# Eve SOTA gap → task → evidence → exit matrix — 2026-09
**Initiative:** `eve-sota-gap-closure-2026-09-01`
**Machine record:**
[`eve-sota-gap-task-evidence-exit/2026-09-01.json`](eve-sota-gap-task-evidence-exit/2026-09-01.json)
**Schema:**
[`eve-sota-gap-task-evidence-exit.schema.json`](eve-sota-gap-task-evidence-exit.schema.json)
**Required verifier:**
[`../../tools/eve-everywhere/verify-gap-task-evidence-exit.mjs`](../../tools/eve-everywhere/verify-gap-task-evidence-exit.mjs)
This record assigns all 211 ledger tasks exactly once as an owner and covers all
18 registered gaps. 94 rows currently have direct evidence and 117 remain
planned/open. Planned evidence is an admission contract, not a completion claim.
## Admission command
```bash
node tools/eve-everywhere/verify-gap-task-evidence-exit.mjs
```
The verifier re-parses the authoritative ledger, recomputes both its whole-file
SHA-256 and the pinned exact-task-requirements SHA-256, and refuses missing or
duplicate owners, unknown task/gap references, gap ownership or admission
dependencies that differ from policy, orphan gaps, open prerequisites beneath a
completed owner, self-dependencies, and dependency cycles. Task IDs can occur
structurally only as `ownerTaskId` or `dependencyTaskIds`.
It also applies an explicit, verifier-pinned proof-scope classification to every
task. A newly added task has no admission policy until that classification is
reviewed and added. Required and planned scopes must match exactly, and every
scope is pinned to an evidence class. Source review, static checks, and unit
behavior therefore cannot masquerade as a real model, coding agent, database,
vector store, browser, mobile, native desktop, DCC, engine, external channel,
multimodal fixture, protocol, load/soak, human, or assistive-technology
observation.
## Gap coverage
| Gap | Registered statement | Owning rows |
| --- | ---------------------------------------------------------- | ----------: |
| G1 | Governed planning, orchestration, and fleet throughput | 23 |
| G2 | Retrieval and knowledge quality | 23 |
| G3 | Second-brain fusion, watchers, and channels | 26 |
| G4 | Metis and future workbench seams | 23 |
| G5 | DCC and creative-suite agency | 50 |
| G6 | Native computer-use admission | 23 |
| G7 | Human-grounded judge validation | 25 |
| G8 | Release-scope evaluation activation | 19 |
| G9 | Interaction, trust, accessibility, and cross-client parity | 35 |
| G10 | Agentic security, authority, and injection resistance | 31 |
| G11 | Memory quality, provenance, and data rights | 33 |
| G12 | Goal-to-verified-delivery competence | 44 |
| G13 | Production reliability and recoverability | 32 |
| G14 | Evaluation completeness | 79 |
| G15 | Multimodal, document, and media competence | 87 |
| G16 | Privacy, data governance, and regulatory readiness | 41 |
| G17 | Protocol and ecosystem interoperability | 39 |
| G18 | Model and router lifecycle performance resilience | 40 |
## Proof-boundary policy
| Scope | Boundary | Required evidence class | Meaning |
| ----------------------------- | -------------------- | -------------------------- | ---------------------------------------------------------------------------------------------------- |
| `source-inspection` | repository | `source-review` | Direct review of the authoritative source, generated artifacts, and ownership. |
| `static-contract` | repository | `automated-static` | Machine validation of a schema, contract, registry, policy, or generated record. |
| `automated-behavior` | logic | `automated-test` | Behavioral tests with an observed red control and green regression. |
| `service-integration` | service | `service-integration` | Integration or contract proof across the real service/tool boundary. |
| `persistence-recovery` | persistence | `service-integration` | Durable write/read, restart, migration, deletion, or restore behavior. |
| `authorization-isolation` | authorization | `service-integration` | Positive and negative actor, tenant, object, purpose, and scope enforcement. |
| `real-model-provider` | model-provider | `live-runtime` | A retained receipt from the resolved real model/provider route. |
| `real-agent-runtime` | coding-agent | `live-runtime` | A real, version-preflighted coding-agent CLI executes the attributed work lane. |
| `real-database` | database | `live-runtime` | Observation against the real database implementation rather than an in-memory substitute. |
| `real-vector-store` | vector-store | `live-runtime` | Build, query, freshness, isolation, deletion, and recovery on the bound vector/index runtime. |
| `real-browser-ui` | browser | `live-runtime` | A real browser journey through the user-facing surface and backing service. |
| `native-desktop-runtime` | desktop | `live-runtime` | OS-level fixture proof with real app, window, input, permission, and end-state checks. |
| `mobile-runtime` | mobile | `live-runtime` | Device or emulator automation through the actual mobile client boundary. |
| `real-dcc-runtime` | dcc | `live-runtime` | A real DCC process with readback and independently checked editable artifacts. |
| `real-engine-runtime` | engine | `live-runtime` | A real engine executable at the required host path with retained runtime evidence. |
| `external-channel-runtime` | channel | `live-runtime` | A chosen, credentialed external channel with identity, delivery, safety, and kill proof. |
| `real-multimodal-runtime` | media | `live-runtime` | Real document, image, audio, video, OCR, or generated-media fixtures cross their operating boundary. |
| `protocol-interop` | protocol | `live-runtime` | Conformance against a real independent protocol implementation where adoption is claimed. |
| `host-capability-observation` | host | `live-runtime` | Fresh host probe of the executable, permission, credential-name, or harness capability. |
| `operational-load-soak` | operations | `operational-exercise` | Supervised load, soak, queue drain, or game-day evidence at the operating boundary. |
| `fault-recovery` | recovery | `automated-test` | Injected timeout, cancellation, dependency loss, replay, partial result, or restart failure. |
| `security-adversarial` | security | `security-exercise` | Adversarial and benign-paired tests through the authoritative prompt/tool/runtime seam. |
| `privacy-data-rights` | privacy | `governance-review` | Classification, minimization, provider-use, retention, deletion, and rights evidence. |
| `performance-quality` | measurement | `measurement` | Pre-registered quality, latency, reliability, resource, or cost measurement. |
| `governance-decision` | governance | `governance-review` | An attributable ADR, risk acceptance, legal applicability, or operator decision record. |
| `human-outcomes` | human-evaluation | `human-review` | Independent human labels, rubric results, or operator outcome evidence. |
| `manual-assistive-tech` | assistive-technology | `human-review` | Named screen-reader or assistive-technology manual interoperability evidence. |
| `independent-verification` | verification | `independent-verification` | A verifier independent of the implementer checks the actual end state or artifact. |
| `docs-render-integrity` | documentation | `automated-static` | Markdown/rendered-HTML parity, freshness, links, and source-path integrity. |
## Owner matrix
The exact requirement and full exit sentence are retained in the machine record.
