reason-codes: canonical KYE™ vocabulary

README.md

@@ -0,0 +1,34 @@
+---
+license: apache-2.0
+language: en
+pretty_name: KYE Protocol™ Reason Codes
+tags:
+- reason-codes
+- ai-governance
+- vocabulary
+- meaning-continuity
+- admissibility
+size_categories:
+- n<1K
+---
+
+# KYE Protocol™ — Reason Codes
+
+Canonical reason-code dictionary used across KYE™ decisions, drift
+events, obligations and recommendations. AI tooling that consumes
+KYE™ decisions can rely on these codes as a stable vocabulary.
+
+Includes:
+
+- **Decision values** (continuity, meaning, admissibility)
+- **Drift types** (constraint_loss, context_loss, memory_conflict,
+  timing_drift, handoff_drift, incentive_drift, state_transition_drift,
+  interpretation_drift, assumption_injection, meaning_compression,
+  meaning_expansion)
+- **Obligations** (notify_principal, request_clarification,
+  prevent_commit_until_reconfirmed, generate_evidence_pack, …)
+- **Inadmissibility classes** for the KYE Action Admissibility Profile™
+
+## License
+
+Apache-2.0. Source of truth: [github.com/KYE-Protocol](https://github.com/KYE-Protocol).

vocabulary/README.md

@@ -0,0 +1,35 @@
+# KYE™ Vocabulary
+
+The KYE™ vocabulary defines stable names for entities, relationships, actions, lifecycle states, obligations, data classes, and reason codes.
+
+| File | Contents |
+|---|---|
+| `entity-types.md` | Entity type names |
+| `relationships.md` | Relationship type names |
+| `actions.md` | Action names |
+| `lifecycle-states.md` | Lifecycle state names |
+| `obligations.md` | Obligation names |
+| `data-classes.md` | Data classification names |
+| `reason-codes.md` | Reason code names |
+| `continuity-discoverability.md` | Continuity decision values, drift types, discovery modes, risk-discovery types |
+| `ontology.md` | Ontology domains, predicates, mapping types, required objects, signals |
+| `operating-model.md` | Operating-model journey stages, gate types, commit-boundary archetypes, risk tiers, lifecycle states |
+| `assurance-card.md` | Assurance-card lifecycle stages, subject types, review triggers, human-involvement types, off-boarding actions, cascade-revocation scopes |
+| `formal-rules.md` | Formal-rule families, normative operators, decision outputs, permission/obligation/power/exception types, conflict types, resolution strategies |
+| `action-admissibility.md` | Admissibility decision values, inadmissibility classes, required checks, signals |
+
+These are the **public** vocabulary documents. They define names so that KYE artifacts can be written, exchanged, and discussed across implementations.
+
+The semantics — how names map to validation rules, policy evaluation, lifecycle enforcement, audit propagation, proof construction, and transparency submission — are part of the normative specification, which is published separately.
+
+## Stability
+
+Vocabulary names are intended to be stable. Changes follow this discipline:
+
+- additions are additive and backwards-compatible
+- renames are not permitted; new names supersede old names with explicit aliasing in the normative specification
+- removals require a deprecation cycle
+
+## Profile-specific vocabulary
+
+Profile-specific vocabulary (for example, payment actions, healthcare obligations, or treasury approval types) lives in profile documents. Where those profiles are released publicly, they will appear under their own directories. Until then, profile vocabulary is not published in this repository.

vocabulary/action-admissibility.md

@@ -0,0 +1,80 @@
+# KYE™ Action Admissibility vocabulary
+
+Names introduced by **KYE Action Admissibility Profile™**
+(`kye-action-admissibility-v1`) — the upstream pre-action layer.
+This document lists names only. Admissibility-scoring algorithms,
+prohibited-action-class detection heuristics, intent-ambiguity
+detection, source / data admissibility decision rules, agent-proposal
+quarantine triggers and sector admission packs are part of the patent
+track and are not published here.
+
+> **KYE™ does not only attribute delegated actions after they exist.
+> It checks whether proposed actions are admissible before they enter
+> the authority pipeline.**
+
+## Decision values
+
+| Value | Meaning |
+|---|---|
+| `admit`                     | Admissible. Route onward to authority decisioning. |
+| `reject`                    | Structurally invalid or prohibited. Do not enter pipeline. |
+| `require_clarification`     | Intent ambiguous. Request principal clarification first. |
+| `require_human_review`      | Admissible only after human reviewer admits. |
+| `quarantine`                | Pattern matches a watched class. Freeze and escalate. |
+| `route_to_authority_check`  | Synonym for `admit` — explicit routing instruction. |
+
+## Inadmissibility classes
+
+What admissibility actively detects:
+
+`invalid_intent` · `ambiguous_intent` · `out_of_scope_proposal` ·
+`prohibited_action_class` · `disallowed_data_source` ·
+`inadmissible_evidence` · `unsafe_tool_path` · `coercion_signal` ·
+`incentive_conflict` · `continuity_break` ·
+`policy_ineligible_action` · `missing_authority_context` ·
+`missing_principal` · `missing_accountable_owner` ·
+`unsupported_jurisdiction`.
+
+## Required objects
+
+| Object | Schema |
+|---|---|
+| Profile descriptor | `action-admissibility-profile.json` |
+| Proposed Action | `proposed-action.json` |
+| Admission Gate™ | `admission-gate.json` |
+| Admissibility Decision | `admissibility-decision.json` |
+| Admissibility Evidence | `admissibility-evidence.json` |
+
+## Required checks
+
+`intent_present` · `intent_clear` · `principal_resolved` ·
+`actor_resolved` · `accountable_owner_known` · `capability_known` ·
+`resource_known` · `data_source_allowed` · `prohibited_action_class` ·
+`continuity_preserved` · `formal_rule_block` · `evidence_ready` ·
+`jurisdiction_supported` · `risk_state_acceptable` ·
+`no_coercion_signal` · `no_incentive_conflict` ·
+`no_unsafe_tool_path`.
+
+## Reason codes
+
+See `reason-codes.md` under the **Action Admissibility profile reason
+codes** group (20 codes).
+
+## Admissibility signals
+
+`kye.admissibility.proposed_action_created` ·
+`kye.admissibility.check_started` ·
+`kye.admissibility.admitted` ·
+`kye.admissibility.rejected` ·
+`kye.admissibility.requires_clarification` ·
+`kye.admissibility.requires_human_review` ·
+`kye.admissibility.quarantined` ·
+`kye.admissibility.routed_to_authority_check` ·
+`kye.admissibility.evidence_generated`.
+
+## Patent-safe boundary
+
+Admissibility-scoring algorithm, prohibited-action-class detection
+heuristics, intent-ambiguity detection rules, source / data
+admissibility decision rules, agent-proposal quarantine triggers and
+sector admission packs are **not** in this repository.

vocabulary/actions.md

@@ -0,0 +1,93 @@
+# KYE™ Action Vocabulary
+
+Actions are stable identifiers for governed operations. This document lists action names recognized by KYE™.
+
+This document defines **names only**. Action semantics, side effects, and required obligations are part of the normative specification.
+
+## Entity lifecycle
+
+```
+entity.register
+entity.verify
+entity.suspend
+entity.stop
+entity.quarantine
+entity.revoke
+entity.archive
+entity.tombstone
+entity.transfer
+```
+
+## Identifiers and relationships
+
+```
+identifier.bind
+identifier.verify
+identifier.revoke
+relationship.create
+relationship.revoke
+```
+
+## Delegations and scopes
+
+```
+delegation.create
+delegation.approve
+delegation.suspend
+delegation.revoke
+scope.create
+scope.attenuate
+scope.revoke
+access_right.grant
+access_right.revoke
+```
+
+## Credentials and attestations
+
+```
+credential.issue
+credential.verify
+credential.present
+credential.revoke
+attestation.issue
+attestation.verify
+attestation.revoke
+workload.attest
+workload.bind
+```
+
+## Runtime actions (illustrative)
+
+```
+document.read
+document.render
+document.export
+record.read
+record.update
+tool.invoke
+model.invoke
+workflow.execute
+```
+
+## Audit, proof, and transparency
+
+```
+audit.export
+proof.generate
+transparency.submit
+transparency.verify
+```
+
+## Signals
+
+```
+signal.publish
+signal.subscribe
+signal.ack
+```
+
+## Naming rules
+
+- Action names are lowercase ASCII, dot-separated, in `noun.verb` form.
+- Action names are stable identifiers.
+- Profile-specific actions (for example, payment actions) are listed in the relevant profile dictionary.

vocabulary/assurance-card.md

@@ -0,0 +1,89 @@
+# KYE™ Assurance Card vocabulary
+
+Names introduced by **KYE Assurance Card Profile™**
+(`kye-assurance-card-v1`) — the lifecycle assurance layer that turns
+KYE™ runtime evidence into a living assurance record. This document
+lists names only. Assurance-card generation, runtime-evidence binding,
+review-automation, sector-pack content and use-case library
+recommendation algorithms are part of the patent track and are not
+published here.
+
+> **System cards describe the assurance case. KYE™ proves what
+> happened. KYE makes assurance executable.**
+
+## Lifecycle stages
+
+Eight ordered stages every assurance-card subject progresses through.
+
+`design` · `pilot` · `deploy` · `monitor` · `incident_review` ·
+`scope_change_review` · `retention_review` · `decommission`.
+
+## Required objects
+
+| Object | Schema |
+|---|---|
+| Profile descriptor | `assurance-card-profile.json` |
+| Assurance Card | `assurance-card.json` |
+| Human Involvement Plan | `human-involvement-plan.json` |
+| Provenance Evidence | `provenance-evidence.json` |
+| Assurance Review Cycle | `assurance-review-cycle.json` |
+| Decommissioning Plan | `decommissioning-plan.json` |
+
+## Subject types
+
+`ai_agent` · `workflow` · `service` · `model` · `tool` · `connector` ·
+`system`.
+
+## Lifecycle states
+
+`proposed` · `design` · `pilot` · `controlled_production` ·
+`production` · `deprecated` · `suspended` · `revoked`.
+
+## Review triggers
+
+Ten named triggers any review cycle MAY listen on:
+
+`scheduled` · `scope_change` · `new_capability` · `incident` ·
+`risk_state_change` · `model_update` · `authority_change` ·
+`supplier_change` · `licence_change` · `retention` · `decommission`.
+
+## Human involvement types
+
+`influence` · `direct` · `limit` · `approve` · `override` · `review`.
+
+## Human involvement stages
+
+`design` · `pilot` · `deploy` · `pre_commit` · `post_commit` ·
+`incident` · `scope_change` · `retention` · `decommission`.
+
+## Off-boarding actions
+
+`revoke_authority` · `quarantine_credentials` · `rotate_keys` ·
+`archive_evidence` · `notify_owner` · `notify_supplier` ·
+`update_catalog` · `remove_from_runtime`.
+
+## Cascade-revocation scopes
+
+`entity_only` · `entity_and_descendants` · `tenant_wide`.
+
+## Provenance verification methods
+
+`self_attested` · `third_party_attestation` · `signed_evidence` ·
+`registry_check`.
+
+## Reason codes
+
+See `reason-codes.md` under the **Assurance Card profile reason
+codes** group (17 codes).
+
+## Assurance signals
+
+See `kye-assurance-card-v1.md` §6. Five families: assurance-card
+lifecycle, human involvement, provenance, review cycle, decommissioning.
+
+## Patent-safe boundary
+
+The assurance-card generation engine, runtime-evidence binding
+heuristics, review-automation engine, sector-pack content (defence,
+public-sector, payments, health) and use-case library recommendation
+algorithm are **not** in this repository.

vocabulary/authority-finality.md

@@ -0,0 +1,50 @@
+# KYE Protocol™ — Authority Finality™ (vocabulary)
+
+> **Status:** Normative concept. v1.0. Apache 2.0.
+
+## Definition
+
+**Authority Finality™** is the *property* that, for any action taken
+by any entity governed by KYE Protocol™, the runtime can prove — with
+public-key-verifiable evidence — *who or what acted, on behalf of
+whom, using which capability, under what authority, in what state, and
+with what audit trail*.
+
+The structural artefact this property is asserted over is the
+**KYE Chain of Authority™** ([vocabulary](./chain-of-authority.md)) —
+the linked, attenuating delegation chain from root principal to actor.
+Authority Finality™ answers *is the chain final and provable?*;
+KYE Chain of Authority™ is the chain itself.
+
+Authority Finality™ is **technical and evidentiary**. It does not
+replace legal agreements, signatures, or regulatory obligations; it
+provides the substrate that lets those obligations be **proved** after
+the fact.
+
+## Six properties of an Authority-Final action
+
+A KYE-conformant gateway exposes, for every adjudicated action, the
+following six properties bound by a single `policy_decision_id`:
+
+| Property | Field | Source |
+|---|---|---|
+| Who or what is acting | `actor.entity_id` | `entity` profile |
+| On behalf of whom | top of `delegation_chain` | `delegation` profile |
+| Using which capability | `capability_id` (when invocation) | `capability` profile |
+| Under what authority | `delegation_id` + `scope_id` | `core` profile |
+| In what state | composed 6-tuple per `state-composition.json` | `state` profile |
+| With what audit trail | `audit_event_id` + `proof_bundle_id` | `audit` + `transparency` profiles |
+
+## Use in copy
+
+> KYE Protocol™ helps establish **Authority Finality™** for AI-agent
+> actions — creating a replayable evidence layer for accountability,
+> compliance, dispute resolution, and legally defensible audit trails.
+
+## Trademark
+
+**Authority Finality™** is a trademark of the KYE Protocol™ project.
+It identifies the property defined here. Conformant implementations
+may use the mark to claim the property; forks may not.
+
+— KYE Protocol™ project, 2026.

vocabulary/badge-levels.md

@@ -0,0 +1,71 @@
+# KYE Conformance Badge Levels (canonical)
+
+> **Status:** Normative. v1.0. Apache 2.0. Companion to
+> the `conformance-certification` rule pack + `badge-levels` dictionary (post-§29 landing).
+
+KYE distinguishes **conformant** (passes a test suite) from **certified**
+(reviewed + listed in the public registry). Six tiers plus four
+profile-scoped variants.
+
+## Trust hierarchy
+
+| Level | Badge                              | Awarded by             | Renewal  |
+|-------|------------------------------------|------------------------|----------|
+| L0    | KYE Declared / Compatible          | Self                   | n/a      |
+| L1    | **KYE Self-Tested™**               | Self (open suite)      | quarterly |
+| L2    | **KYE Self-Attested™**             | Self (signed)          | quarterly |
+| L3    | **KYE Conformant™** + variants     | Program (verified run) | annual |
+| L4    | **KYE Certified™**                 | Program + reviewer   | annual   |
+
+## L3 variants (issued separately)
+
+- KYE Core Conformant™
+- KYE Authority Conformant™
+- KYE Capability Conformant™
+- KYE Evidence Conformant™
+- KYE Recovery Conformant™
+- KYE Payments Conformant™
+- KYE Healthcare Conformant™
+- KYE Sector Profile Conformant™ *(parameterised — banking / energy / aviation / …)*
+- KYE EU AI Act Profile Ready™
+
+## Permitted commercial wording
+
+| Wording                  | Permitted? | Notes                                                |
+|--------------------------|-----------|------------------------------------------------------|
+| **KYE Conformant™**      | ✓          | Backed by a passing run against a published suite    |
+| **KYE Certified™**       | ✓          | Only after program review + signed registry record |
+| **KYE Self-Tested™**     | ✓          | Free badge after running the open conformance pack   |
+| **KYE Self-Attested™**   | ✓          | Free badge after a signed self-audit declaration     |
+| **KYE Ready™**           | ✓          | Soft go-to-market label for partner ecosystem        |
+| **Powered by KYE™**      | ✓          | Uses KYE APIs / SDKs (no certification claim)        |
+| KYE Compliant            | ⚠ avoid    | Reads as legal / regulatory; reserve for sector use  |
+
+## Required evidence per badge
+
+| Badge                 | Evidence required                                                |
+|-----------------------|------------------------------------------------------------------|
+| KYE Self-Tested™      | `conformance-run.json` (passed)                                  |
+| KYE Self-Attested™    | `self-audit-run.json` + `self-attestation.json` (signed)         |
+| KYE Conformant™ (L3)  | `conformance-run.json` (verified) + `certification-record.json`  |
+| KYE Certified™ (L4)   | L3 evidence + manual review + signed `certification-record.json` listed in registry |
+
+Every badge MUST be tied to: protocol version · profile-set version ·
+test-suite version · product / version · expiry · public verification
+URL · revocation status · signed certification record. Permanent
+badges are **not permitted**.
+
+## Use in copy
+
+> **KYE Conformance & Certification** lets customers verify that an
+> implementation correctly supports entity authority, state,
+> delegation, capability governance, decisions, audit trails,
+> evidence packs, and Decision Maps™.
+>
+> **Govern the governance layer.** KYE implementations can
+> continuously audit their own engines, schemas, profiles, registries,
+> policies, decisions, audit trails, evidence packs, and Decision
+> Maps™ — then publish signed self-attestations or pursue official
+> KYE certification.
+
+— KYE Protocol™ project, 2026.

