC-2026-015/016 and the audit-core registration; both blockers cleared Origin and evidence path both landed today. T07 origin: railiance-apps deployed decisions.coulomb.social and corrected the hostname in this repo — not the decide.coulomb.social this workplan proposed. Verified here rather than taken on report: both paths 200, TLS verify 0, Let's Encrypt cert valid to 2026-12-09. T08: audit-core registered the source with every field as proposed and landed the detection half. INFD-IN-0003 closed. Their refinements booked — reconciliation on the high-volume class too, since rate detects a stream stopping but never a stream missing the particular renders that mattered, which is exactly our threat model; and PR-12, the custody locator must be a stable non-secret identifier because redact scans data and an existence declaration arriving without its pointer looks complete while being useless. INFD-IN-0004 ruled as GH-DEC-2026-015: gate-house reversed itself and nesting is permitted for this pair. The decisive ground was not the cycle argument we led with — our binding slice canonicalizes principal and target, two of the five digest fields, so co-reference left us performing a partial recomputation of one act in a second vocabulary, closer to the translation R3 forbade than nesting is. Our ordering objection was withdrawn as mistaken. The permission is conditioned and NOT ACTIVE until approval-engine states its presentation exclusion as normative and tested. layer.yaml is deliberately unchanged and carries nesting_permission_active false — we do not activate on our own initiative. GH-DEC-2026-016 ruled NC-03. Its §5 is live rather than hypothetical and is booked as PR-11: principal_type: human is a property of the client registration, structurally the same shape as the gap-route tenant, so a human-in-the-loop control must not be discharged on it as verified humanity. T07 stays progress: the submission to key-cape is written but unsent, blocked by the local permission classifier rather than by any repository. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01V3W1dQG7GFFM9d94jFx7iR Assistant: claude-code Assistant-Model: opus Assistant-Process: 1565372@bnt-lap001 Assistant-Session: 16bb2f25-b34c-49ef-8e94-5fec3567a568
2026-09-10 19:19:10 +02:00
#
Activate nesting per GH-DEC-2026-015: view_hash carries binding.digest approval-engine met the condition. Verified here rather than taken on report: their docs/approval-claim.md carries "Presentation exclusion — GH-DEC-2026-015 §4" in normative language, and I ran tests/test_claim_contract.py::test_presentation_changes_cannot_change_the_approved_act myself — 1 passed. That test pins the digest input set from BOTH sides, and the narrowing half is what makes it real: without it a digest over four fields, or over a constant, would pass the widening half perfectly. view_hash now carries binding.digest and the act-scope is no longer independently canonicalized here, so the act has exactly one canonicalization computed by the layer that owns it. approval_binding_digest is validated for shape and refused without its approval id — it is carried, never computed. The three published vectors are unchanged: they do not carry the new key, so pick omits it. Asserted, not assumed. The cycle condition did not disappear, its protection moved — from refusing nesting to approval-engine's normative exclusion. layer.yaml carries it as cycle_condition with a test, so a future widening meets a rule rather than silence. One thing not assumed. Both gate-house and approval-engine said our binding slice canonicalizes principal and target, two of their five fields. target plainly is act material and is now dropped. But their principal is the party ON WHOSE BEHALF the approval was issued, while ours is the person being BOUND — the approver. Different roles, and dropping ours would remove who was shown this from view_hash and gut the promise. Kept it, declared principal_role_overlap open in layer.yaml, tested that changing the approver still moves view_hash, and raised it rather than silently resolving it either way. L0/L2 are unaffected: with no approval there is no digest to defer to, and test_act_scope_still_binds_when_there_is_no_carried_digest pins that. 100 tests pass. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01V3W1dQG7GFFM9d94jFx7iR Assistant: claude-code Assistant-Model: opus Assistant-Process: 1565372@bnt-lap001 Assistant-Session: 16bb2f25-b34c-49ef-8e94-5fec3567a568