The compact view below exposes every owner, admission edge, proof boundary, and
evidence state for review.
| Owner | Gaps | State | Dependencies | Required proof scopes | Evidence |
| ----- | ------------------------------------------------------------------------------- | --------- | --------------------------------------------------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ---------------- |
| 0.1 | G1, G3, G8, G9, G10, G11 | completed | — | `source-inspection`, `static-contract`, `docs-render-integrity` | available-direct |
| 0.2 | G1, G2, G3, G4, G5, G6, G7, G8, G9, G10, G11, G12, G13, G14, G15, G16, G17, G18 | completed | 0.1 | `source-inspection`, `static-contract`, `host-capability-observation`, `performance-quality` | available-direct |
| 0.3 | G1, G2, G3, G4, G5, G6, G7, G8, G9, G10, G11, G12, G13, G14, G15, G16, G17, G18 | completed | 0.2 | `source-inspection`, `static-contract`, `governance-decision` | available-direct |
| 0.4 | G1, G2, G3, G4, G5, G6, G7, G8, G9, G10, G11, G12, G13, G14, G15, G16, G17, G18 | completed | 0.2 | `source-inspection`, `static-contract`, `governance-decision` | available-direct |
| 0.5 | G1, G2, G3, G4, G5, G6, G7, G8, G9, G10, G11, G12, G13, G14, G15, G16, G17, G18 | completed | 0.3, 0.4 | `source-inspection`, `static-contract`, `automated-behavior` | available-direct |
| 0.6 | G1, G2, G3, G4, G5, G6, G7, G8, G9, G10, G11, G12, G13, G14, G15, G16, G17, G18 | completed | 0.5 | `source-inspection`, `static-contract`, `automated-behavior` | available-direct |
| 0.7 | G2, G3, G5, G6, G11, G13, G14, G15, G16, G17, G18 | completed | 0.2, 0.6 | `source-inspection`, `static-contract`, `host-capability-observation` | available-direct |
| 0.8 | G1, G2, G3, G4, G5, G6, G7, G8, G9, G10, G11, G12, G13, G14, G15, G16, G17, G18 | completed | 0.3, 0.4, 0.5, 0.6, 0.7 | `source-inspection`, `static-contract`, `automated-behavior`, `governance-decision`, `human-outcomes` | available-direct |
| 1.1 | G4 | completed | 0.6 | `source-inspection`, `static-contract`, `privacy-data-rights` | available-direct |
| 1.2 | G4 | completed | 1.1 | `source-inspection`, `static-contract`, `governance-decision` | available-direct |
| 1.3 | G4 | completed | 1.1, 1.2 | `source-inspection`, `static-contract`, `service-integration` | available-direct |
| 1.4 | G4 | completed | 1.3 | `source-inspection`, `automated-behavior`, `service-integration`, `authorization-isolation`, `fault-recovery` | available-direct |
| 1.5 | G4 | completed | 1.1, 1.2, 1.4 | `source-inspection`, `static-contract`, `automated-behavior`, `service-integration`, `authorization-isolation`, `fault-recovery` | available-direct |
| 1.6 | G4 | completed | 1.3, 1.4, 1.5 | `source-inspection`, `static-contract`, `automated-behavior`, `real-model-provider`, `performance-quality` | available-direct |
| 1.7 | G4 | completed | 1.3, 1.4, 1.5, 1.6 | `source-inspection`, `static-contract`, `automated-behavior`, `service-integration`, `authorization-isolation`, `fault-recovery`, `security-adversarial` | available-direct |
| 1.8 | G4 | completed | 1.1, 1.7 | `source-inspection`, `static-contract`, `automated-behavior`, `service-integration` | available-direct |
| 1.9 | G4 | completed | 0.8, 1.8 | `source-inspection`, `static-contract`, `automated-behavior`, `governance-decision` | available-direct |
| 2.1 | G1, G12 | completed | 0.6 | `source-inspection`, `static-contract`, `governance-decision` | available-direct |
| 2.2 | G1, G12 | completed | 2.1 | `source-inspection`, `static-contract`, `governance-decision` | available-direct |
| 2.3 | G1, G12 | completed | 2.2 | `source-inspection`, `automated-behavior`, `service-integration`, `persistence-recovery`, `host-capability-observation`, `fault-recovery` | available-direct |
| 2.4 | G1, G12 | completed | 2.2, 2.3 | `source-inspection`, `automated-behavior`, `persistence-recovery`, `fault-recovery` | available-direct |
| 2.5 | G1, G12 | completed | 2.2, 2.3, 2.4 | `source-inspection`, `automated-behavior`, `service-integration`, `real-browser-ui`, `performance-quality` | available-direct |
| 2.6 | G1, G12 | completed | 2.2, 2.3 | `source-inspection`, `automated-behavior`, `authorization-isolation`, `fault-recovery`, `security-adversarial` | available-direct |
| 2.7 | G1, G12 | completed | 2.1, 2.2, 2.3, 2.4, 2.5, 2.6 | `source-inspection`, `automated-behavior`, `persistence-recovery`, `fault-recovery`, `security-adversarial` | available-direct |
| 2.8 | G1, G12 | open | 2.7, 4.8, 12.7, 13.7 | `source-inspection`, `service-integration`, `real-agent-runtime`, `operational-load-soak`, `fault-recovery`, `security-adversarial`, `performance-quality`, `independent-verification` | planned |
| 3.1 | G2, G18 | open | 0.6 | `source-inspection`, `static-contract`, `security-adversarial`, `privacy-data-rights`, `human-outcomes` | planned |
| 3.2 | G2, G18 | completed | 0.6, 15.1 | `source-inspection`, `static-contract`, `real-model-provider`, `fault-recovery`, `privacy-data-rights`, `performance-quality`, `governance-decision` | available-direct |
| 3.3 | G2, G18 | completed | 3.2 | `source-inspection`, `static-contract`, `persistence-recovery`, `authorization-isolation`, `real-vector-store`, `privacy-data-rights` | available-direct |
| 3.4 | G2, G18 | completed | 3.3 | `source-inspection`, `automated-behavior`, `service-integration`, `real-vector-store`, `fault-recovery` | available-direct |
| 3.5 | G2, G18 | open | 3.1, 3.4 | `source-inspection`, `automated-behavior`, `real-vector-store`, `performance-quality`, `human-outcomes` | planned |
| 3.6 | G2, G18 | completed | 3.3, 3.4 | `source-inspection`, `automated-behavior`, `service-integration`, `authorization-isolation`, `real-vector-store`, `fault-recovery`, `security-adversarial`, `privacy-data-rights` | available-direct |
| 3.7 | G2, G18 | open | 3.5, 3.6 | `source-inspection`, `static-contract`, `security-adversarial`, `performance-quality`, `governance-decision` | planned |
| 3.8 | G2, G18 | open | 3.7, 13.6, 14.4, 15.5 | `source-inspection`, `automated-behavior`, `persistence-recovery`, `real-vector-store`, `fault-recovery`, `privacy-data-rights`, `performance-quality` | planned |