vocabulary/chain-of-authority.md

@@ -0,0 +1,69 @@
+# KYE Protocol™ — KYE Chain of Authority™ (vocabulary)
+
+> **Status:** Normative concept. v1.0. Apache 2.0.
+
+## Definition
+
+**KYE Chain of Authority™** is the *structural artefact* of authority
+under KYE Protocol™: the linked, attenuating delegation chain that
+runs from the root principal down to the entity actually performing
+an action.
+
+Each link is a signed delegation that:
+
+- names the entity it delegates from and to,
+- carries a capability and a scope,
+- attenuates (never broadens) the parent's scope,
+- is bound to a state-vector at the moment of signing,
+- is hash-linked into the audit chain.
+
+The chain itself is the artefact. **Authority Finality™** is the
+*property* of that chain — that the chain is provably terminal,
+public-key-verifiable, and replayable. Two concepts, one structure.
+
+## Narrative parallel
+
+Chain-of-custody for *documents* is what courts already accept. KYE
+Chain of Authority™ is the equivalent for *actions*: a signed,
+linked, end-to-end record of who delegated what to whom, under what
+scope, in what state, with what audit trail.
+
+> KYE™ produces the KYE Chain of Authority™ the way digital
+> signatures produce chain-of-custody for documents. Courts already
+> accept signed documents; KYE™ gives them signed delegations.
+
+## How the chain is built
+
+| Position | Field | Source |
+|---|---|---|
+| Root | `principal_entity_id` of the topmost grant | `core` profile |
+| Each link | `delegation_id` with `parent_delegation_id`, `capability_id`, `scope_id`, `state_snapshot` | `delegation` profile |
+| Terminal | `actor.entity_id` | `entity` profile |
+| Verifiable | each link is COSE/JWS-signed; chain is hash-linked into the audit chain | `transparency` profile |
+
+## Use in copy
+
+> Before the bank authorises the payment, the gateway checks the
+> entire **KYE Chain of Authority™** — root principal → business →
+> agent → action — and emits the signed Decision Map™ + Evidence Pack™.
+
+## Trademark
+
+**KYE Chain of Authority™** is a trademark of the KYE Protocol™
+project. The KYE prefix is intentional: it mitigates a §2(e)
+descriptiveness refusal that would face the bare phrase
+"Chain of Authority" in USPTO / UKIPO / EUIPO filings.
+
+Conformant implementations may use the mark to refer to the artefact
+they produce. Forks may not.
+
+## Relationship to Authority Finality™
+
+| | KYE Chain of Authority™ | Authority Finality™ |
+|---|---|---|
+| Type | Structural artefact | Property / outcome |
+| Question it answers | *What is the chain?* | *Is the chain final and provable?* |
+| Plural? | Yes — many chains per gateway | No — single property of any chain |
+| Schema | Composed from `delegation` chain + `state-composition.json` | Asserted by the runtime over the composed chain |
+
+— KYE Protocol™ project, 2026.

vocabulary/continuity-discoverability.md

@@ -0,0 +1,119 @@
+# KYE™ Continuity + Discoverability vocabulary
+
+Names introduced by **KYE Continuity Profile™** (`kye-continuity-v1`) and
+**KYE Discoverability Profile™** (`kye-discoverability-v1`). This document
+lists names only. Detection algorithms, scoring weights, thresholds, and
+graph-traversal mechanisms are part of the normative specification and the
+patent track, and are not published here.
+
+## KYE Continuity Profile™
+
+### Continuity decision values
+
+| Value | Meaning |
+|---|---|
+| `continuity_preserved` | Interpreted intent matches declared intent; authority + state + scope intact; no material drift detected. |
+| `continuity_degraded` | One or more dimensions show drift; action proceeds with constraints + obligation list. |
+| `continuity_broken` | Material drift on at least one dimension; action MUST NOT proceed without re-anchoring. |
+
+The companion `continuity_score` is a numeric in `[0, 1]`. The mapping from
+score to decision value is part of the normative specification.
+
+### Drift types
+
+Ten named drift categories. Detection rules are not published.
+
+| Name | Surface |
+|---|---|
+| `intent_drift` | Interpreted goal diverges from declared goal. |
+| `authority_drift` | Authority record at decision time differs from authority at delegation time. |
+| `scope_drift` | Action lies outside intersected scope along the chain. |
+| `state_drift` | Authority / principal / actor state changed since grant. |
+| `capability_drift` | Capability invoked is not the capability granted. |
+| `execution_drift` | Runtime behaviour diverges from declared plan. |
+| `incentive_drift` | Pressure / incentive context conflicts with declared intent. |
+| `oversight_drift` | Required oversight chain is missing or stale. |
+| `evidence_drift` | Evidence references are missing, mutated, or unverifiable. |
+| `delegation_drift` | Delegation chain has been broken, attenuated past tolerance, or re-rooted. |
+
+### Reason codes (continuity)
+
+Continuity-status reason codes are listed in `reason-codes.md` under the
+**Continuity** group. Names follow the `<dimension>_drift_detected`
+convention (e.g., `intent_drift_detected`, `scope_drift_detected`).
+
+### Required objects
+
+| Object | Role |
+|---|---|
+| `KYEContinuityDecision` | The signed continuity verdict for a single decision point. |
+| `KYEAgencyDriftEvent` | One record per detected drift; hash-chained into the audit ledger. |
+| `Continuity Decision Map™` | The replayable artefact bound to a continuity decision. |
+| `Continuity Evidence Pack™` | Signed bundle composing intent + interpretation + state + pressure + incentive + decision + execution + drift events. |
+
+### Continuity signals
+
+`kye.intent.declared`, `kye.intent.drift_detected`,
+`kye.agency_drift.detected`, `kye.agency_drift.review_required`,
+`kye.agency_drift.resolved`, `kye.agency_drift.quarantined`.
+
+## KYE Discoverability Profile™
+
+### Discovery modes
+
+| Mode | Surface |
+|---|---|
+| `directory_lookup` | Resolve a known KYE URN to its current authority record + state. |
+| `path_finder` | Find the delegation path(s) from a principal to an actor for a capability. |
+| `graph_walk` | Traverse the authority graph under a discovery policy with row-level masking. |
+| `risk_discovery` | Non-mutating traversal returning the *would-result* set for a candidate action (stale / over-permissioned / pre-revocation blast radius). |
+| `connector_discovery` | Locate Connector Profile™ implementations + their conformance state. |
+| `evidence_finder` | Resolve audit-event + evidence-pack pointers bound to a decision. |
+
+### Risk discovery types
+
+| Type | Returns |
+|---|---|
+| `stale` | Authority records that have not been exercised within a profile-defined window. |
+| `over_permissioned` | Authority records whose granted scope materially exceeds observed exercise. |
+| `pre_revocation_blast_radius` | The downstream set that *would* be quarantined if a target authority were revoked. |
+
+The detection windows, scoring weights, and side-effect-level thresholds are
+part of the normative specification.
+
+### Required objects
+
+| Object | Role |
+|---|---|
+| `KYE Authority Directory™` | The directory index of cryptographically-bound authority records. |
+| `KYE Discovery Console™` | The operator surface over the directory + path finder + risk discovery. |
+| `KYE Authority Path Finder™` | The path-resolution engine over the authority graph. |
+| `KYE Evidence Finder™` | The resolver over audit-event + evidence-pack indexes. |
+| `KYE Connector Discovery Hub™` | The discovery surface for Connector Profile™ implementations. |
+
+### Reason codes (discovery)
+
+Discovery reason codes are listed in `reason-codes.md` under the
+**Discovery** group. Names follow the `discovery_<mode>_<outcome>`
+convention.
+
+### Discovery signals
+
+`kye.discovery.queried`, `kye.discovery.masked`,
+`kye.discovery.federation_traversed`, `kye.discovery.policy_denied`,
+`kye.risk.stale_detected`, `kye.risk.over_permissioned_detected`,
+`kye.risk.blast_radius_computed`.
+
+## Cross-profile composition
+
+KYE Continuity Profile™ and KYE Discoverability Profile™ MAY compose with
+any KYE Connector Profile™. The profiles do not redefine identity, scope
+attenuation, or audit-chain semantics; they extend the runtime decision
+surface and the discovery surface respectively.
+
+## Patent-safe boundary
+
+Algorithm content for drift detection, score computation, blast-radius
+traversal, intent-interpretation binding, and pre-revocation pruning is
+**not** in this repository. See the public landing pages for the profile
+overviews; counsel-controlled material remains private.

vocabulary/data-classes.md

@@ -0,0 +1,22 @@
+# KYE™ Data Classification Vocabulary
+
+Data classification labels are used in scopes, policies, and obligations to constrain what data classes an action may touch.
+
+This document defines **names only**. Tagging rules and policy semantics are part of the normative specification.
+
+| Class | Meaning |
+|---|---|
+| `public` | Public information |
+| `internal` | Internal to the organization |
+| `confidential` | Confidential information |
+| `restricted` | Restricted access required |
+| `secret` | Secret information |
+| `classified` | Classified under formal classification regimes |
+| `pii` | Personally identifiable information |
+| `ephi` | Electronic protected health information |
+| `cardholder_data` | Payment cardholder data |
+| `financial_data` | Financial data |
+| `operational_technology_data` | Operational technology data |
+| `defense_data` | Defense-related data |
+| `export_controlled_data` | Export-controlled data |
+| `safety_critical_data` | Safety-critical data |

vocabulary/entity-types.md

@@ -0,0 +1,177 @@
+# KYE™ Entity Types
+
+This document lists the entity types KYE Protocol™ recognizes. Type names are stable identifiers used in entity records, policy inputs, and dictionaries.
+
+This document defines **names only**. The lifecycle, validation rules, and attestation requirements for each type are part of the normative specification, which is published separately.
+
+## Hierarchy
+
+Every KYE™ entity resolves into one canonical containment tree (v3-final, locked 2026-05-14). The tree is parent-required at every node. Model, Tool, External App, and Audit Stream are **first-class Workspace-level entities** in v3 — they are not subtypes of Principal.
+
+```
+Tenant                                          (kye:tenant:…)    — billing + isolation root
+  ├── Legal Entity                              (kye:lglent:…)    — registered company / regulator-facing legal person
+  ├── Billing Account                           (kye:bill:…)      — Stripe customer + payment method + tax id
+  ├── Domains                                   (kye:dom:…)       — verified DNS domains, branded URLs
+  ├── Policies                                  (kye:pol:…)       — active Operating Model™, rule bundles, attestations
+  └── Workspaces                                (kye:wsp:…)        — environment / dataspace (prod, sandbox, region-specific)
+        ├── Projects                            (kye:prj:…)       — units of work; scoped budgets + quotas
+        ├── Teams                               (kye:team:…)      — RBAC + ownership group
+        ├── Resources                           (kye:res:…)       — D1 tables, R2 buckets, queues, connectors
+        ├── Principals                          (kye:prin:…)      — the acting parties (Human / System / Agent)
+        ├── Models                              (kye:model:…)     — inference artefacts; versioned
+        ├── Tools                               (kye:tool:…)      — named callables; declared side-effects
+        ├── External Apps                       (kye:extapp:…)    — 3rd-party integrations; partner connectors
+        └── Audit Streams                       (kye:audstream:…) — append-only event destinations (SIEM, S3, Kafka)
+```
+
+Each entity carries a `state` field governed by a registered state machine (see `public/vocabulary/state-machines.md`). Every entity also carries a `state_machine_id` identifying which machine governs it. Relationships between entities are explicit typed edges (see `public/vocabulary/relationships.md`).
+
+### Tenant-level siblings (5)
+
+| Sibling | Stable type | ID class | Required parent | Purpose |
+|---|---|---|---|---|
+| Legal Entity     | `legal_entity`    | `lglent`    | `tenant_id` | registered legal person; regulator-facing record (company number, country, jurisdiction). One Tenant MAY hold multiple Legal Entities (group structure, foreign subsidiaries). |
+| Billing Account  | `billing_account` | `bill`      | `tenant_id` | Stripe customer id, default payment method, billing email, tax id, currency. One Tenant MAY hold multiple Billing Accounts (split billing per business unit). |
+| Domains          | `domain`          | `dom`       | `tenant_id` | verified DNS domains the Tenant owns (`acme.com`, `acme-uk.com`). Used for branded sender, federation routing, SSO. |
+| Policies         | `policy_bundle`   | `pol`       | `tenant_id` | active Operating Model™, rule bundles, attestation policies, sub-processor list. One Tenant MAY hold multiple Policy Bundles (one per regime: PRA / DORA / EU AI Act). |
+| Workspaces       | `workspace`       | `wsp`       | `tenant_id` | environment / dataspace. One Tenant MAY hold many Workspaces (prod / sandbox / eu-west-1 / consumer-bank / corporate-bank). |
+
+### Workspace-level siblings (8)
+
+| Sibling | Stable type | ID class | Required parent | Purpose |
+|---|---|---|---|---|
+| Projects      | `project`      | `prj`       | `workspace_id` | unit of work; scoped budget + quota; one Operating Model fragment per Project |
+| Teams         | `team`         | `team`      | `workspace_id` | RBAC + ownership group; who-owns-what; bounded delegation surface; approval routing |
+| Resources     | `resource`     | `res`       | `workspace_id` | D1 tables, R2 buckets, queues, connectors that live inside the workspace. Resources are leaves — they don't contain other entities |
+| Principals    | `principal`    | `prin`      | `workspace_id`; `team_id` optional | the acting party — referenced as `actor_entity_id` / `principal_entity_id` in every Decision Map™ row |
+| Models        | `model`        | `model`     | `workspace_id` | an inference artefact (LLM, vision model, embedding model); carries `provider`, `family`, `version`, `weights_hash`. First-class in v3. |
+| Tools         | `tool`         | `tool`      | `workspace_id` | a named callable (function, MCP tool, API endpoint); carries `function_signature`, `declared_capabilities[]`, `side_effects[]`. First-class in v3. |
+| External Apps | `external_app` | `extapp`    | `workspace_id` | a 3rd-party integration or partner connector acting on behalf of the tenant; carries `partner_org_id`, `connector_kind`, `oauth_subject`. First-class in v3. |
+| Audit Streams | `audit_stream` | `audstream` | `workspace_id` | an append-only event-log destination the Workspace writes to (SIEM, S3, Splunk, Kafka topic); governed by an authority grant, not an IAM binding. New in v3. |
+
+### Principal subtypes (v3: 4 subtypes)
+
+In v3, `principal_class` has exactly four values — `human`, `system`,
+`agent`, and `external_app` (the schema enum on
+`kye.entity.principal.v1`, "reduced from six to four"). The v2 values
+`model` and `tool` are **deprecated** — promoted to standalone entity
+classes (`kye:model:…`, `kye:tool:…`). `external_app` is retained as a
+principal_class (a 3rd-party integration acting on behalf of the tenant)
+and ALSO exists as a standalone entity class (`kye:extapp:…`, see the
+table above). See `constitution/DEVIATIONS.md` for the migration window.
+
+Every Principal MUST resolve to exactly one subtype via its `principal_class` field:
+
+| Subclass | Examples | Subclass-specific fields |
+|---|---|---|
+| `human`  | person, employee, contractor, operator, approver, auditor, regulator, customer, patient, support_user | full_name, email_hash, idp_subject, role |
+| `system` | workload, service_account, api_client, serverless_function, container, vm, cluster, runtime, queue_consumer, webhook_endpoint, connector | service_principal_id, runs_in_workspace, owning_team_id |
+| `agent`  | ai_agent, agent_runtime | model_entity_id (→ `kye:model:…`), tool_entity_ids[] (→ `kye:tool:…`), operating_loop_kind |
+| `external_app` | 3rd-party integration / partner connector acting on behalf of the tenant | partner_org_id, connector_kind, oauth_subject |
+
+Note that in v3 an `agent` Principal references its Model via a
+`kye:model:…` entity ID and its Tools via `kye:tool:…` entity IDs —
+it does not embed them as subtypes.
+
+The hierarchy is enforced at schema time (`kye.entity.{tenant, legal_entity, billing_account, domain, policy_bundle, workspace, project, team, resource, principal, model, tool, external_app, audit_stream}.v1`) and at runtime by the PDP. KYE Onboarding Agent™ auto-creates a default Workspace + Team + admin Principal for every newly-granted Tenant so existing API keys, decisions, and audit rows always resolve to a valid parent chain.
+
+## Human entities
+
+| Type | Description |
+|---|---|
+| `person` | A natural person |
+| `employee` | A person employed by a legal entity |
+| `contractor` | A person engaged under a service contract |
+| `operator` | A person operating a system or workflow |
+| `approver` | A person empowered to approve actions |
+| `auditor` | A person performing audit review |
+| `regulator` | A regulatory authority representative |
+| `customer` | A consumer of a product or service |
+| `patient` | A patient in a healthcare context |
+| `support_user` | A support-context user |
+
+## Organizational entities
+
+| Type | Description |
+|---|---|
+| `business` | A commercial organization |
+| `legal_entity` | A registered legal entity |
+| `subsidiary` | A subsidiary of a parent company |
+| `parent_company` | A parent company |
+| `department` | A department within an organization |
+| `team` | A team within an organization |
+| `public_authority` | A public-sector authority |
+| `processor` | A data processor |
+| `subprocessor` | A sub-processor |
+| `supplier` | A supplier or vendor |
+| `relying_party` | A relying party in an identity exchange |
+
+## Technical entities
+
+| Type | Description |
+|---|---|
+| `workload` | A running computational unit |
+| `service_account` | A non-human service identity |
+| `api_client` | An API client |
+| `serverless_function` | A serverless function |
+| `container` | A container instance |
+| `vm` | A virtual machine |
+| `cluster` | A compute cluster |
+| `runtime` | A runtime environment |
+| `connector` | An integration connector |
+| `webhook_endpoint` | A webhook receiver |
+| `queue_consumer` | A message-queue consumer |
+
+## AI entities
+
+| Type | Description |
+|---|---|
+| `ai_agent` | An AI agent |
+| `agent_runtime` | An AI agent runtime |
+| `model` | A model |
+| `model_version` | A specific model version |
+| `prompt_template` | A prompt template |
+| `memory_store` | An agent memory store |
+| `vector_index` | A vector index |
+| `tool` | A tool or plugin |
+| `guardrail` | A guardrail component |
+| `evaluator` | An evaluator component |
+| `workflow` | A multi-step workflow |
+
+## Credential entities
+
+| Type | Description |
+|---|---|
+| `key` | A cryptographic key |
+| `api_key` | An API key |
+| `jwt` | A JWT-formatted credential |
+| `certificate` | An X.509 certificate |
+| `x509_svid` | A SPIFFE X.509 SVID |
+| `jwt_svid` | A SPIFFE JWT SVID |
+| `verifiable_credential` | A W3C verifiable credential |
+| `delegation_credential` | A delegation credential |
+| `scope_credential` | A scope credential |
+| `approval_credential` | An approval credential |
+
+## Data and resource entities
+
+| Type | Description |
+|---|---|
+| `dataset` | A dataset |
+| `database` | A database |
+| `table` | A database table |
+| `file` | A file |
+| `document` | A document |
+| `record` | A record |
+| `payment_instruction` | A payment instruction |
+| `clinical_record` | A clinical record |
+| `cardholder_object` | A cardholder data object (regulated) |
+| `evidence_pack` | A compliance evidence pack |
+| `policy_bundle` | A policy bundle |
+
+## Naming rules
+
++ Type names are lowercase ASCII with underscores.
++ Type names are stable identifiers and must not be reused for unrelated semantics.
++ Adding a new type requires a vocabulary proposal. Mechanism details are out of scope for this document.