2026-09-10 20:58:22 +02:00
# ACTIVATED 2026-09-10, after verifying the condition here rather than taking
# it on report: approval-engine/docs/approval-claim.md carries "Presentation
# exclusion — GH-DEC-2026-015 §4" in normative language, and
# tests/test_claim_contract.py::test_presentation_changes_cannot_change_the_approved_act
# pins the input set from BOTH sides — widening (presentation material leaves
# the digest unchanged) and narrowing (each of the five act fields changes it).
# The narrowing half matters: without it a digest over four fields, or over a
# constant, would pass the widening half perfectly. Run and confirmed passing.
Apply GH-DEC-2026-015/016 and the audit-core registration; both blockers cleared Origin and evidence path both landed today. T07 origin: railiance-apps deployed decisions.coulomb.social and corrected the hostname in this repo — not the decide.coulomb.social this workplan proposed. Verified here rather than taken on report: both paths 200, TLS verify 0, Let's Encrypt cert valid to 2026-12-09. T08: audit-core registered the source with every field as proposed and landed the detection half. INFD-IN-0003 closed. Their refinements booked — reconciliation on the high-volume class too, since rate detects a stream stopping but never a stream missing the particular renders that mattered, which is exactly our threat model; and PR-12, the custody locator must be a stable non-secret identifier because redact scans data and an existence declaration arriving without its pointer looks complete while being useless. INFD-IN-0004 ruled as GH-DEC-2026-015: gate-house reversed itself and nesting is permitted for this pair. The decisive ground was not the cycle argument we led with — our binding slice canonicalizes principal and target, two of the five digest fields, so co-reference left us performing a partial recomputation of one act in a second vocabulary, closer to the translation R3 forbade than nesting is. Our ordering objection was withdrawn as mistaken. The permission is conditioned and NOT ACTIVE until approval-engine states its presentation exclusion as normative and tested. layer.yaml is deliberately unchanged and carries nesting_permission_active false — we do not activate on our own initiative. GH-DEC-2026-016 ruled NC-03. Its §5 is live rather than hypothetical and is booked as PR-11: principal_type: human is a property of the client registration, structurally the same shape as the gap-route tenant, so a human-in-the-loop control must not be discharged on it as verified humanity. T07 stays progress: the submission to key-cape is written but unsent, blocked by the local permission classifier rather than by any repository. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01V3W1dQG7GFFM9d94jFx7iR Assistant: claude-code Assistant-Model: opus Assistant-Process: 1565372@bnt-lap001 Assistant-Session: 16bb2f25-b34c-49ef-8e94-5fec3567a568
2026-09-10 19:19:10 +02:00
reruled_by: GH-DEC-2026-015
nesting_permitted_when: >-
approval-engine states the presentation exclusion from binding.digest as
Activate nesting per GH-DEC-2026-015: view_hash carries binding.digest approval-engine met the condition. Verified here rather than taken on report: their docs/approval-claim.md carries "Presentation exclusion — GH-DEC-2026-015 §4" in normative language, and I ran tests/test_claim_contract.py::test_presentation_changes_cannot_change_the_approved_act myself — 1 passed. That test pins the digest input set from BOTH sides, and the narrowing half is what makes it real: without it a digest over four fields, or over a constant, would pass the widening half perfectly. view_hash now carries binding.digest and the act-scope is no longer independently canonicalized here, so the act has exactly one canonicalization computed by the layer that owns it. approval_binding_digest is validated for shape and refused without its approval id — it is carried, never computed. The three published vectors are unchanged: they do not carry the new key, so pick omits it. Asserted, not assumed. The cycle condition did not disappear, its protection moved — from refusing nesting to approval-engine's normative exclusion. layer.yaml carries it as cycle_condition with a test, so a future widening meets a rule rather than silence. One thing not assumed. Both gate-house and approval-engine said our binding slice canonicalizes principal and target, two of their five fields. target plainly is act material and is now dropped. But their principal is the party ON WHOSE BEHALF the approval was issued, while ours is the person being BOUND — the approver. Different roles, and dropping ours would remove who was shown this from view_hash and gut the promise. Kept it, declared principal_role_overlap open in layer.yaml, tested that changing the approver still moves view_hash, and raised it rather than silently resolving it either way. L0/L2 are unaffected: with no approval there is no digest to defer to, and test_act_scope_still_binds_when_there_is_no_carried_digest pins that. 100 tests pass. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01V3W1dQG7GFFM9d94jFx7iR Assistant: claude-code Assistant-Model: opus Assistant-Process: 1565372@bnt-lap001 Assistant-Session: 16bb2f25-b34c-49ef-8e94-5fec3567a568
2026-09-10 20:58:22 +02:00
normative and tested.
nesting_permission_active: true
nesting_activated_at: "2026-09-10"
nesting_condition_evidence:
doc: approval-engine/docs/approval-claim.md#presentation-exclusion
test: tests/test_claim_contract.py::test_presentation_changes_cannot_change_the_approved_act
evidence_record: approval-engine/docs/evidence/2026-09-10-presentation-exclusion.json
# OPEN, raised with approval-engine rather than assumed. Their `principal` is
# the party ON WHOSE BEHALF the approval was issued; ours is the person being
# BOUND — the approver. Different roles, so this repository still commits its
# own principal in view_hash. Dropping it would remove *who was shown this*
# and gut the promise. If the two are the same field, ours drops too.
principal_role_overlap: open
# The cycle condition remains the thing to protect, and it is now protected by
# approval-engine's normative exclusion rather than by refusing nesting.
cycle_condition: >-
Mutual containment. view_hash carries binding.digest; binding.digest MUST
NOT cover presentation material, so containment stays one-directional and
the GH-DEC-2026-008 cycle cannot arise. If that exclusion is ever relaxed,
this linkage must be revisited before the widening ships — a fail-closed
consumer obeying a cyclic claim denies permanently.