| 4.1 | G10, G16, G17 | completed | 0.4 | `source-inspection`, `static-contract`, `governance-decision` | available-direct |
| 4.2 | G10, G16, G17 | completed | 4.1 | `source-inspection`, `static-contract`, `automated-behavior`, `service-integration`, `authorization-isolation`, `security-adversarial` | available-direct |
| 4.3 | G10, G16, G17 | completed | 4.1, 4.2 | `source-inspection`, `static-contract`, `automated-behavior`, `real-model-provider`, `security-adversarial`, `performance-quality` | available-direct |
| 4.4 | G10, G16, G17 | completed | 4.1 | `source-inspection`, `static-contract`, `automated-behavior`, `service-integration`, `authorization-isolation`, `security-adversarial` | available-direct |
| 4.5 | G10, G16, G17 | completed | 4.1, 4.4 | `source-inspection`, `automated-behavior`, `authorization-isolation`, `native-desktop-runtime`, `real-dcc-runtime`, `host-capability-observation`, `fault-recovery`, `security-adversarial` | available-direct |
| 4.6 | G10, G16, G17 | completed | 4.1, 4.4 | `source-inspection`, `static-contract`, `automated-behavior`, `service-integration`, `authorization-isolation`, `protocol-interop`, `fault-recovery`, `security-adversarial`, `privacy-data-rights` | available-direct |
| 4.7 | G10, G16, G17 | completed | 4.4, 4.5 | `source-inspection`, `automated-behavior`, `service-integration`, `authorization-isolation`, `fault-recovery`, `independent-verification` | available-direct |
| 4.8 | G10, G16, G17 | completed | 4.1, 4.2, 4.3, 4.4, 4.5, 4.6, 4.7 | `source-inspection`, `automated-behavior`, `service-integration`, `real-model-provider`, `security-adversarial`, `performance-quality` | available-direct |
| 5.1 | G5, G14, G15 | completed | 0.6 | `source-inspection`, `static-contract` | available-direct |
| 5.2 | G5, G14, G15 | open | 5.1, 15.1 | `source-inspection`, `static-contract`, `automated-behavior`, `service-integration`, `real-model-provider`, `fault-recovery` | planned |
| 5.3 | G5, G14, G15 | completed | 0.7 | `source-inspection`, `automated-behavior`, `real-dcc-runtime`, `host-capability-observation`, `independent-verification` | available-direct |
| 5.4 | G5, G14, G15 | completed | 5.1, 5.3 | `source-inspection`, `static-contract`, `automated-behavior`, `service-integration`, `authorization-isolation`, `security-adversarial` | available-direct |
| 5.5 | G5, G14, G15 | open | 4.8, 5.2, 5.4, 12.7, 13.7, 14.7 | `source-inspection`, `static-contract`, `automated-behavior`, `service-integration`, `authorization-isolation`, `real-dcc-runtime`, `protocol-interop`, `security-adversarial` | planned |
| 5.6 | G5, G14, G15 | open | 5.5 | `source-inspection`, `service-integration`, `real-dcc-runtime`, `independent-verification` | planned |
| 5.7 | G5, G14, G15 | open | 5.6 | `source-inspection`, `automated-behavior`, `persistence-recovery`, `real-dcc-runtime`, `fault-recovery`, `performance-quality`, `independent-verification` | planned |
| 5.8 | G5, G14, G15 | open | 5.5, 5.7 | `source-inspection`, `automated-behavior`, `service-integration`, `real-dcc-runtime`, `fault-recovery`, `security-adversarial`, `independent-verification` | planned |
| 5.9 | G5, G14, G15 | completed | 0.7 | `source-inspection`, `static-contract`, `automated-behavior`, `host-capability-observation`, `privacy-data-rights`, `governance-decision` | available-direct |
| 5.10 | G5, G14, G15 | open | 0.8, 4.8, 5.12, 5.13, 5.14, 5.33, 12.7, 13.7, 14.7 | `source-inspection`, `automated-behavior`, `service-integration`, `authorization-isolation`, `real-agent-runtime`, `real-engine-runtime`, `host-capability-observation`, `fault-recovery`, `security-adversarial`, `privacy-data-rights`, `performance-quality`, `human-outcomes`, `independent-verification` | planned |
| 5.11 | G5, G14, G15 | open | 0.8, 5.8, 5.10, 5.12, 5.15, 5.16, 5.17, 5.18, 5.19 | `source-inspection`, `automated-behavior`, `service-integration`, `authorization-isolation`, `real-model-provider`, `native-desktop-runtime`, `mobile-runtime`, `real-dcc-runtime`, `real-engine-runtime`, `real-multimodal-runtime`, `fault-recovery`, `security-adversarial`, `privacy-data-rights`, `performance-quality`, `human-outcomes`, `independent-verification` | planned |
| 5.12 | G5, G14, G15 | open | 0.8, 5.9 | `source-inspection`, `static-contract`, `host-capability-observation`, `governance-decision`, `human-outcomes` | planned |
| 5.13 | G5, G14, G15 | open | 4.8, 5.12, 12.7, 13.7, 14.7, 15.1 | `source-inspection`, `automated-behavior`, `service-integration`, `authorization-isolation`, `real-engine-runtime`, `fault-recovery`, `security-adversarial`, `independent-verification` | planned |
| 5.14 | G5, G14, G15 | open | 5.13, 5.34, 5.35 | `source-inspection`, `automated-behavior`, `real-engine-runtime`, `fault-recovery`, `performance-quality`, `human-outcomes`, `independent-verification` | planned |
| 5.15 | G5, G14, G15 | open | 4.8, 5.12, 12.7, 13.7, 14.7 | `source-inspection`, `automated-behavior`, `service-integration`, `authorization-isolation`, `real-engine-runtime`, `fault-recovery`, `security-adversarial`, `human-outcomes`, `independent-verification` | planned |
| 5.16 | G5, G14, G15 | open | 4.8, 5.12, 12.7, 13.7, 14.7 | `source-inspection`, `automated-behavior`, `service-integration`, `real-engine-runtime`, `fault-recovery`, `performance-quality`, `human-outcomes`, `independent-verification` | planned |
| 5.17 | G5, G14, G15 | open | 4.8, 5.12, 12.7, 13.7, 14.7, 17.4, 17.5 | `source-inspection`, `automated-behavior`, `service-integration`, `real-model-provider`, `real-engine-runtime`, `real-multimodal-runtime`, `fault-recovery`, `privacy-data-rights`, `performance-quality`, `human-outcomes`, `independent-verification` | planned |
| 5.18 | G5, G14, G15 | open | 4.8, 5.12, 12.7, 13.7, 14.7 | `source-inspection`, `automated-behavior`, `native-desktop-runtime`, `mobile-runtime`, `real-engine-runtime`, `fault-recovery`, `performance-quality`, `human-outcomes`, `independent-verification` | planned |