vocabulary/formal-rules.md

@@ -0,0 +1,125 @@
+# KYE™ Formal Rules vocabulary
+
+Names introduced by **KYE Formal Rules Profile™**
+(`kye-formal-rules-v1`) — the rights, obligations and governance
+layer. This document lists names only. Rule-conflict-detection
+algorithms, rule-proof internals, rule-compilation engine, contract-
+to-authority extraction, semantic obligation mapping and sector-rule-
+pack content are part of the patent track and are not published here.
+
+> **A rule should not only be written. It should be enforceable,
+> discoverable, revocable, provable and replayable.**
+
+## Rule families
+
+Six normative families. Every `KYEFormalRule` declares exactly one.
+
+| Family | Operator | Meaning |
+|---|---|---|
+| `permission`  | `P`   | Subject MAY perform the action under stated authority + scope + state. |
+| `obligation`  | `O`   | Subject MUST perform the required action when the trigger condition applies. |
+| `prohibition` | `F`   | Subject MUST NOT perform the prohibited action under stated conditions. |
+| `power`       | `Pow` | Subject HAS authority to create, modify, revoke, waive or override a normative state. |
+| `immunity`    | `Imm` | Subject / state CANNOT be altered by the referenced actor or rule. |
+| `exception`   | `Ex`  | Rule is displaced or modified under enumerated exceptional conditions. |
+
+A seventh family `meta_governance` is recorded under
+`KYEGovernanceRule` and describes who may change rules, which rules
+dominate, and how conflicts resolve.
+
+## Normative operators
+
+Compact notation used inside `normative_effect.operator`:
+
+| Operator | Family | Meaning |
+|---|---|---|
+| `P`   | permission  | "may" |
+| `O`   | obligation  | "must" |
+| `F`   | prohibition | "must not" |
+| `Pow` | power       | "has authority to" |
+| `Imm` | immunity    | "cannot be altered by" |
+| `Ex`  | exception   | "displaced by, in conditions" |
+
+KYE™ does not commit to a specific deontic-logic syntax in the public
+surface. The operators above are the canonical product-friendly
+abbreviations.
+
+## Decision outputs
+
+Beyond Core's `{allow, deny, require_approval, require_step_up,
+quarantine}`, the Formal Rules Profile™ adds:
+
+`prohibited` · `obligation_created` · `obligation_satisfied` ·
+`obligation_breached` · `override_required` · `evidence_required` ·
+`human_review_required`.
+
+## Permission types
+
+`delegated_capability_permission` · `role_based_permission` ·
+`scope_based_permission` · `regulatory_permission` ·
+`contractual_permission`.
+
+## Obligation types
+
+`pre_execution_approval` · `post_action_review` ·
+`evidence_emission` · `notification` · `data_retention` ·
+`data_deletion` · `regulatory_report` · `remediation` · `custom`.
+
+## Obligation states
+
+`pending` · `satisfied` · `breached` · `waived` · `expired` ·
+`disputed` · `remediated` · `superseded`.
+
+## Power types
+
+`grant_power` · `revocation_power` · `modification_power` ·
+`waiver_power` · `override_power` · `delegation_power`.
+
+## Exception types
+
+`emergency_override` · `business_continuity` ·
+`regulatory_instruction` · `incident_response` ·
+`scheduled_maintenance` · `custom`.
+
+## Conflict types
+
+`permission_prohibition_conflict` · `obligation_circularity` ·
+`obligation_without_satisfaction_path` · `unbounded_obligation` ·
+`unrevocable_authority` · `circular_delegation` ·
+`missing_evidence_requirement` · `priority_tie` · `scope_overlap` ·
+`exception_overlap`.
+
+## Resolution strategies
+
+`specific_rule_overrides_general_rule` ·
+`prohibition_overrides_permission` · `later_rule_overrides_earlier` ·
+`higher_priority_wins` · `explicit_governance_override` ·
+`unresolved`.
+
+## Required objects
+
+| Object | Schema |
+|---|---|
+| Profile descriptor | `formal-rules-profile.json` |
+| Formal Rule | `formal-rule.json` |
+| Permission | `permission.json` |
+| Obligation | `obligation.json` |
+| Prohibition | `prohibition.json` |
+| Power | `power.json` |
+| Exception | `exception.json` |
+| Governance Rule | `governance-rule.json` |
+| Rule Conflict | `rule-conflict.json` |
+| Rule Proof | `rule-proof.json` |
+| Obligation State | `obligation-state.json` |
+
+## Reason codes
+
+See `reason-codes.md` under the **Formal Rules profile reason codes**
+group (33 codes).
+
+## Patent-safe boundary
+
+Rule-conflict-detection algorithm, rule-proof internals, rule-
+compilation engine, contract-to-authority extraction, semantic
+obligation mapping and sector-specific formal-rule packs are **not**
+in this repository.

vocabulary/graph-types-v1.md

@@ -0,0 +1,61 @@
+# KYE Protocol™ — Graph node, edge, and query dictionaries (V1)
+
+> **Status:** Normative. v1.0. Apache 2.0. Companion to
+> the `graph` dictionary + graph-traversal additions to the `runtime-authority` profile (post-§29 landing).
+>
+> KYE is **graph-first** because authority is relational. The
+> dictionaries below are the canonical values for `node_type`,
+> `edge_type`, and `graph_query_type` referenced by `graph-node.json`,
+> `graph-edge.json`, and `decision-map.json`. Specific traversal
+> algorithms are part of the patent track and are not described here.
+
+## 1. `node_type`
+
+`entity` · `human` · `organisation` · `ai_agent` · `api_client` ·
+`wallet` · `device` · `credential` · `capability` · `skill` · `tool` ·
+`workflow` · `prompt` · `model` · `resource` · `policy` ·
+`authority_grant` · `delegation` · `scope` · `state` ·
+`payload_artifact` · `decision` · `audit_event` · `evidence_pack` ·
+`compliance_control` · `taxonomy_term` · `metadata_schema`
+
+## 2. `edge_type`
+
+`owns` · `controls` · `maintains` · `issues` · `approves` ·
+`delegates_to` · `acts_on_behalf_of` · `grants_authority_to` ·
+`has_authority` · `authorises_capability` · `uses_capability` ·
+`depends_on` · `calls` · `accesses` · `bound_to_credential` ·
+`has_state` · `governed_by_policy` · `requires_approval_from` ·
+`evaluated_by` · `allowed_by` · `denied_by` · `produced_decision` ·
+`produced_audit_event` · `included_in_evidence_pack` ·
+`maps_to_framework_control` · `supersedes` · `revokes` ·
+`quarantines` · `recovers`
+
+## 3. `graph_query_type`
+
+`authority_path` · `delegation_path` · `capability_dependency` ·
+`decision_map` · `evidence_graph` · `risk_propagation` ·
+`blast_radius` · `compliance_mapping` · `lineage` · `ownership` ·
+`policy_coverage`
+
+## 4. Trademarked graph projections
+
+| Term                  | Score | Use                           |
+|-----------------------|-------|-------------------------------|
+| **Authority Graph™**  | 97    | Core KYE concept              |
+| **Decision Map™**     | 96    | Per-decision explainability   |
+| **Evidence Graph™**   | 95    | Decision ↔ evidence linkage   |
+| **Blast Radius Map™** | 94    | Compromise impact analysis    |
+| **Compliance Map™**   | 94    | Framework control projection  |
+
+## 5. Use in copy
+
+> **KYE is graph-first because authority is relational.** It maps
+> every entity, delegation, capability, credential, policy, state,
+> decision, and evidence object into a traversable Authority Graph™.
+>
+> Every KYE decision comes with a **Decision Map™** — a
+> replayable graph showing who or what acted, for whom, using which
+> capability, under what authority, in what state, and with what
+> evidence.
+
+— KYE Protocol™ project, 2026.

vocabulary/healthcare-break-glass-categories.md

@@ -0,0 +1,76 @@
+# KYE Protocol™ — Healthcare break-glass categories
+
+> **Status:** Normative. v1.0. Apache 2.0. Companion to
+> `internal` and `internal`.
+
+A break-glass grant in the Healthcare profile MUST carry a
+category-coded `reason_code` from the list below. Categories drive the
+post-hoc-review SLA, the audit retention class, and the regulator
+notification rule.
+
+## 1. Emergency clinical categories
+
+`healthcare.bg.emergency_treatment` —
+imminent threat to life or limb; care provider invoking break-glass
+without prior consent. Post-hoc review: ≤ 24 h.
+
+`healthcare.bg.unconscious_patient` —
+patient cannot give consent; authorised next-of-kin not yet reached.
+Post-hoc review: ≤ 48 h.
+
+`healthcare.bg.public_health_emergency` —
+declared public-health emergency (e.g. mass-casualty event) authorising
+broader access. Post-hoc review: ≤ 72 h; regulator notified.
+
+`healthcare.bg.psychiatric_hold` —
+involuntary hold under jurisdictional mental-health statute. Post-hoc
+review: ≤ 24 h; legal counsel CC'd.
+
+## 2. Care-coordination categories
+
+`healthcare.bg.care_handoff_emergency` —
+unplanned transfer of patient (e.g. interhospital). Post-hoc review:
+≤ 24 h.
+
+`healthcare.bg.dependent_care` —
+parent / guardian acting for a minor or dependent without contemporaneous
+consent. Post-hoc review: ≤ 7 days.
+
+## 3. Investigative categories
+
+`healthcare.bg.research_waiver` —
+research access under an IRB / ethics-board waiver; authority bound to
+the waiver attestation. Post-hoc review: ≤ 30 days.
+
+`healthcare.bg.fraud_investigation` —
+internal investigation of suspected fraud; authority bound to a
+compliance-officer attestation. Post-hoc review: ≤ 30 days.
+
+`healthcare.bg.law_enforcement_subpoena` —
+court-issued subpoena; authority bound to subpoena attestation.
+Post-hoc review: ≤ 30 days; legal counsel CC'd.
+
+`healthcare.bg.regulatory_audit` —
+regulator-issued audit demand. Post-hoc review: ≤ 30 days.
+
+## 4. Operational categories
+
+`healthcare.bg.system_recovery` —
+recovery of records during a system outage where normal authority is
+unreachable. Post-hoc review: ≤ 48 h.
+
+`healthcare.bg.data_subject_request_extraordinary` —
+extraordinary data-subject access request (e.g. emergency disclosure to
+the patient themselves). Post-hoc review: ≤ 7 days.
+
+## 5. Forbidden categories (MUST be rejected)
+
+`healthcare.bg.curiosity` ·
+`healthcare.bg.unauthorised_research` ·
+`healthcare.bg.commercial_disclosure` ·
+`healthcare.bg.media_disclosure`
+
+A break-glass request whose `reason_code` is in §5 MUST be rejected
+with `reason_code=break_glass_category_forbidden`.
+
+— KYE Protocol™ project, 2026.

vocabulary/lifecycle-states.md

@@ -0,0 +1,77 @@
+# KYE™ Lifecycle States
+
+Every KYE Protocol™ entity carries a `state` field governed by a
+registered state machine (see `public/vocabulary/state-machines.md`).
+This document lists the **canonical state names** that the core KYE
+vocabulary pre-defines.
+
+State transition rules, guard evaluation, atomicity requirements, and
+propagation behaviors are part of the normative specification and are
+not published in this repository.
+
+## State machine linkage
+
+Each entity record holds three state-related fields:
+
+| Field | Description |
+|---|---|
+| `state` | The current state value (one of the names below, or a custom name defined by the entity's machine) |
+| `state_machine_id` | URN of the registered state machine governing this entity (e.g., `kye:sm:kyeprotocol.com:core.principal.v1`) |
+| `state_machine_version` | Semver of the machine version in force |
+
+The full state event log lives in the append-only `state_events` table.
+See `public/vocabulary/state-machines.md` for the event record fields.
+
+## States
+
+| State | Stability class | Meaning |
+|---|---|---|
+| `discovered` | transient | Entity has been discovered but not yet registered |
+| `registered` | transient | Entity has been registered with an immutable ID |
+| `pending_verification` | transient | Entity is awaiting verification |
+| `verified` | stable | Entity has been verified |
+| `active` | stable | Entity is active and may participate in governed actions |
+| `limited` | stable | Entity is operating under reduced scope |
+| `under_review` | transient | Entity is under review |
+| `approval_required` | transient | Entity actions require approval |
+| `suspended` | stable | Entity is temporarily prevented from acting |
+| `stopped` | stable | Entity has been stopped by a stop-cascade event |
+| `quarantined` | stable | Entity has been quarantined pending investigation |
+| `revoked` | stable | Entity has been revoked |
+| `transferred` | terminal | Entity has been transferred to a new trust domain |
+| `archived` | terminal | Entity has been archived |
+| `tombstoned` | terminal | Entity is tombstoned and cannot be reactivated |
+
+## Entity-class to state machine mapping
+
+Each entity class links to its canonical state machine in the
+KYE State Library™:
+
+| Entity class | Default state machine ID |
+|---|---|
+| `tenant` | `kye:sm:kyeprotocol.com:core.tenant.v1` |
+| `legal_entity` | `kye:sm:kyeprotocol.com:core.legal_entity.v1` |
+| `billing_account` | `kye:sm:kyeprotocol.com:core.billing_account.v1` |
+| `domain` | `kye:sm:kyeprotocol.com:core.domain.v1` |
+| `policy_bundle` | `kye:sm:kyeprotocol.com:core.policy_bundle.v1` |
+| `workspace` | `kye:sm:kyeprotocol.com:core.workspace.v1` |
+| `project` | `kye:sm:kyeprotocol.com:core.project.v1` |
+| `team` | `kye:sm:kyeprotocol.com:core.team.v1` |
+| `resource` | `kye:sm:kyeprotocol.com:core.resource.v1` |
+| `principal` | `kye:sm:kyeprotocol.com:core.principal.v1` |
+| `model` | `kye:sm:kyeprotocol.com:lib.ai_governance.model.v1` |
+| `tool` | `kye:sm:kyeprotocol.com:lib.ai_governance.tool.v1` |
+| `external_app` | `kye:sm:kyeprotocol.com:core.external_app.v1` |
+| `audit_stream` | `kye:sm:kyeprotocol.com:core.audit_stream.v1` |
+
+Tenants may adopt sector-specific library machines or derive tightened
+variants (see `public/vocabulary/state-library.md`).
+
+## High-level transition expectations
+
+The full transition graph is normative and out of scope for this document. Two high-level expectations:
+
+- A `tombstoned` entity has no outbound transitions.
+- A `revoked` entity does not return directly to `active`; recovery requires an explicit re-registration or successor process.
+
+The mechanisms enforcing these expectations are part of the normative specification.