Declare the layer per GH-DEC-2026-012; close T02 Gate House ruled all three questions within a day, attributing the speed to the request being filed before the architecture with candidate answers and their costs. R1 PEP-shaped, confirmed as proposed. The ruling settles the shape; the layer stays ours to declare, so layer.yaml is written in this repository's voice rather than transcribed from the reply. R2 yes to a presentation claim, no second catalog row, under three limits now declared in layer.yaml and tested. Limit 2 — the claim must never be an input to the decision it presents for — is load-bearing: our self-dealing argument was accepted because it holds, not despite it. Limit 3 drives architecture, since here the actor being audited and the evidence source are the same component. R3 (b) with the authority rule: binding digest authoritative for what the request is, view_hash only for what was shown, neither substitutable, and a disagreement between them is a finding against the presenting surface rather than a fact about the request. Linkage is co-reference; nesting was refused because it reproduces the GH-DEC-2026-008 hash cycle. Built to v0.8 obligation 3 rather than migrating later: axis enumerated, unknown resolves to fail_closed, absent distinguishable from unknown in the record, and published-equals-shipped asserted by test rather than claimed. Every stance is fail_closed, which is a conclusion not a shortcut — ops-warden can justify fail_open on a continuity argument that does not exist here. GH-DEC-2026-010 inherited as a declared gap in four documents: a decision cannot today be proven to have come from access-engine. The decision path must not be described as validated while FLEX-WP-0024 is open. 46 tests pass. T05 and T07 unblocked. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01V3W1dQG7GFFM9d94jFx7iR Assistant: claude-code Assistant-Model: opus Assistant-Process: 1565372@bnt-lap001 Assistant-Session: 16bb2f25-b34c-49ef-8e94-5fec3567a568
2026-09-09 22:25:35 +02:00
# §5 applies to Staff. This is a browser-facing surface with no Tooling contact.
tooling_contacts: []
# §11 — record non-Tooling clients so the check is total.
non_tooling_clients: []
intended_non_tooling_clients:
- target: approval-engine
layer: engine
rationale: >-
GET /v1/approvals/{id} and /claim (approval:read) to render; POST
/v1/approvals/{id}/entries (approval:approve) to record a binding. Never
/consume. Requirements: approval-engine/docs/approver-surface-requirements.md.
- target: access-engine
layer: engine
rationale: >-
Decision consumed before rendering an approval to a person. A 200 from
approval-engine is not entitlement. This surface consumes a decision and
never renders one.
- target: key-cape
layer: engine
rationale: >-
Identity. Authorization-code + PKCE browser client. Identity is imported,
never invented here.
- target: audit-core
layer: engine
rationale: >-
Evidence destination for presentation records and dispositions. Must be an
independent path per limit L3-independent-evidence-path.
- target: state-hub
layer: not-catalogued
rationale: >-
Progress events. Outside §5 by the v0.5 scope rule. Recorded, not policed.
# §9.6 — presentation evidence is load-bearing: it is the only record of what a
# human was shown before binding. Atomicity and attestation cover accident and
# later tampering, never a compromised source.
evidence:
kind: load-bearing
residual: compromised-surface-presents-x-attests-y
residual_closed: false
custody: same-bound-as-every-other-source # §16 decided: no stronger archive
note: >-
GH-DEC-2026-012 states the residual is not closed in those words, and this
repository is not credited with closing it. Same disposition as
approval-engine's equivalent residual for adversarial omission at a
compromised source.
# INHERITED DECLARED GAP — GH-DEC-2026-010.
#
# Obligation 1 now requires a decision be ATTRIBUTABLE to access-engine. No
# consumer can satisfy that today: flex-auth's decision envelope is unsigned.
# This is a declared §13 gap tracked as FLEX-WP-0024, not a clean path this
# repository can walk.
#
# Stated here, and in SCOPE.md and ArchitectureBlueprint.md, because
# GH-DEC-2026-012 requires it be said in this repository's own documents rather
# than describing validation as complete.
inherited_gaps:
- id: GH-DEC-2026-010-attributability
obligation: 1
gap: >-
A decision consumed from access-engine cannot today be proven to have come
from access-engine — the envelope is unsigned.
tracked_by: FLEX-WP-0024
consequence_here: >-
This surface's record can show that a decision was obtained and what it
said. It cannot yet show it was access-engine that said it. Validation of
the decision path MUST NOT be described as complete while this is open.
status: open
declared_shapes:
"5.1": []
"5.2": []
"5.3": []