| 5.19 | G5, G14, G15 | open | 5.12, 5.29 | `source-inspection`, `static-contract`, `governance-decision`, `independent-verification` | planned |
| 5.20 | G5, G14, G15 | open | 5.3 | `source-inspection`, `automated-behavior`, `real-dcc-runtime`, `fault-recovery`, `independent-verification` | planned |
| 5.21 | G5, G14, G15 | open | 5.9, 5.12 | `source-inspection`, `static-contract`, `real-dcc-runtime`, `real-engine-runtime`, `host-capability-observation`, `governance-decision` | planned |
| 5.22 | G5, G14, G15 | open | 4.5, 4.6, 4.7, 5.21 | `source-inspection`, `automated-behavior`, `service-integration`, `authorization-isolation`, `real-dcc-runtime`, `fault-recovery`, `security-adversarial`, `independent-verification` | planned |
| 5.23 | G5, G14, G15 | open | 5.22 | `source-inspection`, `static-contract`, `automated-behavior`, `real-dcc-runtime`, `fault-recovery`, `independent-verification` | planned |
| 5.24 | G5, G14, G15 | open | 5.23, 14.7 | `source-inspection`, `automated-behavior`, `persistence-recovery`, `real-dcc-runtime`, `real-multimodal-runtime`, `fault-recovery`, `privacy-data-rights`, `independent-verification` | planned |
| 5.25 | G5, G14, G15 | open | 5.5, 5.24, 12.10, 14.7 | `source-inspection`, `automated-behavior`, `service-integration`, `authorization-isolation`, `real-model-provider`, `real-dcc-runtime`, `fault-recovery`, `security-adversarial`, `independent-verification` | planned |
| 5.26 | G5, G14, G15 | open | 5.5, 5.13, 14.7, 17.4 | `source-inspection`, `automated-behavior`, `service-integration`, `authorization-isolation`, `real-dcc-runtime`, `real-engine-runtime`, `real-multimodal-runtime`, `fault-recovery`, `privacy-data-rights`, `independent-verification` | planned |
| 5.27 | G5, G14, G15 | open | 5.14, 5.20, 5.26 | `source-inspection`, `automated-behavior`, `persistence-recovery`, `real-dcc-runtime`, `real-engine-runtime`, `fault-recovery`, `independent-verification` | planned |
| 5.28 | G5, G14, G15 | open | 5.24, 5.26, 13.3 | `source-inspection`, `automated-behavior`, `service-integration`, `persistence-recovery`, `real-dcc-runtime`, `real-engine-runtime`, `fault-recovery`, `independent-verification` | planned |
| 5.29 | G5, G14, G15 | open | 5.25, 5.28, 12.9 | `source-inspection`, `automated-behavior`, `authorization-isolation`, `real-dcc-runtime`, `fault-recovery`, `security-adversarial`, `performance-quality`, `independent-verification` | planned |
| 5.30 | G5, G14, G15 | open | 5.13, 5.21 | `source-inspection`, `static-contract`, `automated-behavior`, `real-engine-runtime`, `protocol-interop`, `host-capability-observation`, `governance-decision`, `independent-verification` | planned |
| 5.31 | G5, G14, G15 | open | 5.5, 5.13, 5.30 | `source-inspection`, `automated-behavior`, `service-integration`, `authorization-isolation`, `real-engine-runtime`, `protocol-interop`, `fault-recovery`, `security-adversarial`, `independent-verification` | planned |
| 5.32 | G5, G14, G15 | open | 5.31 | `source-inspection`, `static-contract`, `automated-behavior`, `service-integration`, `authorization-isolation`, `real-engine-runtime`, `security-adversarial`, `performance-quality`, `independent-verification` | planned |
| 5.33 | G5, G14, G15 | open | 5.32, 15.3 | `source-inspection`, `static-contract`, `automated-behavior`, `service-integration`, `real-model-provider`, `real-engine-runtime`, `fault-recovery`, `performance-quality`, `independent-verification` | planned |
| 5.34 | G5, G14, G15 | open | 5.13 | `source-inspection`, `static-contract`, `automated-behavior`, `persistence-recovery`, `real-engine-runtime`, `fault-recovery`, `independent-verification` | planned |
| 5.35 | G5, G14, G15 | open | 5.13 | `source-inspection`, `static-contract`, `automated-behavior`, `real-engine-runtime`, `performance-quality`, `governance-decision`, `independent-verification` | planned |
| 6.1 | G6, G10, G14 | completed | 0.6 | `source-inspection`, `static-contract`, `governance-decision` | available-direct |
| 6.2 | G6, G10, G14 | completed | 6.1 | `source-inspection`, `static-contract`, `automated-behavior`, `service-integration`, `real-model-provider`, `fault-recovery`, `privacy-data-rights` | available-direct |
| 6.3 | G6, G10, G14 | completed | 6.1, 6.2 | `source-inspection`, `static-contract`, `automated-behavior`, `security-adversarial` | available-direct |
| 6.4 | G6, G10, G14 | completed | 0.7, 6.1 | `source-inspection`, `automated-behavior`, `native-desktop-runtime`, `host-capability-observation`, `fault-recovery`, `privacy-data-rights` | available-direct |
| 6.5 | G6, G10, G14 | completed | 6.3, 6.4 | `source-inspection`, `automated-behavior`, `authorization-isolation`, `native-desktop-runtime`, `fault-recovery`, `security-adversarial`, `performance-quality`, `independent-verification` | available-direct |
| 6.6 | G6, G10, G14 | open | 4.8, 6.2, 6.3, 6.4, 6.5, 12.7, 13.7, 14.7 | `source-inspection`, `automated-behavior`, `service-integration`, `authorization-isolation`, `real-model-provider`, `native-desktop-runtime` | planned |
| 6.7 | G6, G10, G14 | open | 6.6 | `source-inspection`, `automated-behavior`, `persistence-recovery`, `real-model-provider`, `native-desktop-runtime`, `fault-recovery`, `security-adversarial`, `performance-quality` | planned |
| 6.8 | G6, G10, G14 | open | 6.7 | `source-inspection`, `automated-behavior`, `real-model-provider`, `native-desktop-runtime`, `security-adversarial` | planned |
| 7.1 | G3, G11, G12, G13, G14, G15, G16, G17, G18 | completed | 0.6 | `source-inspection`, `static-contract` | available-direct |
| 7.2 | G3, G11, G12, G13, G14, G15, G16, G17, G18 | open | 7.1, 12.2 | `source-inspection`, `static-contract`, `automated-behavior`, `authorization-isolation`, `real-model-provider`, `fault-recovery`, `security-adversarial`, `performance-quality` | planned |
| 7.3 | G3, G11, G12, G13, G14, G15, G16, G17, G18 | completed | 7.1 | `source-inspection`, `static-contract`, `persistence-recovery`, `authorization-isolation`, `fault-recovery`, `privacy-data-rights` | available-direct |