vocabulary/meaning-continuity.md

@@ -0,0 +1,137 @@
+# KYE™ Meaning Continuity vocabulary
+
+Names introduced by **KYE Meaning Continuity™**
+(`kye-meaning-continuity-v1`) — the meaning-preservation sub-profile
+of **KYE Continuity Profile™** v1.1.0. This document lists names
+only. Meaning-score computation, constraint-loss detection across
+handoffs, drift-cascade routing, reconfirmation-flow trigger
+thresholds, evidence-pack composition rules and sector-specific
+meaning packs are part of the patent track and are not published
+here.
+
+> **KYE Continuity Profile™ preserves the chain. KYE Meaning
+> Continuity™ preserves the meaning along the chain.**
+
+## Decision values
+
+| Value | Meaning |
+|---|---|
+| `meaning_preserved`         | Meaning preserved end-to-end. Route onward. |
+| `meaning_degraded`          | Non-material drift detected. Onward with note. |
+| `meaning_broken`            | Material meaning loss detected. Block. |
+| `meaning_unknown`           | Insufficient evidence to score. Pause. |
+| `require_clarification`     | Ambiguity in interpretation. Ask the principal. |
+| `require_reconfirmation`    | Material constraint changed. Ask the principal to reconfirm. |
+| `require_human_review`      | Drift cannot be resolved automatically. Escalate. |
+| `quarantine`                | Pattern matches a watched drift class. Freeze. |
+
+## Routing outcomes
+
+`pause_admission` · `quarantine_proposed_action` ·
+`route_to_admissibility` · `route_to_authority_check`.
+
+## Drift types
+
+What meaning continuity actively detects:
+
+| Drift type | Meaning |
+|---|---|
+| `constraint_loss`         | A material constraint declared by the principal was dropped during interpretation or handoff. |
+| `context_loss`            | The contextual record under which the intent was declared has been lost or unreferenced. |
+| `memory_conflict`         | Memory used during interpretation conflicts with the principal's declared intent. |
+| `timing_drift`            | Time elapsed between intent and evaluation has changed the meaning of the action. |
+| `handoff_drift`           | Meaning has drifted across one or more handoff boundaries. |
+| `incentive_drift`         | Optimisation goal in effect at evaluation time differs from the principal's declared goal. |
+| `state_transition_drift`  | A state transition between intent and evaluation has changed the meaning. |
+| `interpretation_drift`    | The interpreted goal differs materially from the declared goal. |
+| `assumption_injection`    | Assumptions not declared by the principal have been added during interpretation. |
+| `meaning_compression`     | Constraints or scope have been compressed below what the principal declared. |
+| `meaning_expansion`       | Constraints or scope have been expanded beyond what the principal declared. |
+
+## Continuity modules (Continuity v1.1.0)
+
+`intent_continuity` · `meaning_continuity` · `context_continuity` ·
+`memory_continuity` · `incentive_continuity` · `timing_continuity` ·
+`handoff_continuity` · `state_continuity` · `execution_continuity`.
+
+## Required objects
+
+| Object | Schema |
+|---|---|
+| Meaning Continuity Context  | `meaning-continuity-context.json` |
+| Handoff Trace               | `handoff-trace.json` |
+| Meaning Drift Event         | `meaning-drift-event.json` |
+| Meaning Continuity Decision | `meaning-continuity-decision.json` |
+
+## Reason codes
+
+`meaning_preserved` · `meaning_degraded` · `meaning_broken` ·
+`meaning_unknown` · `meaning_drift_detected` ·
+`authorised_meaning_preserved` · `authorised_meaning_lost` ·
+`constraint_lost` · `constraint_changed` · `constraint_added` ·
+`assumption_added` · `context_lost` ·
+`context_changed_since_intent` · `memory_conflict_detected` ·
+`memory_stale` · `memory_missing` ·
+`incentive_conflict_detected` · `timing_changed_meaning` ·
+`handoff_boundary_drift` · `handoff_integrity_degraded` ·
+`state_transition_changed_meaning` ·
+`principal_reconfirmation_required` ·
+`meaning_evidence_incomplete`.
+
+## Required obligations
+
+`pause_admission` · `request_clarification` ·
+`request_reconfirmation` · `prevent_commit_until_reconfirmed` ·
+`emit_meaning_drift_event` · `include_in_evidence_pack` ·
+`notify_principal` · `escalate_to_human_review` ·
+`quarantine_actor` · `bind_to_admissibility_decision`.
+
+## Handoff types
+
+`human_to_agent_instruction` · `agent_to_agent_delegation` ·
+`agent_to_sub_agent_delegation` · `agent_to_tool_request` ·
+`tool_to_agent_response` · `agent_to_human_clarification` ·
+`agent_to_human_reconfirmation`.
+
+## Meaning continuity signals
+
+`kye.meaning_continuity.check_started` ·
+`kye.meaning_continuity.check_completed` ·
+`kye.meaning_continuity.preserved` ·
+`kye.meaning_continuity.degraded` ·
+`kye.meaning_continuity.broken` ·
+`kye.meaning_continuity.requires_reconfirmation` ·
+`kye.meaning_continuity.requires_human_review` ·
+`kye.meaning_continuity.evidence_pack_generated`.
+
+## Meaning drift signals
+
+`kye.meaning_drift.detected` ·
+`kye.meaning_drift.resolved` ·
+`kye.meaning_drift.quarantined` ·
+`kye.meaning_drift.reconfirmation_requested` ·
+`kye.meaning_drift.reconfirmation_completed`.
+
+## Handoff signals
+
+`kye.handoff.trace_created` ·
+`kye.handoff.boundary_crossed` ·
+`kye.handoff.drift_detected` ·
+`kye.handoff.integrity_degraded`.
+
+## Memory / context / timing supporting signals
+
+`kye.memory_context.used` ·
+`kye.memory_context.conflict_detected` ·
+`kye.memory_context.stale` ·
+`kye.context_snapshot.changed` ·
+`kye.timing.meaning_changed`.
+
+## Patent-safe boundary
+
+Meaning-score computation, constraint-loss detection across
+handoffs, drift-cascade routing rules, reconfirmation-flow trigger
+thresholds, evidence-pack composition rules and sector-specific
+meaning packs are **not** in this repository. The intelligent
+runtime is part of the normative specification held privately and
+is delivered as a commercial product.

vocabulary/obligations.md

@@ -0,0 +1,41 @@
+# KYE™ Obligation Vocabulary
+
+Obligations are named requirements that may attach to a policy decision. This document lists obligation names recognized by KYE™.
+
+The way obligations are evaluated, enforced, and proved is part of the normative specification.
+
+## Audit and proof
+
+| Obligation | Meaning |
+|---|---|
+| `audit.emit` | An audit event must be emitted |
+| `proof.required` | A proof artifact must be produced |
+| `transparency.submit` | A signed statement must be submitted to a transparency service |
+| `trace.capture.required` | Execution trace must be captured |
+
+## Data handling
+
+| Obligation | Meaning |
+|---|---|
+| `redaction.required` | Output must be redacted |
+| `watermark.required` | Output must be watermarked |
+| `citable.output.required` | Output must include citations |
+| `external.send.block` | Output must not be sent externally |
+| `memory.write.block` | Memory writes are blocked |
+
+## Approval and review
+
+| Obligation | Meaning |
+|---|---|
+| `approval.required` | Human approval is required |
+| `human.review.required` | Human review is required |
+| `dry_run.only` | Action is permitted in dry-run only |
+| `break_glass.justification.required` | Break-glass justification is required |
+
+## Monitoring
+
+| Obligation | Meaning |
+|---|---|
+| `monitor.realtime` | Real-time monitoring is required |
+| `tool.allowlist.enforce` | Tool allowlist must be enforced |
+| `high_risk.decision.block` | High-risk decisions must be blocked |

vocabulary/ontology.md

@@ -0,0 +1,117 @@
+# KYE™ Ontology vocabulary
+
+Names introduced by **KYE Ontology Profile™** (`kye-ontology-v1`). This
+document lists names only. Semantic-resolution algorithms,
+false-equivalence detection heuristics, cross-profile mapping
+reconciliation, tenant overlay resolution and risk-weighted ontology
+traversal are part of the normative specification and the patent track,
+and are not published here.
+
+> **Schemas make data valid. Ontologies make data meaningful.**
+
+## Ontology domains
+
+Twelve domains. Each KYE term declares exactly one.
+
+| Domain | Surface |
+|---|---|
+| `entity` | Any nameable subject — human, org, agent, model, tool, device, dataset, credential, instrument, asset. |
+| `authority` | Authority kinds — delegation, mandate, consent, approval, permission, entitlement, licence, power_of_attorney, role_authority, regulatory_authority, contractual_authority, emergency_authority. |
+| `capability` | Named verbs — payment_initiation, card_purchase, data_access, contract_signing, clinical_access, tool_invocation, infrastructure_command, credential_rotation, settlement_approval, evidence_export. |
+| `scope` | Constraint kinds — amount_limit, time_window, merchant_category, jurisdiction, data_class, environment, resource_class, approval_threshold, purpose_limitation, retention_limit. |
+| `state` | State dimensions — entity_state, authority_state, delegation_state, credential_state, capability_state, risk_state, recovery_state, continuity_state, discovery_state, certification_state. |
+| `decision` | Verdict kinds — allow, allow_with_constraints, require_approval, deny, continuity_*. |
+| `evidence` | Artefact kinds — audit_event, decision_map, evidence_pack, payload_hash, signature, state_snapshot, intent_trace, continuity_context, discovery_audit_event. |
+| `continuity` | Drift types and continuity dimensions. |
+| `discoverability` | Discovery modes, risk-discovery types, masking classes. |
+| `connector` | Connector Profile™ family kinds. |
+| `sector` | Sector namespaces — payments, open_finance, legal, health, pensions, cyber, critical_infrastructure, sovereign_ai, telecom, pharma_gxp. |
+| `certification` | Conformance / certification artefacts. |
+
+## Predicates
+
+Stable predicate dictionary. New predicates require an RFC.
+
+`grants` · `limits` · `inherits_from` · `requires` · `requires_authority` ·
+`requires_scope` · `requires_state` · `requires_evidence` ·
+`requires_approval_from` · `expires_at` · `revokes` · `supersedes` ·
+`depends_on` · `constrains` · `narrows` · `blocks` · `allows` ·
+`requires_step_up` · `requires_human_review` · `triggers` ·
+`triggers_signal` · `invalidates_authority` · `proves` · `supports` ·
+`binds_to` · `references` · `verifies` · `replays` · `maps_to_control` ·
+`acts_on_behalf_of` · `has_authority` · `uses_capability` ·
+`evidenced_by` · `equivalent_to` · `related_not_identical` ·
+`not_equivalent_to` · `aliased_by` · `subsumes` · `subsumed_by`.
+
+## Mapping types
+
+External-system → KYE term mappings declare exactly one of:
+
+| Type | Meaning |
+|---|---|
+| `equivalent` | Source and KYE term are interchangeable for runtime purposes. |
+| `related_not_identical` | Overlap exists; source DOES NOT prove the KYE term alone; additional KYE objects MUST be presented (`requires_additional_kye_objects`). |
+| `not_equivalent` | The source term MUST NOT be treated as the KYE term. |
+| `aliased_by` | Label-level alias only; same KYE term. |
+| `subsumes` | Source narrower than the KYE term. |
+| `subsumed_by` | Source broader than the KYE term. |
+
+A mapping with `related_not_identical` MUST list
+`requires_additional_kye_objects` so the policy engine can deny when the
+source term is presented alone (reason code:
+`external_term_not_equivalent`).
+
+## Required objects
+
+| Object | Schema | Role |
+|---|---|---|
+| Ontology profile descriptor | `ontology-profile.json` | Profile manifest. |
+| Ontology term | `ontology-term.json` | One canonical term. |
+| Ontology relationship | `ontology-relationship.json` | (subject, predicate, object) triple with constraints. |
+| Ontology mapping | `ontology-mapping.json` | External-system → KYE term mapping with mapping_type + required objects. |
+| Semantic assertion | `semantic-assertion.json` | Decision-bound semantics, hash-chained into the audit ledger. |
+
+## Serializations
+
+| Layer | Format | Status |
+|---|---|---|
+| Runtime API | JSON Schema 2020-12 | required |
+| Semantic interoperability | JSON-LD context | recommended |
+| External research / KG integration | RDF / OWL | optional |
+
+JSON-LD context published alongside this document (`jsonld-context.json`).
+
+## Reason codes
+
+See `reason-codes.md` under the **Ontology profile reason codes** group.
+
+## Ontology signals
+
+Term lifecycle: `kye.ontology.term.created` ·
+`kye.ontology.term.updated` · `kye.ontology.term.deprecated` ·
+`kye.ontology.term.revoked`.
+
+Relationship lifecycle: `kye.ontology.relationship.created` ·
+`kye.ontology.relationship.updated` ·
+`kye.ontology.relationship.deprecated`.
+
+Mapping lifecycle: `kye.ontology.mapping.created` ·
+`kye.ontology.mapping.validated` ·
+`kye.ontology.mapping.conflict_detected` ·
+`kye.ontology.mapping.deprecated`.
+
+Semantic assertion: `kye.semantic_assertion.created` ·
+`kye.semantic_assertion.verified` ·
+`kye.semantic_assertion.conflict_detected`.
+
+Semantic graph: `kye.semantic_graph.updated` ·
+`kye.semantic_graph.rebuilt` ·
+`kye.semantic_graph.validation_failed`.
+
+## Patent-safe boundary
+
+The semantic-resolution algorithm, false-equivalence detection
+heuristics, cross-profile reconciliation engine, tenant overlay
+resolution and risk-weighted ontology traversal are **not** in this
+repository. See the public landing page for the profile overview;
+counsel-controlled material remains private.

vocabulary/operating-model.md

@@ -0,0 +1,96 @@
+# KYE™ Operating Model vocabulary
+
+Names introduced by **KYE Operating Model Profile™**
+(`kye-operating-model-v1`) — the enterprise adoption layer of KYE™.
+This document lists names only. Readiness scoring engine, risk-tiering
+heuristics, side-effect classification, commit-boundary detection,
+authority-gate recommendation and runtime enforcement internals are part
+of the normative specification and the patent track, and are not
+published here.
+
+> **KYE™ turns AI-governance operating models into runtime authority
+> decisions and replayable evidence.**
+
+## Journey stages
+
+Ten ordered stages every operating-model implementation enforces.
+
+`intake` · `assess` · `classify` · `map_authority` · `place_gates` ·
+`configure_runtime` · `execute` · `evidence` · `review` · `improve`.
+
+## Required objects
+
+| Object | Schema |
+|---|---|
+| Profile descriptor | `operating-model-profile.json` |
+| Use-case intake | `use-case-intake.json` |
+| Readiness assessment | `readiness-assessment.json` |
+| Entity Authority Record™ | `entity-authority-record.json` |
+| Authority Gate™ | `authority-gate.json` |
+| Commit Boundary™ | `commit-boundary.json` |
+| Review path | `review-path.json` |
+| Training record | `training-record.json` |
+| Adoption evidence pack | `adoption-evidence-pack.json` |
+| Governed catalogue entry | `governed-catalog-entry.json` |
+
+## Authority-gate types
+
+Eight named types. Implementations MAY add `custom`.
+
+| Type | Surface |
+|---|---|
+| `payment_execution`         | Before a prepared payment becomes an executed payment. |
+| `external_message`          | Before an outbound message (email, customer-facing notification, signal handler) leaves the tenant. |
+| `contract_signature`        | Before a contract / e-signature is sealed. |
+| `clinical_action`           | Before a clinical action is performed (orders, escalations, triage). |
+| `infrastructure_command`    | Before an infrastructure command is issued (deploy, rotate, terminate). |
+| `data_export`               | Before a data export leaves the tenant. |
+| `credential_rotation`       | Before a credential is rotated. |
+| `evidence_export`           | Before an evidence pack is exported beyond the tenant. |
+| `data_access`               | Before sensitive data is read. |
+| `custom`                    | Implementer-defined gate. |
+
+## Commit-boundary archetypes
+
+| Before the boundary | After the boundary |
+|---|---|
+| `draft email`            | `send email` |
+| `payment.prepare`        | `payment.execute` |
+| `recommend refund`       | `issue refund` |
+| `draft contract`         | `sign contract` |
+| `suggest clinical step`  | `perform clinical step` |
+| `plan infrastructure change` | `execute infrastructure command` |
+| `recommend access grant` | `issue access grant` |
+| `propose policy edit`    | `apply policy edit` |
+
+## Decision policy values
+
+`allow` · `deny` · `require_approval` · `require_step_up` ·
+`quarantine`.
+
+## Risk tiers
+
+`low` · `medium` · `high` · `critical`.
+
+## Lifecycle states
+
+`proposed` · `pilot` · `controlled_production` · `production` ·
+`deprecated` · `suspended` · `revoked`.
+
+## Reason codes
+
+See `reason-codes.md` under the **Operating Model profile reason codes**
+group (19 codes).
+
+## Operating-model signals
+
+See `kye-operating-model-v1.md` §6. Five families: use-case lifecycle,
+readiness lifecycle, authority-record lifecycle, authority-gate
+lifecycle, commit-boundary lifecycle, and catalogue lifecycle.
+
+## Patent-safe boundary
+
+The readiness-scoring algorithm, risk-tier classification thresholds,
+side-effect classification rules, commit-boundary detection algorithm,
+authority-gate recommendation engine, and runtime gate-enforcement
+internals are **not** in this repository.