| 7.4 | G3, G11, G12, G13, G14, G15, G16, G17, G18 | open | 4.8, 7.2, 7.3, 12.7, 13.7, 14.7 | `source-inspection`, `automated-behavior`, `service-integration`, `persistence-recovery`, `authorization-isolation`, `real-database`, `fault-recovery` | planned |
| 7.5 | G3, G11, G12, G13, G14, G15, G16, G17, G18 | open | 7.4 | `source-inspection`, `automated-behavior`, `service-integration`, `persistence-recovery`, `authorization-isolation`, `real-database`, `fault-recovery`, `security-adversarial` | planned |
| 7.6 | G3, G11, G12, G13, G14, G15, G16, G17, G18 | completed | 7.1 | `source-inspection`, `static-contract`, `authorization-isolation`, `fault-recovery`, `security-adversarial`, `governance-decision` | available-direct |
| 7.7 | G3, G11, G12, G13, G14, G15, G16, G17, G18 | open | 3.7, 7.6, 9.4 | `source-inspection`, `static-contract`, `automated-behavior`, `service-integration`, `persistence-recovery`, `authorization-isolation`, `real-vector-store`, `fault-recovery`, `privacy-data-rights` | planned |
| 7.8 | G3, G11, G12, G13, G14, G15, G16, G17, G18 | open | 4.8, 7.2, 7.5, 7.6, 7.7, 12.7, 13.7, 14.7 | `source-inspection`, `static-contract`, `security-adversarial`, `privacy-data-rights`, `performance-quality`, `governance-decision` | planned |
| 7.9 | G3, G11, G12, G13, G14, G15, G16, G17, G18 | open | 7.6, 14.7, 17.7 | `source-inspection`, `static-contract`, `automated-behavior`, `service-integration`, `persistence-recovery`, `authorization-isolation`, `security-adversarial`, `privacy-data-rights` | planned |
| 7.10 | G3, G11, G12, G13, G14, G15, G16, G17, G18 | open | 7.9, 12.9 | `source-inspection`, `automated-behavior`, `persistence-recovery`, `real-model-provider`, `real-dcc-runtime`, `performance-quality`, `human-outcomes`, `independent-verification` | planned |
| 8.1 | G9 | completed | 0.6 | `source-inspection`, `static-contract` | available-direct |
| 8.2 | G9 | completed | 8.1 | `source-inspection`, `static-contract`, `governance-decision` | available-direct |
| 8.3 | G9 | completed | 8.1, 8.2 | `source-inspection`, `static-contract`, `automated-behavior`, `service-integration`, `real-browser-ui`, `fault-recovery`, `privacy-data-rights` | available-direct |
| 8.4 | G9 | open | 8.1, 8.2 | `source-inspection`, `static-contract`, `automated-behavior`, `service-integration`, `mobile-runtime`, `fault-recovery`, `performance-quality` | planned |
| 8.5 | G9 | open | 8.2 | `source-inspection`, `automated-behavior`, `service-integration`, `persistence-recovery`, `authorization-isolation`, `real-browser-ui`, `mobile-runtime`, `fault-recovery`, `performance-quality`, `independent-verification` | planned |
| 8.6 | G9 | completed | 8.2 | `source-inspection`, `static-contract`, `automated-behavior`, `real-browser-ui`, `security-adversarial`, `performance-quality`, `governance-decision` | available-direct |
| 8.7 | G9 | completed | 8.1, 8.2 | `source-inspection`, `static-contract`, `automated-behavior`, `real-browser-ui`, `fault-recovery`, `privacy-data-rights`, `performance-quality`, `human-outcomes` | available-direct |
| 8.8 | G9 | open | 8.3, 8.4, 8.5, 8.6, 8.7 | `source-inspection`, `static-contract`, `automated-behavior`, `real-browser-ui`, `manual-assistive-tech` | planned |
| 8.9 | G9 | open | 8.1, 8.2, 8.3, 8.4, 8.5, 8.6, 8.7, 8.8, 8.20 | `source-inspection`, `automated-behavior`, `service-integration`, `real-model-provider`, `real-browser-ui`, `mobile-runtime`, `fault-recovery`, `privacy-data-rights`, `human-outcomes`, `independent-verification` | planned |
| 8.10 | G9 | open | 8.2, 8.5, 17.9, 17.12 | `source-inspection`, `automated-behavior`, `service-integration`, `persistence-recovery`, `authorization-isolation`, `real-browser-ui`, `human-outcomes` | planned |
| 8.11 | G9 | open | 8.10, 17.13 | `source-inspection`, `automated-behavior`, `service-integration`, `authorization-isolation`, `real-browser-ui`, `mobile-runtime`, `real-dcc-runtime`, `fault-recovery`, `human-outcomes`, `manual-assistive-tech`, `independent-verification` | planned |
| 8.12 | G9 | completed | 8.1, 8.2 | `source-inspection`, `static-contract`, `governance-decision`, `docs-render-integrity` | available-direct |
| 8.13 | G9 | open | 8.2, 8.3, 8.12, 14.8, 17.1 | `source-inspection`, `automated-behavior`, `service-integration`, `persistence-recovery`, `authorization-isolation`, `real-browser-ui`, `mobile-runtime`, `security-adversarial`, `privacy-data-rights` | planned |
| 8.14 | G9 | open | 8.2, 8.12 | `source-inspection`, `static-contract`, `automated-behavior`, `real-browser-ui`, `mobile-runtime`, `security-adversarial`, `manual-assistive-tech` | planned |
| 8.15 | G9 | open | 8.5, 8.6, 8.14 | `source-inspection`, `static-contract`, `automated-behavior`, `service-integration`, `persistence-recovery`, `authorization-isolation`, `real-browser-ui`, `mobile-runtime`, `security-adversarial`, `human-outcomes` | planned |
| 8.16 | G9 | open | 8.5, 8.12 | `source-inspection`, `automated-behavior`, `service-integration`, `persistence-recovery`, `authorization-isolation`, `real-browser-ui`, `mobile-runtime`, `fault-recovery`, `privacy-data-rights` | planned |
| 8.17 | G9 | open | 8.12, 9.2, 14.8 | `source-inspection`, `automated-behavior`, `service-integration`, `persistence-recovery`, `authorization-isolation`, `real-database`, `real-browser-ui`, `mobile-runtime`, `privacy-data-rights` | planned |
| 8.18 | G9 | open | 8.12, 15.7 | `source-inspection`, `static-contract`, `automated-behavior`, `real-browser-ui`, `mobile-runtime`, `governance-decision` | planned |
| 8.19 | G9 | open | 8.5, 8.15, 13.3 | `source-inspection`, `automated-behavior`, `service-integration`, `persistence-recovery`, `real-browser-ui`, `mobile-runtime`, `fault-recovery` | planned |
| 8.20 | G9 | open | 8.8, 8.13, 8.14, 8.15, 8.16, 8.17, 8.18, 8.19 | `source-inspection`, `automated-behavior`, `real-browser-ui`, `mobile-runtime`, `manual-assistive-tech`, `independent-verification`, `docs-render-integrity` | planned |
| 9.1 | G11 | completed | 0.6 | `source-inspection`, `static-contract`, `automated-behavior`, `real-database`, `fault-recovery` | available-direct |
| 9.2 | G11 | completed | 9.1 | `source-inspection`, `automated-behavior`, `service-integration`, `persistence-recovery`, `authorization-isolation`, `real-database`, `real-browser-ui`, `privacy-data-rights` | available-direct |
| 9.3 | G11 | open | 9.1 | `source-inspection`, `static-contract`, `automated-behavior`, `real-model-provider`, `real-database`, `performance-quality`, `human-outcomes` | planned |
| 9.4 | G11 | completed | 9.1, 9.2 | `source-inspection`, `static-contract`, `automated-behavior`, `service-integration`, `persistence-recovery`, `authorization-isolation`, `fault-recovery`, `security-adversarial` | available-direct |
| 9.5 | G11 | completed | 9.2, 9.4 | `source-inspection`, `automated-behavior`, `service-integration`, `persistence-recovery`, `authorization-isolation`, `real-database`, `real-vector-store`, `fault-recovery`, `security-adversarial`, `privacy-data-rights` | available-direct |
| 9.6 | G11 | open | 3.7, 7.7, 9.4 | `source-inspection`, `automated-behavior`, `service-integration`, `persistence-recovery`, `authorization-isolation`, `real-vector-store`, `fault-recovery`, `privacy-data-rights` | planned |
| 9.7 | G11 | open | 9.2, 9.3, 9.4, 9.5, 9.6 | `source-inspection`, `static-contract`, `automated-behavior`, `real-model-provider`, `real-database`, `real-vector-store`, `fault-recovery`, `security-adversarial`, `privacy-data-rights`, `performance-quality` | planned |
| 10.1 | G8 | completed | 0.6 | `source-inspection`, `static-contract`, `automated-behavior` | available-direct |
| 10.2 | G8 | completed | 10.1 | `source-inspection`, `static-contract`, `automated-behavior` | available-direct |
| 10.3 | G8 | completed | 10.2 | `source-inspection`, `automated-behavior`, `service-integration` | available-direct |
| 10.4 | G8 | open | 10.1, 10.2, 10.3 | `source-inspection`, `static-contract`, `automated-behavior`, `real-model-provider`, `performance-quality` | planned |
| 11.1 | G12 | completed | 0.6 | `source-inspection`, `static-contract`, `security-adversarial`, `privacy-data-rights` | available-direct |
| 11.2 | G12 | completed | 2.1 | `source-inspection`, `static-contract`, `automated-behavior`, `performance-quality` | available-direct |
| 11.3 | G12 | completed | 0.6 | `source-inspection`, `static-contract`, `automated-behavior`, `host-capability-observation`, `security-adversarial`, `privacy-data-rights`, `performance-quality` | available-direct |
| 11.4 | G12 | open | 4.8, 11.2, 11.3 | `source-inspection`, `static-contract`, `automated-behavior`, `security-adversarial`, `privacy-data-rights`, `performance-quality`, `independent-verification` | planned |
| 11.5 | G12 | completed | 11.3, 11.11 | `source-inspection`, `automated-behavior`, `persistence-recovery`, `host-capability-observation`, `fault-recovery`, `security-adversarial` | available-direct |
| 11.6 | G12 | open | 11.1, 11.4, 11.5, 11.12, 12.2 | `source-inspection`, `automated-behavior`, `real-agent-runtime`, `operational-load-soak`, `performance-quality`, `human-outcomes`, `independent-verification` | planned |
| 11.7 | G12 | open | 2.8, 11.6, 11.11 | `source-inspection`, `service-integration`, `persistence-recovery`, `real-agent-runtime`, `operational-load-soak`, `fault-recovery`, `independent-verification` | planned |
| 11.8 | G12 | open | 0.8, 5.11, 11.7, 12.8, 17.6 | `source-inspection`, `service-integration`, `persistence-recovery`, `real-model-provider`, `real-agent-runtime`, `real-database`, `real-browser-ui`, `native-desktop-runtime`, `mobile-runtime`, `real-dcc-runtime`, `real-engine-runtime`, `real-multimodal-runtime`, `operational-load-soak`, `fault-recovery`, `performance-quality`, `human-outcomes`, `independent-verification` | planned |
| 11.9 | G12 | completed | 11.2 | `source-inspection`, `static-contract`, `automated-behavior`, `service-integration`, `persistence-recovery`, `authorization-isolation`, `real-database`, `fault-recovery` | available-direct |
| 11.10 | G12 | completed | 11.9 | `source-inspection`, `static-contract`, `automated-behavior`, `service-integration`, `persistence-recovery`, `authorization-isolation`, `real-database` | available-direct |
| 11.11 | G12 | completed | 2.7, 11.10 | `source-inspection`, `automated-behavior`, `service-integration`, `persistence-recovery`, `authorization-isolation`, `real-agent-runtime`, `real-database`, `fault-recovery`, `independent-verification` | available-direct |
| 11.12 | G12 | open | 11.2, 13.2 | `source-inspection`, `static-contract`, `automated-behavior`, `service-integration`, `privacy-data-rights`, `performance-quality`, `human-outcomes`, `independent-verification` | planned |
| 12.1 | G7, G14 | completed | 0.8 | `source-inspection`, `static-contract` | available-direct |
| 12.2 | G7, G14 | completed | 0.3, 12.1 | `source-inspection`, `static-contract`, `governance-decision` | available-direct |
| 12.3 | G7, G14 | open | 0.3 | `source-inspection`, `automated-behavior`, `real-model-provider`, `performance-quality`, `human-outcomes` | planned |
| 12.4 | G7, G14 | completed | 12.2 | `source-inspection`, `static-contract`, `automated-behavior`, `fault-recovery`, `performance-quality` | available-direct |
| 12.5 | G7, G14 | completed | 0.4 | `source-inspection`, `static-contract`, `privacy-data-rights`, `governance-decision` | available-direct |
| 12.6 | G7, G14 | open | 12.2, 12.4, 12.5 | `source-inspection`, `automated-behavior`, `service-integration`, `operational-load-soak`, `privacy-data-rights`, `performance-quality`, `human-outcomes` | planned |
| 12.7 | G7, G14 | open | 4.8, 12.1, 12.2, 12.3, 12.4, 12.5 | `source-inspection`, `static-contract`, `automated-behavior`, `fault-recovery`, `security-adversarial`, `performance-quality` | planned |
| 12.8 | G7, G14 | open | 0.8, 12.2, 12.7 | `source-inspection`, `static-contract`, `automated-behavior`, `fault-recovery`, `performance-quality`, `independent-verification` | planned |
| 12.9 | G7, G14 | open | 0.3, 12.1, 12.2, 17.7 | `source-inspection`, `static-contract`, `performance-quality`, `governance-decision`, `human-outcomes` | planned |
| 12.10 | G7, G14 | open | 12.7, 12.8, 12.9 | `source-inspection`, `static-contract`, `automated-behavior`, `independent-verification` | planned |