vocabulary/payload-states.md

@@ -0,0 +1,88 @@
+# KYE Protocol™ — Payload artefact states
+
+> **Status:** Normative. v1.0. Apache 2.0. Companion to
+> the `action-admissibility` rule pack + `pep` profile and
+> `internal`.
+>
+> A payload artefact carries **state** but never **authority**. The
+> states below are the canonical values for the `state` field of the
+> KYE Payload Artefact schema.
+
+## 1. Lifecycle states
+
+`created` — payload object instantiated by the actor; not yet signed.
+
+`signed` — payload is in a state the Gateway will accept for verification.
+Construction details are part of the patent track and not disclosed here.
+
+`submitted` — sent by the actor to the Gateway.
+
+`received` — the Gateway has the bytes.
+
+`verified` — payload-trust checks completed; the payload is eligible
+for adjudication. The specific checks are part of the patent track.
+
+`bound_to_decision` — verified and adjudicated; the decision URN is
+recorded on the artefact.
+
+`executed` — the capability ran; the side-effect was committed.
+
+`failed` — execution faulted after a successful decision binding.
+
+`archived` — payload retained per retention policy; no further
+mutation permitted.
+
+## 2. Denial states
+
+`rejected` — one of the verification checks failed; see `reason_code`
+for which.
+
+`expired` — receive time is past `expires_at`.
+
+`replayed` — the payload-trust profile rejected this submission as
+non-fresh.
+
+`tampered` — the payload-trust profile rejected this submission for an
+integrity violation.
+
+## 3. Required deny reason codes
+
+`payload_signature_missing` ·
+`payload_signature_invalid` ·
+`payload_hash_mismatch` ·
+`payload_canonicalization_unsupported` ·
+`payload_nonce_missing` ·
+`payload_nonce_reused` ·
+`payload_expired` ·
+`payload_actor_mismatch` ·
+`payload_capability_mismatch` ·
+`payload_decision_binding_missing`
+
+## 4. Required signal types
+
+Every state transition MUST emit the corresponding signal:
+
+| Transition | Signal type |
+|---|---|
+| `submitted → received` | `payload.received` |
+| `received → verified` | `payload.verified` |
+| `received → rejected` | `payload.rejected` |
+| `verified → bound_to_decision` | `payload.bound_to_decision` |
+| `bound_to_decision → executed` | `payload.executed` |
+| `executed → failed` | `payload.failed` |
+| `* → expired` | `payload.expired` |
+| `* → replayed` | `payload.replay_detected` |
+| `* → tampered` | `payload.tampered` |
+| `* → archived` | `payload.archived` |
+
+## 5. Use in copy
+
+> **KYE governs acting entities, principal entities, capability
+> entities, and evidence artefacts.** Payloads are *evidence
+> artefacts* that bind signed requests to authority decisions and
+> audit trails.
+>
+> KYE binds signed payloads to entity authority, capability state,
+> policy decisions, and replayable audit evidence.
+
+— KYE Protocol™ project, 2026.

vocabulary/payments-decision-codes.md

@@ -0,0 +1,111 @@
+# KYE Protocol™ — Payments decision and lifecycle codes
+
+> **Status:** Normative. v1.0. Apache 2.0. Companion to
+> `internal` + `internal`.
+
+This file enumerates the public payment decision codes, hold reasons,
+reversal reasons, and dispute outcomes used by the KYE-Payments profile
+and its overlays (EU, Card, High-Assurance).
+
+## 1. sPDP decisions
+
+`payment.allow` · `payment.allow_with_constraints` ·
+`payment.require_approval` · `payment.deny`
+
+## 2. Deny reason codes
+
+`payment.deny.amount_above_authority_cap` ·
+`payment.deny.currency_not_in_scope` ·
+`payment.deny.rail_not_in_scope` ·
+`payment.deny.beneficiary_not_in_allowlist` ·
+`payment.deny.beneficiary_in_denylist` ·
+`payment.deny.sanctions_screen_failed` ·
+`payment.deny.kyc_step_up_required` ·
+`payment.deny.aml_threshold_exceeded` ·
+`payment.deny.payment_authority_revoked` ·
+`payment.deny.delegation_revoked` ·
+`payment.deny.attestation_stale` ·
+`payment.deny.risk_state_blocked` ·
+`payment.deny.daily_velocity_exceeded` ·
+`payment.deny.weekly_velocity_exceeded` ·
+`payment.deny.monthly_velocity_exceeded`
+
+## 3. Approval-required reason codes
+
+`payment.approval.amount_above_minor_threshold` ·
+`payment.approval.amount_above_major_threshold` ·
+`payment.approval.cross_border_first_time` ·
+`payment.approval.high_risk_corridor` ·
+`payment.approval.beneficiary_first_time` ·
+`payment.approval.psd2_sca_required` ·
+`payment.approval.regulatory_dual_control`
+
+## 4. Hold reason codes
+
+`payment.hold.risk_review` · `payment.hold.sanctions_screen` ·
+`payment.hold.kyc_step_up` · `payment.hold.fraud_review` ·
+`payment.hold.regulatory_block` · `payment.hold.rail_pending` ·
+`payment.hold.cutoff_window` · `payment.hold.dispute_pending`
+
+## 5. Reversal reason codes
+
+`payment.reversal.initiator_request` ·
+`payment.reversal.fraud_detected` ·
+`payment.reversal.duplicate_submission` ·
+`payment.reversal.amount_correction` ·
+`payment.reversal.beneficiary_invalid` ·
+`payment.reversal.regulatory_finding` ·
+`payment.reversal.sanctions_match`
+
+## 6. Dispute reason codes
+
+`payment.dispute.unauthorized` ·
+`payment.dispute.fraud_card_present` ·
+`payment.dispute.fraud_card_not_present` ·
+`payment.dispute.product_not_received` ·
+`payment.dispute.product_not_as_described` ·
+`payment.dispute.duplicate` ·
+`payment.dispute.amount_disagreement` ·
+`payment.dispute.credit_not_processed` ·
+`payment.dispute.cancelled_recurring` ·
+`payment.dispute.regulatory_chargeback`
+
+## 7. Charge-back outcomes
+
+`payment.charged_back.evidence_insufficient` ·
+`payment.charged_back.regulatory_finding` ·
+`payment.charged_back.merchant_no_response` ·
+`payment.charged_back.fraud_confirmed`
+
+## 8. Dispute-resolved outcomes
+
+`payment.dispute_resolved.evidence_sufficient` ·
+`payment.dispute_resolved.merchant_credit_issued` ·
+`payment.dispute_resolved.cardholder_withdrew` ·
+`payment.dispute_resolved.arbitration_in_favour`
+
+## 9. Reconciliation outcomes
+
+`payment.reconciled.matched` ·
+`payment.reconciled.matched_with_variance` ·
+`payment.reconciled.unmatched` ·
+`payment.reconciled.late`
+
+## 10. Settlement events
+
+`payment.settlement.posted` ·
+`payment.settlement.deferred` ·
+`payment.settlement.failed` ·
+`payment.settlement.partial`
+
+## 11. ISO 20022 message-class alignment
+
+Used by the High-Assurance overlay only:
+
+`pacs.008` (FIToFICustomerCreditTransfer) ·
+`pacs.009` (FinancialInstitutionCreditTransfer) ·
+`pacs.004` (PaymentReturn) ·
+`pacs.002` (FIToFIPaymentStatusReport) ·
+`camt.054` (BankToCustomerDebitCreditNotification)
+
+— KYE Protocol™ project, 2026.

vocabulary/reason-codes.md

@@ -0,0 +1,187 @@
+# KYE™ Reason Code Vocabulary
+
+Reason codes accompany decisions, signals, and lifecycle changes. This document lists reason codes recognized by KYE™.
+
+This document defines **names only**.
+
+| Code | Meaning |
+|---|---|
+| `policy_violation` | Action violated policy |
+| `credential_revoked` | Credential was revoked |
+| `delegation_revoked` | Delegation was revoked |
+| `assurance_below_minimum` | Assurance level below minimum |
+| `risk_threshold_exceeded` | Risk threshold exceeded |
+| `external_export_blocked` | External export blocked |
+| `workload_attestation_failed` | Workload attestation failed |
+| `manual_incident_response` | Manual incident response action |
+| `legal_hold_applied` | Legal hold applied |
+| `consent_withdrawn` | Consent withdrawn |
+
+## Payload Trust profile reason codes
+
+| Code | Meaning |
+|---|---|
+| `payload_signature_missing` | Payload signature absent |
+| `payload_signature_invalid` | Payload signature does not verify against declared credential |
+| `payload_hash_mismatch` | `payload_hash` does not match the canonicalised bytes |
+| `payload_canonicalization_unsupported` | Gateway does not support the declared canonicalisation suite |
+| `payload_nonce_missing` | Freshness window present but nonce empty |
+| `payload_nonce_reused` | Nonce already seen inside the freshness window |
+| `payload_expired` | Receive time is past `expires_at` |
+| `payload_actor_mismatch` | `actor_entity_id` does not match the credential's subject |
+| `payload_capability_mismatch` | `capability_id` is not granted to the actor |
+| `payload_decision_binding_missing` | Action attempted without a successful `bound_to_decision` transition |
+
+## Taxonomy & Metadata profile reason codes
+
+| Code | Meaning |
+|---|---|
+| `taxonomy_term_unknown` | Term value is not registered in the referenced taxonomy |
+| `taxonomy_term_deprecated` | Term resolved to an active taxonomy but the term itself is `status=deprecated` |
+| `taxonomy_version_mismatch` | Mapped value uses a taxonomy version not accepted by the policy |
+| `metadata_schema_unknown` | `metadata_schema_id` is not registered |
+| `metadata_required_field_missing` | A required field defined by the metadata schema is absent |
+| `metadata_field_taxonomy_violation` | A field value is not from the schema's declared taxonomy |
+| `metadata_classification_drift` | Classification field changed without re-binding evidence |
+| `data_class_not_allowed` | Resource carries a data class that is not permitted for the actor's profile |
+
+## Graph profile reason codes
+
+| Code | Meaning |
+|---|---|
+| `graph_authority_path_missing` | No valid authority path from actor to action / resource |
+| `graph_delegation_path_missing` | No active delegation path between actor and principal |
+| `graph_delegation_path_disputed` | Delegation path contains a node in `disputed` state |
+| `graph_capability_dependency_compromised` | A capability dependency is in `state=compromised` |
+| `graph_capability_dependency_high_risk` | A capability dependency carries `risk_state=high` |
+| `graph_blast_radius_too_broad` | Compromise blast radius exceeds the configured safety bound |
+| `graph_traversal_depth_exceeded` | Query depth limit reached without a complete result |
+| `graph_node_not_found` | Referenced node URN is not registered |
+| `graph_edge_revoked` | Edge has been revoked since the last cache snapshot |
+
+## Ontology profile reason codes
+
+| Code | Meaning |
+|---|---|
+| `ontology_term_missing` | Referenced KYE term is not registered |
+| `ontology_term_deprecated` | Referenced KYE term is `status=deprecated` |
+| `ontology_mapping_missing` | A required external-system → KYE mapping was not found |
+| `ontology_mapping_conflict` | Two mappings make incompatible claims for the same source term |
+| `semantic_equivalence_rejected` | Asserted equivalence violates a `not_equivalent_to` rule |
+| `semantic_relationship_ambiguous` | Resolution returned multiple candidate predicates |
+| `semantic_resolution_failed` | Resolver could not produce a canonical KYE term |
+| `semantic_assertion_verified` | Semantic assertion verified successfully |
+| `semantic_assertion_conflict` | Semantic assertion conflicts with a hash-chained prior assertion |
+| `external_term_not_equivalent` | Source term presented as KYE authority without required additional KYE objects |
+| `profile_term_mismatch` | Term used outside its declared profile_refs |
+| `sector_term_requires_mapping` | Cross-sector decision requires a validated mapping; none present |
+| `ontology_policy_denied` | Action denied by an ontology-level policy rule |
+
+## Operating Model profile reason codes
+
+| Code | Meaning |
+|---|---|
+| `use_case_owner_missing` | Use-case intake lacks an `owner_entity_id` |
+| `use_case_purpose_missing` | Use-case intake lacks a `business_objective` |
+| `data_classification_missing` | Use-case intake lacks `data_classes` |
+| `authority_model_missing` | No `KYEEntityAuthorityRecord` exists for the actor |
+| `review_path_missing` | A `KYEReviewPath` is required but not configured |
+| `evidence_requirements_missing` | Required evidence list is empty for a high-risk capability |
+| `revocation_path_missing` | Authority record lacks a revocation policy |
+| `commit_boundary_missing` | A capability with `external_effect_possible` lacks a `KYECommitBoundary` |
+| `authority_gate_missing` | A capability requires a `KYEAuthorityGate` that does not exist |
+| `human_review_required` | Risk tier mandates human review before commit |
+| `autonomous_execution_not_allowed` | Readiness assessment forbids autonomous execution |
+| `payment_execution_gate_required` | Payment execution must be guarded by a payment gate |
+| `external_send_gate_required` | Outbound message must be guarded by an external-send gate |
+| `clinical_action_gate_required` | Clinical action must be guarded by a clinical-action gate |
+| `legal_commit_gate_required` | Contract / signature must be guarded by a legal-commit gate |
+| `infrastructure_command_gate_required` | Infrastructure command must be guarded by an infrastructure gate |
+| `readiness_incomplete` | Readiness assessment dimensions are incomplete |
+| `pilot_ready_with_controls` | Readiness sufficient for pilot under enumerated controls |
+| `production_ready` | Readiness sufficient for production execution |
+
+## Assurance Card profile reason codes
+
+| Code | Meaning |
+|---|---|
+| `assurance_card_missing` | Subject entity has no associated `KYEAssuranceCard` |
+| `assurance_card_owner_missing` | Assurance card lacks an `owner_entity_id` |
+| `intended_use_undocumented` | `intended_use.summary` is missing |
+| `prohibited_uses_missing` | `intended_use.prohibited_uses` is empty for a high-risk subject |
+| `provenance_missing` | No `KYEProvenanceEvidence` is bound to the assurance card |
+| `human_involvement_plan_missing` | Subject performs commit-level capabilities but no `KYEHumanInvolvementPlan` is bound |
+| `approval_threshold_breach` | Action exceeds an approval threshold without recorded human approval |
+| `review_overdue` | `KYEAssuranceReviewCycle.next_review_due` has passed |
+| `scope_change_review_required` | Authority record / gates / commit-boundaries changed without a triggered review |
+| `incident_review_required` | Incident occurred without a triggered review |
+| `retention_window_expired` | Retention window expired without archival or decommissioning |
+| `decommissioning_plan_missing` | Subject is `deprecated` / `revoked` without an active `KYEDecommissioningPlan` |
+| `assurance_card_runtime_evidence_unbound` | Assurance card lacks references to runtime decision maps / evidence packs |
+| `provenance_supplier_unverified` | A supplier in the provenance evidence is `verified=false` |
+| `provenance_licence_change_unrecorded` | Licence change detected but not recorded as a review trigger |
+| `evidence_replay_failed` | Offline replay of bound evidence failed verification |
+| `bypass_of_human_involvement_attempted` | Runtime detected an attempt to bypass a `KYEHumanInvolvementPlan` involvement point |
+
+## Formal Rules profile reason codes
+
+| Code | Meaning |
+|---|---|
+| `permission_granted` | Standing permission resolved as allow |
+| `permission_missing` | No matching permission for this capability + scope + state |
+| `permission_expired` | Permission's `valid_until` has passed |
+| `permission_revoked` | Permission was revoked by an exercised power |
+| `permission_scope_exceeded` | Action exceeds the permission's declared scope |
+| `obligation_created` | Decision created a new obligation |
+| `obligation_pending` | Obligation has been created but not yet satisfied |
+| `obligation_satisfied` | Obligation satisfaction conditions all met |
+| `obligation_breached` | Obligation deadline passed without satisfaction |
+| `obligation_waived` | Obligation waived by an exercised power |
+| `obligation_expired` | Obligation expired without satisfaction or waiver |
+| `obligation_disputed` | Obligation flagged as disputed pending review |
+| `obligation_not_satisfied` | Action denied because a required obligation is unsatisfied |
+| `prohibition_triggered` | Prohibition condition matched the requested action |
+| `prohibited_action_requested` | Subject requested an action enumerated in `prohibited_capability` |
+| `prohibited_action_without_approval` | Action prohibited unless an approval event is present |
+| `power_missing` | Holder does not hold the required power for this operation |
+| `power_exercised` | Power was successfully exercised |
+| `power_not_valid` | Holder's power is in an invalid state |
+| `power_revoked` | Power has been revoked |
+| `override_requested` | Governance override requested |
+| `override_approved` | Governance override approved |
+| `override_denied` | Governance override denied |
+| `override_expired` | Override expired without re-approval |
+| `override_requires_second_approver` | Override requires a second approver per governance rule |
+| `rule_conflict_detected` | Two or more rules conflict |
+| `permission_prohibition_conflict` | Permission and prohibition both matched |
+| `obligation_without_satisfaction_path` | Obligation has no path to `satisfied` |
+| `unbounded_obligation` | Obligation lacks a deadline or termination condition |
+| `circular_delegation` | Delegation chain forms a cycle |
+| `missing_evidence_requirement` | Rule requires evidence not declared in the rule set |
+| `rule_priority_resolved` | Rule conflict resolved by priority |
+| `specific_rule_overrides_general_rule` | Specific rule won over general rule on resolution |
+
+## Action Admissibility profile reason codes
+
+| Code | Meaning |
+|---|---|
+| `proposed_action_admissible` | Proposed action passes all checked dimensions |
+| `intent_invalid` | Declared intent is structurally invalid |
+| `intent_ambiguous` | Declared / interpreted intent ambiguity exceeds threshold |
+| `intent_missing` | No intent reference attached to the proposed action |
+| `principal_missing` | No principal entity resolves on the proposed action |
+| `actor_missing` | No actor entity resolves on the proposed action |
+| `accountable_owner_missing` | No accountable owner resolves on the proposed action |
+| `capability_unknown` | Requested capability is not registered |
+| `resource_unknown` | Target resource is not registered |
+| `data_source_disallowed` | Source of input data is disallowed by policy |
+| `evidence_inadmissible` | Required evidence references are missing or untrusted |
+| `prohibited_action_class` | Action matches a prohibited class for this actor / sector |
+| `unsafe_tool_path_detected` | Proposed tool path matches an unsafe pattern |
+| `coercion_signal_detected` | Pressure / coercion signal detected upstream of the proposal |
+| `incentive_conflict_detected` | Agent incentive conflicts with declared intent |
+| `continuity_drift_blocks_admission` | Continuity drift detected upstream blocks admission |
+| `policy_ineligible_action` | Proposal ineligible under the active policy set |
+| `formal_rule_blocks_admission` | A formal rule (typically a prohibition) blocks admission |
+| `jurisdiction_unsupported` | No supported jurisdiction binding for the proposed action |
+| `risk_state_blocks_admission` | Current risk state of the actor / principal blocks admission |