| 12.11 | G7, G14 | open | 8.12, 12.1, 12.2 | `source-inspection`, `static-contract`, `performance-quality`, `governance-decision`, `human-outcomes` | planned |
| 13.1 | G13 | completed | 0.3 | `source-inspection`, `static-contract`, `performance-quality`, `governance-decision` | available-direct |
| 13.2 | G13 | completed | 13.1 | `source-inspection`, `automated-behavior`, `service-integration`, `persistence-recovery`, `authorization-isolation`, `protocol-interop`, `security-adversarial`, `privacy-data-rights` | available-direct |
| 13.3 | G13 | completed | 13.1 | `source-inspection`, `static-contract`, `automated-behavior`, `service-integration`, `persistence-recovery`, `fault-recovery` | available-direct |
| 13.4 | G13 | completed | 13.2, 13.3 | `source-inspection`, `service-integration`, `persistence-recovery`, `real-model-provider`, `real-database`, `real-vector-store`, `operational-load-soak`, `fault-recovery`, `performance-quality` | available-direct |
| 13.5 | G13 | open | 13.2, 13.3 | `source-inspection`, `automated-behavior`, `service-integration`, `persistence-recovery`, `authorization-isolation`, `real-model-provider`, `real-database`, `real-vector-store`, `real-browser-ui`, `real-dcc-runtime`, `external-channel-runtime`, `protocol-interop`, `fault-recovery`, `security-adversarial` | planned |
| 13.6 | G13 | completed | 13.1 | `source-inspection`, `automated-behavior`, `persistence-recovery`, `real-database`, `real-vector-store`, `fault-recovery`, `privacy-data-rights` | available-direct |
| 13.7 | G13 | open | 13.1, 13.2, 13.3, 13.4, 13.5, 13.6 | `source-inspection`, `automated-behavior`, `persistence-recovery`, `operational-load-soak`, `fault-recovery`, `performance-quality`, `independent-verification` | planned |
| 14.1 | G16 | completed | 0.4 | `source-inspection`, `static-contract`, `privacy-data-rights` | available-direct |
| 14.2 | G16 | completed | 14.1 | `source-inspection`, `static-contract`, `privacy-data-rights`, `performance-quality`, `governance-decision` | available-direct |
| 14.3 | G16 | completed | 14.1 | `source-inspection`, `automated-behavior`, `service-integration`, `real-multimodal-runtime`, `security-adversarial`, `privacy-data-rights` | available-direct |
| 14.4 | G16 | completed | 3.3, 9.2, 14.1 | `source-inspection`, `automated-behavior`, `service-integration`, `persistence-recovery`, `real-database`, `real-vector-store`, `fault-recovery`, `privacy-data-rights` | available-direct |
| 14.5 | G16 | completed | 14.1 | `source-inspection`, `static-contract`, `automated-behavior`, `service-integration`, `persistence-recovery`, `authorization-isolation`, `security-adversarial`, `privacy-data-rights` | available-direct |
| 14.6 | G16 | completed | 14.1, 14.2 | `source-inspection`, `static-contract`, `privacy-data-rights`, `governance-decision`, `human-outcomes` | available-direct |
| 14.7 | G16 | completed | 14.1, 14.2, 14.3, 14.4, 14.5, 14.6 | `source-inspection`, `static-contract`, `privacy-data-rights`, `governance-decision` | available-direct |
| 14.8 | G16 | completed | 14.1, 14.3, 14.4, 14.5 | `source-inspection`, `static-contract`, `automated-behavior`, `persistence-recovery`, `authorization-isolation`, `privacy-data-rights`, `governance-decision` | available-direct |
| 15.1 | G18 | completed | 0.6 | `source-inspection`, `static-contract`, `service-integration`, `real-model-provider`, `fault-recovery`, `privacy-data-rights`, `performance-quality`, `governance-decision` | available-direct |
| 15.2 | G18 | completed | 15.1 | `source-inspection`, `static-contract`, `automated-behavior`, `service-integration`, `real-model-provider`, `fault-recovery`, `security-adversarial` | available-direct |
| 15.3 | G18 | completed | 15.1 | `source-inspection`, `automated-behavior`, `service-integration`, `persistence-recovery`, `authorization-isolation`, `real-model-provider`, `fault-recovery`, `security-adversarial`, `performance-quality` | available-direct |
| 15.4 | G18 | completed | 15.1 | `source-inspection`, `automated-behavior`, `service-integration`, `real-model-provider`, `privacy-data-rights`, `performance-quality` | available-direct |
| 15.5 | G18 | open | 15.2, 15.3, 15.4 | `source-inspection`, `static-contract`, `automated-behavior`, `service-integration`, `persistence-recovery`, `real-model-provider`, `fault-recovery`, `performance-quality` | planned |
| 15.6 | G18 | completed | 15.1, 15.2, 15.3, 15.4 | `source-inspection`, `automated-behavior`, `service-integration`, `persistence-recovery`, `authorization-isolation`, `fault-recovery`, `security-adversarial`, `performance-quality` | available-direct |
| 15.7 | G18 | open | 15.1, 15.2 | `source-inspection`, `static-contract`, `automated-behavior`, `real-model-provider`, `performance-quality`, `governance-decision` | planned |
| 16.1 | G17 | completed | 0.4 | `source-inspection`, `static-contract`, `governance-decision` | available-direct |
| 16.2 | G17 | completed | 0.4 | `source-inspection`, `static-contract`, `automated-behavior`, `service-integration`, `protocol-interop`, `fault-recovery` | available-direct |
| 16.3 | G17 | completed | 16.2 | `source-inspection`, `static-contract`, `automated-behavior`, `service-integration`, `authorization-isolation`, `protocol-interop`, `fault-recovery`, `security-adversarial` | available-direct |
| 16.4 | G17 | completed | 0.4 | `source-inspection`, `static-contract`, `security-adversarial`, `governance-decision` | available-direct |
| 16.5 | G17 | open | 4.8, 12.7, 13.7, 14.7, 16.2, 16.3, 16.4 | `source-inspection`, `static-contract`, `automated-behavior`, `service-integration`, `authorization-isolation`, `security-adversarial`, `privacy-data-rights` | planned |
| 16.6 | G17 | open | 4.8, 12.7, 13.7, 14.7, 16.1, 16.2, 16.3, 16.4, 16.5 | `source-inspection`, `static-contract`, `automated-behavior`, `authorization-isolation`, `protocol-interop`, `fault-recovery`, `security-adversarial` | planned |
| 17.1 | G15 | open | 0.6 | `source-inspection`, `static-contract`, `privacy-data-rights`, `governance-decision` | planned |