vocabulary/redaction-fields.md

@@ -0,0 +1,52 @@
+# KYE Protocol™ — Telemetry redaction field list
+
+> **Status:** Normative. v1.0. Apache 2.0. Companion to
+> the canonical telemetry dictionary (`public/vocabulary/telemetry.md`).
+
+Every telemetry-emitting Gateway MUST replace the values of the JSON
+paths below with a per-tenant redaction token before export.
+
+## 1. PII
+
+`pii.full_name` · `pii.first_name` · `pii.last_name` ·
+`pii.email` · `pii.phone` · `pii.address.line1` ·
+`pii.address.line2` · `pii.address.postal_code` ·
+`pii.address.country` · `pii.dob` · `pii.gov_id` ·
+`pii.passport_number` · `pii.tax_id` · `pii.ssn` ·
+`pii.driving_licence` · `pii.biometric_template`
+
+## 2. Financial
+
+`payment.account_number` · `payment.iban` · `payment.routing_number` ·
+`payment.card_pan` · `payment.card_cvv` · `payment.card_expiry` ·
+`payment.beneficiary.name` · `payment.beneficiary.address` ·
+`treasury.wallet_address` · `custody.private_key_handle`
+
+## 3. Healthcare
+
+`healthcare.mrn` · `healthcare.diagnosis_codes` ·
+`healthcare.medication_list` · `healthcare.lab_results` ·
+`healthcare.notes_freetext` · `healthcare.insurance_id` ·
+`healthcare.next_of_kin`
+
+## 4. Authentication / secrets
+
+`auth.password` · `auth.password_hash` · `auth.totp_secret` ·
+`auth.recovery_code` · `auth.api_key` · `auth.bearer_token` ·
+`auth.refresh_token` · `auth.session_id` · `auth.client_secret` ·
+`auth.private_key_pem`
+
+## 5. Capability invocations
+
+`capability.input.user_message` · `capability.output.completion` ·
+`capability.input.attached_file_bytes` ·
+`capability.tool_arguments.user_supplied`
+
+## 6. Free-form fields
+
+Any field whose JSON path matches `*.notes` · `*.comment` ·
+`*.message` · `*.body` and that is not on a structured-allowlist
+(see `public/vocabulary/structured-allowlist.md`) is treated as
+free-form and MUST be redacted.
+
+— KYE Protocol™ project, 2026.

vocabulary/relationships.md

@@ -0,0 +1,88 @@
+# KYE™ Relationship Vocabulary
+
+Relationships are typed edges between KYE Protocol™ entities. This document lists the relationship types KYE™ recognizes.
+
+This document defines **names only**. Validation rules, evidence requirements, and lifecycle constraints are part of the normative specification.
+
+## v3 canonical relationship edges (locked 2026-05-14)
+
+The five canonical edges cover the most common cross-entity relationships
+in the v3 hierarchy. Every edge is explicit — it is stored as a
+`kye:rel:…` record and governed by an authority grant.
+
+| Edge type | Typical source | Typical target | Meaning |
+|---|---|---|---|
+| `member_of` | Principal (`kye:prin:…`) | Team (`kye:team:…`) | The principal is a member of the team and inherits its RBAC grants. |
+| `acts_in` | Principal (`kye:prin:…`) | Workspace (`kye:wsp:…`) | The principal is authorized to act within the workspace. |
+| `granted_access_to` | Principal / Agent / External App | Resource / Model / Tool | The source holds an authority grant to use the target. |
+| `uses` | Agent Principal (`kye:prin:…`) | Model (`kye:model:…`) or Tool (`kye:tool:…`) | The agent invokes the model or tool during its operation. |
+| `applies_to` | Policy Bundle (`kye:pol:…`) | Workspace / Project / Team | The policy bundle is in force for the target scope. |
+
+These five edges are the minimum a conformant implementation MUST store
+as relationship records. Additional edges from the full catalogue below
+are optional extensions.
+
+## Ownership and control
+
+| Relationship | Meaning |
+|---|---|
+| `owned_by` | Source is owned by target |
+| `controlled_by` | Source is controlled by target |
+| `operated_by` | Source is operated by target |
+| `custodied_by` | Source is in custody of target |
+
+## Authority and delegation
+
+| Relationship | Meaning |
+|---|---|
+| `acts_on_behalf_of` | Source acts on behalf of target |
+| `delegated_by` | Source's authority was delegated by target |
+| `authorized_by` | Source's action was authorized by target |
+| `approved_by` | Source was approved by target |
+
+## Issuance and verification
+
+| Relationship | Meaning |
+|---|---|
+| `issued_by` | Source was issued by target |
+| `verified_by` | Source was verified by target |
+| `attested_by` | Source was attested by target |
+| `revoked_by` | Source was revoked by target |
+
+## Use, invocation, and execution
+
+| Relationship | Meaning |
+|---|---|
+| `uses` | Source uses target |
+| `invokes` | Source invokes target |
+| `hosts` | Source hosts target |
+| `deploys` | Source deploys target |
+| `executes` | Source executes target |
+| `accesses` | Source accesses target |
+
+## Provenance
+
+| Relationship | Meaning |
+|---|---|
+| `trained_from` | Source was trained from target |
+| `derived_from` | Source was derived from target |
+
+## Governance and observation
+
+| Relationship | Meaning |
+|---|---|
+| `scoped_to` | Source is scoped to target |
+| `governed_by` | Source is governed by target |
+| `allowed_by` | Source is allowed by target |
+| `denied_by` | Source is denied by target |
+| `observed_by` | Source is observed by target |
+| `monitored_by` | Source is monitored by target |
+
+## Lifecycle and structure
+
+| Relationship | Meaning |
+|---|---|
+| `belongs_to` | Source belongs to target |
+| `transferred_to` | Source was transferred to target |
+| `mapped_to_control` | Source maps to a control identifier |
+| `mapped_to_framework` | Source maps to a control framework |

vocabulary/risk-classes.md

@@ -0,0 +1,34 @@
+# Risk classes
+
+Canonical risk tier vocabulary used by the KYE Risk Engine™. Five tiers, EU-AI-Act-aligned, with explicit per-framework floor mappings.
+
+| Tier | Score range | Meaning | Mandatory controls |
+|---|---|---|---|
+| **`minimal`** | 0–20 | Effectively negligible risk — spam filters, recommendation systems on entertainment, AI-assisted typing. | None. Lifecycle logging recommended. |
+| **`limited`** | 20–50 | Transparency-only risk — must disclose AI involvement to the natural person (EU AI Act Art. 50). | Identification disclosure; audit-chain emission of every interaction. |
+| **`high`** | 50–85 | Conformity assessment required (EU AI Act Annex III: biometric ID, education, employment, essential public services, law enforcement, migration, justice, democratic processes). | Dual-channel sign-off; scenario testing before action; replay-proof generation; human-in-the-loop for irreversibles. |
+| **`unacceptable`** | 85–99 | Exceeds high-tier controls — only deployable with regulator-approved derogation. | Every high-tier control PLUS regulator notification, ≤1-hour kill-switch SLA, dual-signed Operating Model amendment. |
+| **`prohibited`** | 99–100 | The Risk Engine MUST block. Mirrors EU AI Act Art. 5 prohibited practices: social scoring by public authorities, real-time remote biometric identification in public spaces, exploitation of vulnerabilities, subliminal manipulation. | Decision Engine returns deny + reason `PROHIBITED_PRACTICE`. No Operating Model amendment can override. |
+
+## Per-framework floor mappings
+
+| Tier | EU AI Act | DORA | GDPR | NIST AI RMF | ISO 42001 | FCA OpRes | PCI DSS | SOX |
+|---|---|---|---|---|---|---|---|---|
+| `minimal` | minimal | non_critical | normal | map_1 | low | non_important | non_chd | non_icfr |
+| `limited` | limited | non_critical | normal | map_2 | low | non_important | non_chd | non_icfr |
+| `high` | high | critical_or_important | high | map_3 | high | important_business_service | chd_present | icfr_relevant |
+| `unacceptable` | unacceptable | critical_or_important | high | map_4 | high | important_business_service | chd_present | icfr_relevant |
+| `prohibited` | prohibited | critical_or_important | high | map_4 | high | important_business_service | chd_present | icfr_relevant |
+
+## Source files
+
+| Artefact | Path |
+|---|---|
+| Canonical dictionary | `internal` |
+| Schema | `internal` |
+| Example | `public/examples/risk-assessment.json` |
+| Runtime engine | `internal` |
+| HTTP agent | `internal` |
+| OpenAPI rail | `internal` |
+| Mechanism | `internal` |
+| Patent | `internal` |

vocabulary/self-audit-areas.md

@@ -0,0 +1,52 @@
+# KYE Self-Audit Canonical Areas (V1)
+
+> **Status:** Normative. v1.0. Apache 2.0. Companion to
+> the `evidence-replay` profile + `compliance-evidence` rule pack (post-§29 landing).
+
+A complete `KYESelfAuditRun` MUST cover the 17 canonical areas below.
+A finding's `area` field is drawn from this list.
+
+## Areas
+
+`schema` · `dictionary` · `profile` · `registry` ·
+`entity_state` · `authority_grant` · `delegation_path` ·
+`capability_state` · `credential_binding` · `policy_binding` ·
+`decision_correctness` · `reason_code_correctness` ·
+`audit_trail_integrity` · `evidence_completeness` ·
+`webhook_delivery_integrity` · `graph_consistency` ·
+`recovery_revocation_correctness`
+
+## Finding severity
+
+`info` · `low` · `medium` · `high` · `critical`
+
+## Self-audit result
+
+`passed` · `passed_with_warnings` · `failed` · `incomplete` · `not_applicable`
+
+## Attestation status
+
+`draft` · `signed` · `published` · `expired` · `revoked` · `superseded`
+
+## Required signal types
+
+```
+kye.self_audit.started
+kye.self_audit.completed
+kye.self_audit.failed
+kye.self_audit.finding_created
+kye.self_audit.critical_finding_detected
+kye.conformance.self_test_passed
+kye.conformance.self_test_failed
+kye.attestation.created
+kye.attestation.signed
+kye.attestation.published
+kye.attestation.expired
+kye.attestation.revoked
+kye.audit.integrity_check_failed
+kye.evidence.completeness_check_failed
+kye.policy.coverage_gap_detected
+kye.decision.replay_mismatch
+```
+
+— KYE Protocol™ project, 2026.

vocabulary/signal-types.md

@@ -0,0 +1,163 @@
+# KYE Protocol™ — Signal types (canonical inventory)
+
+> **Status:** Normative. v1.0. Apache 2.0. Companion to
+> the canonical signals dictionary (`public/vocabulary/signals.md`).
+
+Every signal `signal_type` value MUST be from this canonical
+inventory. A signal whose type is not in this list is a conformance
+failure (`reason_code=signal_type_unknown`).
+
+## 1. Entity lifecycle
+
+`entity.discovered` · `entity.registered` · `entity.verified` ·
+`entity.activated` · `entity.limited` · `entity.under_review` ·
+`entity.suspended` · `entity.quarantined` · `entity.stopped` ·
+`entity.compromised` · `entity.transferred` · `entity.merged` ·
+`entity.deprecated` · `entity.archived` · `entity.tombstoned`
+
+## 2. Authority
+
+`authority.granted` · `authority.elevated` · `authority.limited` ·
+`authority.suspended` · `authority.revoked` · `authority.disputed` ·
+`authority.frozen`
+
+## 3. Delegation
+
+`delegation.requested` · `delegation.approved` · `delegation.activated` ·
+`delegation.attenuated` · `delegation.expired` · `delegation.revoked` ·
+`delegation.parent_revoked` · `delegation.transferred` ·
+`delegation.blocked`
+
+## 4. Credential
+
+`credential.issued` · `credential.presented` · `credential.verified` ·
+`credential.rotating` · `credential.rotated` · `credential.expired` ·
+`credential.revoked` · `credential.compromised` · `credential.lost` ·
+`credential.recovered` · `credential.erased`
+
+## 5. Attestation
+
+`attestation.registered` · `attestation.refreshed` ·
+`attestation.stale` · `attestation.revoked` · `attestation.failed`
+
+## 6. Capability
+
+`capability.registered` · `capability.granted` ·
+`capability.invoked` · `capability.deny` · `capability.require_approval` ·
+`capability.quarantined` · `capability.superseded` ·
+`capability.revoked` · `capability.authority_revoked` ·
+`capability.dependency_blocked` · `capability.notifications_published` ·
+`capability.invocation_cancelled`
+
+## 7. Recovery / break-glass
+
+`recovery.requested` · `recovery.evidence_pending` ·
+`recovery.under_review` · `recovery.window_expired` ·
+`recovery.escalated` · `recovery.approved` · `recovery.rejected` ·
+`recovery.completed` · `recovery.cancelled` ·
+`break_glass.requested` · `break_glass.granted` ·
+`break_glass.used` · `break_glass.expired` · `break_glass.frozen` ·
+`compromise.reported` · `compromise.cascade_completed`
+
+## 8. Risk
+
+`risk.normal` · `risk.elevated` · `risk.high` · `risk.critical` ·
+`risk.blocked` · `risk.cleared` · `risk.threshold_exceeded` ·
+`risk.score_updated`
+
+## 9. State composition
+
+`state.composition_violation` · `state.transition_committed` ·
+`state.transition_rejected`
+
+## 10. Audit / runtime
+
+`audit.event_appended` · `audit.chain_verified` ·
+`audit.point_in_time_replayed` · `runtime.cascade_started` ·
+`runtime.cascade_completed`
+
+## 11. Webhook / signal-bus
+
+`webhook.delivered` · `webhook.failed` · `webhook.retried` ·
+`webhook.dead_lettered` · `webhook.signing_key_rotated` ·
+`signal.replayed` · `signal.subscribe_acknowledged`
+
+## 12. Federation
+
+`federation.entity_imported` · `federation.entity_exported` ·
+`federation.scope_attenuation_violation` ·
+`federation.conflict_resolved` · `federation.replication_lag_high` ·
+`federation.key_archived`
+
+## 13. Transparency
+
+`transparency.statement_appended` · `transparency.receipt_issued` ·
+`transparency.inclusion_verified`
+
+## 14. Telemetry
+
+`telemetry.decision_emitted` · `telemetry.redacted` ·
+`telemetry.export_failed`
+
+## 15. Payments
+
+(per `public/vocabulary/payments-decision-codes.md` §1, §11; emitted as
+signal types `payment.<code>`.)
+
+## 16. Healthcare
+
+`healthcare.consent_presented` · `healthcare.consent_withdrawn` ·
+`healthcare.redaction_applied` · `healthcare.external_send_blocked` ·
+`healthcare.break_glass_invoked` · `healthcare.post_hoc_review_filed`
+
+## 17. Treasury / custody
+
+`treasury.intent_created` · `treasury.rebalanced` ·
+`treasury.reconciled` · `treasury.exception_filed` ·
+`custody.withdraw_requested` · `custody.withdraw_completed` ·
+`custody.attestation_required` · `custody.cold_storage_unsealed`
+
+## 18. Conformance
+
+`conformance.fixture_run_started` · `conformance.fixture_run_completed` ·
+`conformance.report_signed`
+
+## 19. Keys
+
+`keys.rotation_requested` · `keys.rotated` · `keys.archived` ·
+`keys.suite_changed`
+
+## 20. Payload artefacts (KYE Payload Trust Profile™)
+
+`payload.received` · `payload.verified` · `payload.rejected` ·
+`payload.expired` · `payload.replay_detected` · `payload.tampered` ·
+`payload.bound_to_decision` · `payload.executed` · `payload.failed` ·
+`payload.archived`
+
+## 21. Taxonomies (KYE Taxonomy & Metadata Profile™)
+
+`taxonomy.created` · `taxonomy.updated` · `taxonomy.published` ·
+`taxonomy.deprecated` · `taxonomy.term_added` ·
+`taxonomy.term_updated` · `taxonomy.term_deprecated` ·
+`taxonomy.mapping_created` · `taxonomy.mapping_deprecated`
+
+## 22. Metadata bindings (KYE Taxonomy & Metadata Profile™)
+
+`metadata.schema_created` · `metadata.schema_updated` ·
+`metadata.schema_deprecated` · `metadata.validation_failed` ·
+`metadata.classification_changed` · `metadata.binding_created` ·
+`metadata.binding_removed`
+
+## 23. Authority graph (KYE Graph Profile™)
+
+`graph.node_created` · `graph.node_updated` · `graph.node_archived` ·
+`graph.edge_created` · `graph.edge_updated` · `graph.edge_revoked` ·
+`graph.decision_map_emitted` · `graph.blast_radius_computed` ·
+`graph.compliance_map_refreshed` · `graph.traversal_failed`
+
+## 24. Policy administration (KYE PAP — v1.1 preview)
+
+`policy.uploaded` · `policy.activated` · `policy.deactivated` ·
+`policy.version_published` · `policy.rolled_back`
+
+— KYE Protocol™ project, 2026.

vocabulary/signals.md

@@ -0,0 +1,136 @@
+# KYE Protocol™ — Signal types: semantic vocabulary (v1)
+
+> **Status:** Normative public surface. Apache 2.0.
+> Companion to `public/vocabulary/signal-types.md` (full enumeration) and the
+> private signal registry at `internal`.
+>
+> This document describes the **meaning** of each v3 signal type.
+> Implementation internals (signing algorithms, PDP mechanics) are out of scope.
+
+## What is a KYE Protocol™ signal?
+
+A **signal** is an asynchronous, envelope-wrapped event emitted by a KYE Protocol™
+system component whenever a governed state change occurs. Signals enable loosely-coupled
+subscribers — partner systems, audit stores, compliance dashboards — to react to KYE
+lifecycle events without polling.
+
+Every signal envelope carries:
+
+- a globally-unique `signal_id`
+- a `signal_type` token from this vocabulary
+- a `tenant_id` / `workspace_id` routing header
+- a type-specific `payload` object
+- an `idempotency_key` for at-least-once delivery safety
+- a cryptographic `signature` binding the envelope to a verified emitter identity
+
+## Signal type descriptions
+
+### Entity lifecycle signals
+
+**`entity.created`**
+Emitted when a new KYE entity — organisation, agent, model, dataset, or other governed
+class — is first registered in a trust domain. Subscribers receive the full entity
+snapshot so they can initialise derived records without a follow-on read.
+
+**`entity.updated`**
+Emitted when one or more fields of an existing entity record are mutated. The payload
+includes the post-mutation snapshot and a list of changed field paths, enabling
+subscribers to apply targeted updates.
+
+**`entity.deleted`**
+Emitted when a KYE entity is hard-deleted or tombstoned. Downstream systems should
+treat this as authoritative: any cached or derived data for the named entity MUST be
+purged or marked inactive upon receipt.
+
+### State signals
+
+**`state.transitioned`**
+Emitted by the State Engine when an entity successfully fires a transition in its
+governing state machine. The payload names the machine, the before-state, the
+after-state, the transition identifier, and any evidence artefacts that satisfied
+the transition guards.
+
+**`library.adopted`**
+Emitted when a tenant binds a canonical State Library entry to their operating
+environment, creating a tenant-specific state machine derivation. Indicates that the
+tenant's entities governed by that machine are now subject to the library's transition
+rules and evidence requirements.
+
+### Decision signals
+
+**`decision.admitted`**
+Emitted when the policy decision point admits an action for a governed entity. Subscribers
+may use this to track authorised activity for audit, billing, or telemetry purposes.
+
+**`decision.denied`**
+Emitted when the policy decision point denies an action. The payload includes structured
+denial reasons and the policy references that caused the denial. Subscribers such as
+SIEM exporters and the Drift Detector should act on this signal to identify anomalous
+access patterns.
+
+### Drift signal
+
+**`drift.detected`**
+Emitted when an entity's observed behaviour diverges from its declared operating model.
+Drift categories include capability overreach, authority gaps, state inconsistencies,
+policy violations, and identity mismatches. Severity is graded `low | medium | high | critical`.
+
+### Evidence signal
+
+**`evidence.sealed`**
+Emitted when an evidence pack is cryptographically sealed and its integrity hash recorded.
+The payload includes the pack identifier, its SHA-256 digest, and the key identifier of
+the signing key. Subscribers may use this to verify pack integrity independently.
+
+### Incident signals
+
+**`incident.opened`**
+Emitted when a new compliance or operational incident is opened. Carries a severity grade
+and a list of implicated entity identifiers.
+
+**`incident.closed`**
+Emitted when an incident is resolved and closed. Carries the resolution summary and the
+root-cause category to enable trend analysis.
+
+### DSAR signals
+
+**`dsar.requested`**
+Emitted when a Data Subject Access Request (DSAR) is submitted by or on behalf of a data
+subject. Subscribers should initiate the appropriate fulfilment workflow upon receipt.
+Carries the request type (access, erasure, rectification, portability, restriction, objection)
+and the statutory deadline.
+
+**`dsar.fulfilled`**
+Emitted when a DSAR is fully processed and the response dispatched to the data subject.
+The outcome field distinguishes `fulfilled`, `partially_fulfilled`, and `refused` outcomes.
+
+### Revocation signal
+
+**`revocation.cascaded`**
+Emitted by the Revocation Cascade Orchestrator once a root-entity revocation has been
+propagated to all dependent entities (agents, connectors, delegations). The payload lists
+every entity that was revoked as a consequence, enabling subscribers to update derived
+authority records in a single atomic step.
+
+### Compliance and audit signals
+
+**`compliance_card.refreshed`**
+Emitted whenever a tenant's compliance card is recalculated. The payload identifies the
+card, its new version number, and the overall compliance status (`compliant`, `at_risk`,
+or `non_compliant`).
+
+**`audit_pack.assembled`**
+Emitted when the Evidence Pack Assembler has compiled an audit pack from staged artefacts,
+covering a defined time window. The pack is ready for sealing and external submission.
+
+### Infrastructure signal
+
+**`webhook.failed`** _(dead-letter)_
+Emitted to the dead-letter queue when the Webhook Dispatcher has exhausted all delivery
+retries for an outbound webhook. The payload records the original signal identifier,
+the target subscriber, the number of attempts made, and the last error received. This
+signal triggers human review and optional manual replay.
+
+---
+
+_KYE Protocol™ project, 2026. Apache 2.0._

vocabulary/state-library.md

@@ -0,0 +1,102 @@
+# KYE™ State Library Vocabulary
+
+The **KYE State Library™** is the open-source catalogue of reference state
+machines for KYE Protocol™. This document describes the library model in
+plain terms.
+
+This document defines names and the adoption / derivation model only.
+The signing mechanism, integrity-lock construction, and transparency-log
+inclusion are part of the normative specification.
+
+## What the library is
+
+The KYE State Library™ is a curated set of **reference state machines** —
+pre-built graphs of states and transitions that cover common entity
+lifecycle patterns across regulated industries.
+
+Each library entry is:
+
+- **open source** — published under MIT or Apache-2.0
+- **signed** — the entry JSON is signed; the signature is verifiable
+  against the KYE Protocol™ transparency log
+- **versioned** — entries follow `<name>@<semver>`; minor bumps add
+  states or transitions; major bumps are breaking and require a new entry
+- **sector-tagged** — each entry declares one or more `sector` values
+  from the list below
+
+## v1 sector coverage
+
+The v1 library ships 30 starter entries across 9 sectors:
+
+| Sector | Stable ID | Example machines |
+|---|---|---|
+| Banking | `banking` | `retail_account.v1`, `corporate_account.v1`, `credit_facility.v1` |
+| Payments | `payments` | `payment_mandate.v1`, `payment_instruction.v1` |
+| Insurance | `insurance` | `policy_holder.v1`, `claim.v1` |
+| Healthcare | `healthcare` | `patient.v1`, `clinical_record.v1` |
+| Pharma | `pharma` | `trial_participant.v1`, `batch_release.v1` |
+| Logistics | `logistics` | `shipment.v1`, `warehouse_asset.v1` |
+| Energy | `energy` | `grid_asset.v1`, `operator_credential.v1` |
+| Regtech | `regtech` | `regulated_entity.v1`, `compliance_obligation.v1` |
+| AI Governance | `ai_governance` | `model.v1`, `ai_agent.v1`, `audit_stream.v1` |
+
+## Adoption
+
+A tenant **adopts** a library entry by setting `state_machine_id` on an
+entity (or entity class) to the library entry's `kye:sm:…` URN and
+recording the `state_machine_version`. Adoption is a point-in-time
+snapshot — the entity continues to use that version even when the library
+publishes a newer minor version.
+
+Upgrading to a newer version is an explicit operation that adds a
+`state_event` recording the version change.
+
+## Derivation
+
+A tenant may **derive** a tightened variant from a library entry:
+
+1. The derived machine references the parent entry via `parent_machine_id`.
+2. Derivation may only **add** states, **add** guard conditions, or
+   **restrict** allowed transitions. It may not remove terminal classes
+   or remove evidence requirements.
+3. The derived machine is signed by the tenant. The signature is recorded
+   in the tenant's State Registry™.
+4. The derivation delta (the set of additions / restrictions) is stored
+   as a separate signed record alongside the derived machine.
+
+Derivation does **not** transfer the parent library entry's open-source
+licence to proprietary use — the derived delta is the tenant's own work;
+the base entry remains under its original licence.
+
+## Integrity lock
+
+Once a library entry is published (signed + included in the transparency
+log), its content is immutable. A published entry is never edited in
+place. Corrections or additions produce a new semver entry. This ensures
+that any tenant that has adopted `foo@1.2.3` has an unchanging contract
+for the lifetime of that version.
+
+## Naming convention
+
+Library entry IDs follow the pattern:
+
+```
+kye:sm:kyeprotocol.com:lib.<sector>.<name>.<major>
+```
+
+Examples:
+
+```
+kye:sm:kyeprotocol.com:lib.banking.retail_account.v1
+kye:sm:kyeprotocol.com:lib.ai_governance.model.v1
+kye:sm:kyeprotocol.com:lib.healthcare.patient.v1
+```
+
+## What this document does not specify
+
+- the signing algorithm or key management for library entries
+- how library entries are verified against the transparency log
+- the process for submitting new library entries
+- the runtime enforcement of derivation constraints
+
+Those are part of the normative specification.

vocabulary/state-machines.md

@@ -0,0 +1,102 @@
+# KYE™ State Machine Vocabulary
+
+Every KYE Protocol™ entity carries a `state` field governed by a registered
+state machine. This document describes the **state model** in plain terms.
+
+This document defines names and concepts only. The transition-guard
+evaluation logic, evidence-class matching, and audit-chain construction
+are part of the normative specification and are not published here.
+
+## The state model
+
+A **state machine** in KYE Protocol™ is a named, versioned, immutable
+graph of states and transitions. Each machine:
+
+- has a stable `state_machine_id` (a `kye:sm:…` URN)
+- declares a set of **states** — each with a human label and a
+  stability class (`transient`, `stable`, `terminal`)
+- declares a set of **transitions** — each with a `from` state, a
+  `to` state, and an **evidence class** that must be satisfied
+
+An entity record holds:
+
+- `state` — the current state value (string, from the machine's state set)
+- `state_machine_id` — which machine governs this entity
+- `state_machine_version` — the semver of the machine at adoption time
+
+## State event log
+
+Every transition appends a `state_event` record. The record is
+append-only — it is never updated or deleted. Fields:
+
+| Field | Description |
+|---|---|
+| `entity_id` | The entity that transitioned |
+| `from_state` | State before the transition |
+| `to_state` | State after the transition |
+| `transition_id` | Identifies the declared transition in the machine |
+| `triggered_by` | The principal or system that triggered the event |
+| `evidence_ref` | Reference to the evidence pack or decision record that satisfied the guard |
+| `recorded_at` | ISO 8601 timestamp (UTC) |
+
+## Stability classes
+
+| Class | Meaning |
+|---|---|
+| `transient` | Short-lived; the machine expects the entity to leave this state quickly (e.g., `pending_verification`) |
+| `stable` | Normal operating state (e.g., `active`, `limited`) |
+| `terminal` | No outbound transitions (e.g., `tombstoned`) |
+
+## Common state names
+
+The following names are pre-defined in the KYE Protocol™ core vocabulary.
+State machines MAY use any subset and MAY add custom states via derivation.
+
+| State | Stability class | Meaning |
+|---|---|---|
+| `discovered` | transient | Entity seen but not yet registered |
+| `registered` | transient | Registered with an immutable ID; pending first verification |
+| `pending_verification` | transient | Awaiting evidence to advance |
+| `verified` | stable | Verification satisfied |
+| `active` | stable | Participating in governed actions |
+| `limited` | stable | Operating under reduced scope |
+| `under_review` | transient | Review in progress |
+| `approval_required` | transient | Actions require explicit approval |
+| `suspended` | stable | Temporarily prevented from acting |
+| `stopped` | stable | Prevented from acting by a stop-cascade event |
+| `quarantined` | stable | Isolated pending investigation |
+| `revoked` | stable | Authority removed; recovery requires explicit process |
+| `transferred` | terminal | Ownership or trust domain changed |
+| `archived` | terminal | Retained for audit; no active participation |
+| `tombstoned` | terminal | Permanently deactivated; no outbound transitions |
+
+## Machine identity convention
+
+State machine IDs follow the pattern:
+
+```
+kye:sm:<trust-domain>:<name>.<entity-class>.<major>
+```
+
+Examples:
+
+```
+kye:sm:kyeprotocol.com:core.principal.v1
+kye:sm:kyeprotocol.com:core.workspace.v1
+kye:sm:kyeprotocol.com:lib.banking.account.v1
+kye:sm:acme.example:derived.principal.v1
+```
+
+The `core.*` machines are published by KYE Protocol™ and available in
+the KYE State Library™. Tenant-derived machines use the tenant's trust
+domain and record `parent_machine_id` pointing to the library entry they
+derived from.
+
+## What this document does not specify
+
+- the algorithm that evaluates evidence-class guards
+- the signing or integrity mechanism for state events
+- the cross-entity propagation rules when a parent state changes
+- the recovery process after `revoked` or `stopped`
+
+Those are part of the normative specification.

vocabulary/taxonomies-v1.md

@@ -0,0 +1,64 @@
+# KYE Protocol™ — Canonical taxonomies (V1)
+
+> **Status:** Normative. v1.0. Apache 2.0. Companion to
+> the 17 canonical dictionaries (`internal`) (post-§29 landing).
+>
+> KYE is **taxonomy-first**. Authority decisions need shared
+> classification or every implementation invents different words for
+> the same thing. V1 ships with 16 canonical taxonomies. Profiles MAY
+> register additional taxonomies but MUST NOT redefine canonical
+> terms.
+
+## 1. The 16 V1 taxonomies
+
+| # | Taxonomy ID                         | Purpose                                            |
+|---|-------------------------------------|----------------------------------------------------|
+| 1 | `kye:taxonomy:entity_type`          | Entity classes — humans, businesses, agents, services, models, tools, workflows. |
+| 2 | `kye:taxonomy:entity_state`         | Lifecycle states — provisional, active, suspended, quarantined, stopped, tombstoned. |
+| 3 | `kye:taxonomy:capability_type`      | Skill, tool, MCP tool, function, connector, playbook, model_profile, payment_action, api_operation. |
+| 4 | `kye:taxonomy:capability_state`     | active, deprecated, broken, quarantined, revoked, superseded. |
+| 5 | `kye:taxonomy:action_type`          | Verbs the runtime evaluates (read, write, transfer, payment.prepare, …). |
+| 6 | `kye:taxonomy:resource_type`        | Wallet, dataset, document, vehicle, vessel, shipment. |
+| 7 | `kye:taxonomy:data_class`           | Personal, financial, special-category, regulated, public. |
+| 8 | `kye:taxonomy:side_effect_level`    | read / write / move_money / external_send / mutate_critical. |
+| 9 | `kye:taxonomy:risk_state`           | nominal / elevated / watch / denylisted.            |
+| 10| `kye:taxonomy:environment`          | dev / staging / production / regulated_production.  |
+| 11| `kye:taxonomy:decision`             | allow / allow_with_constraints / require_approval / require_step_up / require_human_review / require_recovery / quarantine / deny. |
+| 12| `kye:taxonomy:reason_code`          | Canonical reason codes referenced by every decision. |
+| 13| `kye:taxonomy:evidence_type`        | Audit event, transparency receipt, evidence pack, payload artefact. |
+| 14| `kye:taxonomy:compliance_framework` | EU_AI_ACT, ISO_42001, SOC2, ISO_27001, PCI_DSS, PSD3, DORA, NIS2, HIPAA, NIST_800_207. |
+| 15| `kye:taxonomy:sector`               | Healthcare, financial_services, defence, energy, manufacturing, automotive, maritime, logistics, aviation, public_sector. |
+| 16| `kye:taxonomy:jurisdiction`         | ISO-3166 jurisdictions plus EU as a regulatory bloc. |
+
+## 2. Required signal types
+
+Every taxonomy and metadata mutation MUST emit:
+
+| Event                                | Emitted on                          |
+|--------------------------------------|-------------------------------------|
+| `kye.taxonomy.created`               | New taxonomy registered             |
+| `kye.taxonomy.updated`               | Taxonomy metadata changed           |
+| `kye.taxonomy.published`             | Version moved from draft to active  |
+| `kye.taxonomy.deprecated`            | Term or version deprecated          |
+| `kye.taxonomy.term_added`            | New term added to a taxonomy        |
+| `kye.taxonomy.term_updated`          | Term metadata or status changed     |
+| `kye.taxonomy.term_deprecated`       | Term marked deprecated              |
+| `kye.taxonomy.mapping_created`       | Taxonomy term mapped to framework   |
+| `kye.metadata.schema_created`        | New `KYEMetadataSchema` registered  |
+| `kye.metadata.schema_updated`        | Schema fields or constraints changed |
+| `kye.metadata.validation_failed`     | A metadata block failed validation  |
+| `kye.metadata.classification_changed`| Classification field changed        |
+| `kye.metadata.binding_created`       | Metadata bound to a subject URN     |
+| `kye.metadata.binding_removed`       | Metadata binding removed            |
+
+## 3. Use in copy
+
+> **No authority without context.** KYE uses taxonomies and metadata
+> to make every entity, capability, action, resource, risk, and
+> evidence object governable.
+>
+> KYE is **taxonomy-first** and **metadata-first**, so every authority
+> decision understands the entity, capability, resource, action, risk,
+> sector, jurisdiction, state, and evidence context behind it.
+
+— KYE Protocol™ project, 2026.

vocabulary/trademarks.md

@@ -0,0 +1,156 @@
+# KYE Protocol™ — protected trademark dictionary (SSOT)
+
+This is the **single source of truth** for every protected mark of the
+KYE Protocol™ project. Every public-facing surface (this site, the
+whitepaper, the legal page, READMEs, blog posts, SDK docs) must carry
+the ™ symbol on the **first prominent use** of any of these marks
+within a single page or document. Subsequent uses on the same page
+may drop the symbol.
+
+The audit script `scripts/audit-trademarks.mjs` enforces this against
+`public/site/*.html` on every `npm test` run, so first-prominent-use
+gaps cannot land on `main` undetected.
+
+The structured form of this list is at
+[`trademarks.json`](trademarks.json) — that file is what the audit
+script reads.
+
+## Protected marks (15)
+
+| # | Mark | Class | Notes |
+|---|---|---|---|
+| 1 | **KYE™** | wordmark | Short form; standalone references |
+| 2 | **KYE Protocol™** | wordmark | Full name of the protocol |
+| 3 | **Know Your Entity™** | descriptive wordmark | Source of the KYE acronym |
+| 4 | **Authority Finality™** | concept mark | The protocol-level *property*: an action's chain is provably terminal — six bound fields signed and replayable |
+| 4a | **KYE Chain of Authority™** | concept mark | The *structural artefact*: the linked, attenuating delegation chain from root principal down to the actor. Narrative parallel to chain-of-custody for documents. KYE prefix mitigates §2(e) descriptiveness refusal on the bare phrase. |
+| 5 | **Authority Graph™** | concept mark | Graph projection of authority |
+| 6 | **Decision Map™** | concept mark | Per-decision projection |
+| 7 | **Evidence Graph™** | concept mark | Evidence projection |
+| 8 | **Blast Radius Map™** | concept mark | Compromise-cascade projection |
+| 9 | **Compliance Map™** | concept mark | Framework-control projection |
+| 10 | **KYE Compliance Mapping Rail™** | infrastructure mark | Runtime → control binding |
+| 11 | **KYE Cloud Gateway™** | commercial mark | Hosted commercial offering |
+| 11a | **KYE Reference Gateway™** | implementation mark | The protocol's open-source reference implementation. Ships under commercial licence to design partners. |
+| 11b | **KYE Runtime Gateway™** | implementation mark | Generic name for any production-grade KYE™ Gateway deployment built on the reference. |
+| 12 | **KYE Conformant™** | certification mark | L3 conformance badge |
+| 13 | **KYE Certified™** | certification mark | L4 certification badge |
+| 14 | **KYE Self-Tested™** | certification mark | L1 self-test badge |
+| 15 | **KYE Self-Attested™** | certification mark | L2 self-attestation badge |
+| 16 | **KYE Core Conformant™** | certification mark (capability-scoped) | Core authority + delegation + scope + state + audit + decision-code fixtures. Program in development. |
+| 17 | **KYE Authority Conformant™** | certification mark (capability-scoped) | Authority Finality™ binding tested end-to-end. Program in development. |
+| 18 | **KYE Capability Conformant™** | certification mark (capability-scoped) | Capability registry + grant + invoke + quarantine + cascade fixtures. Program in development. |
+| 19 | **KYE Evidence Conformant™** | certification mark (capability-scoped) | Evidence Pack™ generation, signing, replay and OSCAL projection. Program in development. |
+| 20 | **KYE Sovereign AI Profile™** | profile | Umbrella connector profile family for sovereign / national AI ecosystems. v1.1 preview. |
+| 21 | **KYE Public Sector Profile™** | profile | Agency + civil-servant + citizen-service authority binding. v1.1 preview. |
+| 22 | **KYE National AI Agent Profile™** | profile | Registry + capability + permitted-action binding for AI agents at national scale. v1.1 preview. |
+| 23 | **KYE Government API Authority Profile™** | profile | Pre-authorisation hook for government APIs. v1.1 preview. |
+| 24 | **KYE Public Services Evidence Profile™** | profile | Evidence-pack format for citizen-impacting AI-assisted decisions. v1.1 preview. |
+| 25 | **KYE Sovereign Data Access Profile™** | profile | Dataset + data-space authority + provenance binding. v1.1 preview. |
+| 26 | **KYE Sovereign Model Registry Profile™** | profile | Model entity + risk state + permitted-action binding. v1.1 preview. |
+| 27 | **KYE Cross-Agency Delegation Profile™** | profile | Ministerial + agency + regulatory delegation binding. v1.1 preview. |
+| 28 | **KYE National Sandbox Profile™** | profile | Supervisor + participant authority binding for regulated sandboxes. v1.1 preview. |
+| 29 | **KYE Sovereign AI Authority Gateway™** | app | Runtime authority gateway for sovereign AI ecosystems. Planned commercial app. |
+| 30 | **KYE National AI Agent Registry™** | app | Register and govern AI agents, models, datasets at national scale. Planned commercial app. |
+| 31 | **KYE Public Sector Decision Map™** | app | Replay AI-assisted public-sector decisions. Planned commercial app. |
+| 32 | **KYE Sovereign AI Evidence Pack™** | app | Evidence-pack format + viewer for sovereign AI ecosystems. Planned commercial app. |
+| 33 | **KYE National Sandbox Authority Harness™** | app | Sandbox harness for regulated national AI / open-finance / healthcare programs. Planned commercial app. |
+| 34 | **KYE Authority Wallet™** | app | Mobile authority-control + proof app for AI-agent actions. Create agents, grant scoped authority, approve / deny / revoke, inspect Decision Map™, view evidence packs. |
+| 35 | **KYE Authority Wallet Demo™** | app | Proprietary mobile demo (Expo / React Native). Synthetic data, simulated agent / payment flows. |
+| 36 | **KYE Authority Wallet Pro™** | commercial | Paid production version. Real Gateway integration, enterprise approval workflows, push notifications, evidence archive. Planned. |
+| 37 | **KYE Mobile Approval Gateway™** | commercial | Enterprise mobile-approval module for banks, payments, healthcare, legal. Push-approval-as-part-of-decision. Planned. |
+| 38 | **KYE Continuity Profile™** | profile | Preserves the chain between intent, interpretation, delegated authority, state, decision, execution and evidence. v1.0. |
+| 39 | **Authority Continuity™** | concept | Authority remained continuous from intent to action. |
+| 40 | **Agency Continuity™** | concept | Same property in agentic-systems language. |
+| 41 | **Continuity Decision Map™** | concept | Per-continuity-decision projection. |
+| 42 | **Continuity Evidence Pack™** | concept | Replayable bundle for one continuity check. |
+| 43 | **Delegated Agency Graph™** | concept | Cross-chain visualisation. |
+| 44 | **KYE Continuity Gateway™** | app | Runtime gateway for continuity checks. Planned. |
+| 45 | **KYE Intent Trace App™** | app | Signal → action visualiser. Planned. |
+| 46 | **KYE Agency Drift Monitor™** | app | Drift detection + alerting. Planned. |
+| 47 | **KYE Discoverability Profile™** | profile | Policy-filtered discovery of KYE entities, authorities, capabilities, decisions, evidence, profiles, connectors. v1.0. |
+| 48 | **KYE Authority Directory™** | app | Searchable directory of entities, agents, capabilities, grants and profiles. Planned. |
+| 49 | **KYE Discovery Console™** | app | Enterprise search + investigation console. Planned. |
+| 50 | **KYE Authority Path Finder™** | app | Find principal → delegation → actor → capability paths. Planned. |
+| 51 | **KYE Evidence Finder™** | app | Search and verify decisions, audit events and evidence packs. Planned. |
+| 52 | **KYE Connector Discovery Hub™** | app | Discover profiles, connectors, apps, plugins and certified implementations. Planned. |
+| 53 | **KYE Ontology Profile™** | profile | Semantic layer — shared meaning of entities, authorities, capabilities, scopes, states, decisions, evidence, profiles, connectors and sectors. v1.0. |
+| 54 | **KYE Semantic Layer™** | concept | Product-friendly synonym for KYE Ontology Profile™. |
+| 55 | **KYE Ontology Registry™** | app | Managed registry for terms, relationships, mappings and semantic assertions. Planned. |
+| 56 | **KYE Semantic Authority Mapper™** | app | Maps OAuth scopes, IAM roles, payment mandates, legal delegations, healthcare consents into KYE’s authority model without losing meaning. Planned. |
+| 57 | **KYE Semantic Graph™** | app | Graph view of the ontology + live instances; semantic-path search and risk-ranked traversal. Planned. |
+| 58 | **KYE Ontology Conformance™** | app | Conformance fixture suite + certification track for ontology-correct implementations. Planned. |
+| 59 | **KYE Operating Model Profile™** | profile | Enterprise adoption layer — journey from use-case intake to runtime authority control. v1.0. |
+| 60 | **KYE AI Worker Readiness App™** | app | Readiness assessment for AI workers — pilot / controlled / production. Planned. |
+| 61 | **KYE Entity Authority Record™** | concept | Living governance record per delegated actor. |
+| 62 | **KYE Authority Gates™** | concept | Runtime gates before high-impact AI actions. |
+| 63 | **KYE Authority Gate Designer™** | app | Visual gate composer + simulator. Planned. |
+| 64 | **KYE Commit Boundary™** | concept | Separates recommendation from committed action. |
+| 65 | **KYE Commit Boundary Monitor™** | app | Runtime monitor for recommendation→committed-action transitions. Planned. |
+| 66 | **KYE Governed Entity Catalog™** | app | Discovery surface for governed agents, models, tools, services, connectors, apps and certified implementations. Planned. |
+| 67 | **KYE Cloud Portal™** | app | Enterprise dashboard for the operating-model journey + runtime evidence. Planned. |
+| 68 | **KYE Academy™** | app | Training and certification track for KYE Operating Model Profile™ owners and operators. Planned. |
+| 69 | **KYE Assurance Card Profile™** | profile | Lifecycle assurance layer that turns runtime evidence into a living assurance record. v1.0. |
+| 70 | **KYE Assurance Card™** | concept | Living lifecycle assurance record per delegated entity. |
+| 71 | **KYE Assurance Card Builder™** | app | Generate living assurance cards from KYE runtime objects. Planned. |
+| 72 | **KYE Assurance Card Library™** | app | Discoverable library of governed AI use cases + assurance patterns. Planned. |
+| 73 | **KYE Human Involvement Planner™** | app | Define where authorised humans must review, approve, limit or override AI-enabled actions. Planned. |
+| 74 | **KYE Human Involvement Plan™** | concept | Schema-bound record of human-judgement points across the lifecycle of a delegated AI system. |
+| 75 | **KYE Provenance & Supply Chain Evidence App™** | app | Track datasets, models, tools, suppliers, licences and hardware lineage. Planned. |
+| 76 | **KYE AI Supply Chain & Provenance Profile™** | concept | Provenance + supply-chain dimension of an assurance card subject. |
+| 77 | **KYE Assurance Review Scheduler™** | app | Scheduled + event-triggered + scope-change + incident + retention + decommissioning review cycles. Planned. |
+| 78 | **KYE Assurance Review Cycle™** | concept | Schema-bound review lifecycle for an assurance-card subject. |
+| 79 | **KYE Governed Use Case Library™** | app | Sector-curated library of governed AI use cases linking authority records, gates, commit boundaries, decisions, evidence packs and assurance cards. Planned. |
+| 80 | **KYE Formal Rules Profile™** | profile | Rights, obligations and governance layer — permissions, obligations, prohibitions, powers, exceptions and governance meta-rules as machine-readable authority objects. v1.0. |
+| 81 | **KYE Rights & Obligations Engine™** | app | Runtime engine evaluating permissions, obligations, prohibitions, powers, exceptions and governance rules. Planned. |
+| 82 | **KYE Obligation Ledger™** | app | Append-only ledger tracking every obligation lifecycle. Planned. |
+| 83 | **KYE Rule Prover™** | app | Pre-runtime consistency checker for rule sets. Planned. |
+| 84 | **KYE Rule Compiler™** | app | Compiles formal KYE rules into runtime PDP/PEP policies, authority gates, commit boundaries, signal events and evidence requirements. Planned. |
+| 85 | **KYE Contract-to-Authority Mapper™** | app | Extracts permissions, obligations, prohibitions and powers from contracts into KYE formal rules. Planned. |
+| 86 | **KYE Action Admissibility Profile™** | profile | Upstream pre-action layer — checks whether a proposed action is admissible into the authority pipeline before any authority / formal-rule / commit-boundary check runs. v1.0. |
+| 87 | **KYE Admissibility Engine™** | app | Runtime engine performing the admission check across intent, source, data, policy and rule dimensions. Planned. |
+| 88 | **KYE Admission Gate™** | concept | The gate object the Admissibility Engine evaluates, placed UPSTREAM of authority gates. |
+| 89 | **KYE Pre-Action Filter™** | concept | Product-friendly synonym for the admission surface. |
+
+## Markup convention
+
+In HTML, the canonical first-prominent-use markup is:
+
+```html
+KYE Protocol<span class="tm">™</span>
+```
+
+Plain `™` is also accepted (the audit script counts both). The
+`<span class="tm">` wrapper is preferred for reduced font size on
+the symbol.
+
+## What the audit checks
+
+For each HTML file under `public/site/`, the audit:
+
+1. Finds the **first occurrence** of each mark (case-sensitive).
+2. Verifies that occurrence is immediately followed by either
+   `<span class="tm">™</span>` or `™` (within ~50 characters,
+   to account for nested tags or whitespace).
+3. Skips files where the mark never appears.
+4. Reports any first-prominent use without ™ as a hard failure.
+
+The script is intentionally strict: subsequent uses on the same
+page are not checked, but the first one must be ™-marked.
+
+## When you add a new protected mark
+
+1. Add the mark to `trademarks.json` (and refresh the table above).
+2. Update the universal footer-bottom notice in every page under
+   `public/site/` (search for `are trademarks of the KYE Protocol`).
+3. Update the legal page §2 (`public/site/legal.html#trademark`)
+   and the legal-faq (`public/site/legal-faq.html#trademark-faq`).
+4. Run `npm run test:trademarks` locally to confirm the new mark
+   is correctly flagged everywhere it appears.
+
+---
+
+This file is published under Apache License 2.0 with the rest of the
+public protocol surface. The marks themselves remain trademarks of
+the KYE Protocol™ project; reproducing this dictionary does not
+grant any right to use the marks beyond the [trademark policy](../site/legal.html#trademark).