| 17.2 | G15 | open | 17.1 | `source-inspection`, `automated-behavior`, `service-integration`, `real-multimodal-runtime`, `security-adversarial`, `privacy-data-rights`, `independent-verification` | planned |
| 17.3 | G15 | open | 17.1 | `source-inspection`, `automated-behavior`, `real-model-provider`, `real-multimodal-runtime`, `privacy-data-rights`, `performance-quality` | planned |
| 17.4 | G15 | open | 15.1, 17.1 | `source-inspection`, `static-contract`, `service-integration`, `real-model-provider`, `real-dcc-runtime`, `real-multimodal-runtime`, `fault-recovery`, `privacy-data-rights`, `performance-quality`, `independent-verification` | planned |
| 17.5 | G15 | open | 17.2, 17.3, 17.4 | `source-inspection`, `automated-behavior`, `real-model-provider`, `real-multimodal-runtime`, `performance-quality`, `human-outcomes`, `independent-verification` | planned |
| 17.6 | G15 | open | 4.8, 5.8, 12.7, 13.7, 14.7, 17.2, 17.3, 17.4, 17.5 | `source-inspection`, `service-integration`, `real-model-provider`, `real-dcc-runtime`, `real-multimodal-runtime`, `operational-load-soak`, `fault-recovery`, `privacy-data-rights`, `performance-quality`, `independent-verification` | planned |
| 17.7 | G15 | open | 0.8, 17.1 | `source-inspection`, `static-contract`, `governance-decision`, `human-outcomes` | planned |
| 17.8 | G15 | open | 14.7, 17.2, 17.3, 17.7 | `source-inspection`, `automated-behavior`, `service-integration`, `persistence-recovery`, `authorization-isolation`, `real-model-provider`, `real-browser-ui`, `real-multimodal-runtime`, `security-adversarial`, `privacy-data-rights`, `independent-verification` | planned |
| 17.9 | G15 | open | 17.7, 17.8 | `source-inspection`, `static-contract`, `automated-behavior`, `persistence-recovery`, `real-model-provider`, `real-multimodal-runtime`, `human-outcomes` | planned |
| 17.10 | G15 | open | 15.1, 17.4, 17.9 | `source-inspection`, `automated-behavior`, `service-integration`, `real-model-provider`, `real-multimodal-runtime`, `fault-recovery`, `privacy-data-rights`, `performance-quality`, `independent-verification` | planned |
| 17.11 | G15 | open | 5.25, 17.9, 17.10 | `source-inspection`, `automated-behavior`, `service-integration`, `real-dcc-runtime`, `real-multimodal-runtime`, `independent-verification` | planned |
| 17.12 | G15 | open | 5.28, 17.9, 17.11 | `source-inspection`, `automated-behavior`, `service-integration`, `persistence-recovery`, `authorization-isolation`, `real-model-provider`, `real-dcc-runtime`, `fault-recovery`, `human-outcomes`, `independent-verification` | planned |
| 17.13 | G15 | open | 5.24, 5.26, 14.7 | `source-inspection`, `automated-behavior`, `service-integration`, `persistence-recovery`, `authorization-isolation`, `real-dcc-runtime`, `real-engine-runtime`, `fault-recovery`, `privacy-data-rights`, `independent-verification` | planned |
| 17.14 | G15 | open | 17.7, 17.8 | `source-inspection`, `static-contract`, `automated-behavior`, `real-model-provider`, `real-multimodal-runtime`, `performance-quality`, `human-outcomes`, `independent-verification` | planned |
| 17.15 | G15 | open | 5.6, 5.20, 17.14 | `source-inspection`, `automated-behavior`, `persistence-recovery`, `real-dcc-runtime`, `performance-quality`, `human-outcomes`, `independent-verification` | planned |
| 17.16 | G15 | open | 12.9, 17.10, 17.15 | `source-inspection`, `automated-behavior`, `real-dcc-runtime`, `real-multimodal-runtime`, `performance-quality`, `human-outcomes`, `independent-verification` | planned |
| 17.17 | G15 | open | 5.27, 17.15, 17.16 | `source-inspection`, `automated-behavior`, `authorization-isolation`, `real-browser-ui`, `real-engine-runtime`, `fault-recovery`, `performance-quality`, `human-outcomes`, `independent-verification` | planned |
| 17.18 | G15 | open | 5.29, 7.10, 8.11, 12.10, 17.11 | `source-inspection`, `real-model-provider`, `real-browser-ui`, `real-dcc-runtime`, `real-multimodal-runtime`, `performance-quality`, `human-outcomes`, `independent-verification` | planned |
| 17.19 | G15 | open | 5.29, 7.10, 8.11, 12.10, 17.11 | `source-inspection`, `real-model-provider`, `real-browser-ui`, `real-dcc-runtime`, `real-multimodal-runtime`, `performance-quality`, `human-outcomes`, `independent-verification` | planned |
| 17.20 | G15 | open | 8.11, 12.10, 17.13, 17.17 | `source-inspection`, `real-model-provider`, `real-browser-ui`, `real-dcc-runtime`, `real-engine-runtime`, `real-multimodal-runtime`, `performance-quality`, `human-outcomes`, `independent-verification` | planned |
| 17.21 | G15 | open | 17.18, 17.19, 17.20 | `source-inspection`, `static-contract`, `service-integration`, `governance-decision`, `independent-verification`, `docs-render-integrity` | planned |
| 17.22 | G15 | open | 4.8, 14.7, 15.7, 17.4 | `source-inspection`, `automated-behavior`, `service-integration`, `authorization-isolation`, `real-model-provider`, `real-multimodal-runtime`, `fault-recovery`, `security-adversarial`, `privacy-data-rights`, `performance-quality`, `independent-verification` | planned |
| 17.23 | G15 | open | 8.14, 8.15, 8.19, 17.22 | `source-inspection`, `automated-behavior`, `service-integration`, `persistence-recovery`, `real-browser-ui`, `mobile-runtime`, `real-multimodal-runtime`, `fault-recovery`, `human-outcomes` | planned |
| 17.24 | G15 | open | 8.13, 15.7, 17.2, 17.3 | `source-inspection`, `automated-behavior`, `service-integration`, `real-model-provider`, `real-multimodal-runtime`, `security-adversarial`, `privacy-data-rights`, `independent-verification` | planned |
| 17.25 | G15 | open | 4.5, 4.8, 7.1, 8.15 | `source-inspection`, `automated-behavior`, `service-integration`, `authorization-isolation`, `real-agent-runtime`, `fault-recovery`, `security-adversarial`, `performance-quality`, `independent-verification` | planned |
| 17.26 | G15 | open | 4.8, 7.1, 8.15, 8.18, 8.19 | `source-inspection`, `automated-behavior`, `service-integration`, `real-model-provider`, `real-browser-ui`, `security-adversarial`, `privacy-data-rights`, `performance-quality`, `independent-verification` | planned |
| 17.27 | G15 | open | 12.11, 17.23, 17.24, 17.25, 17.26 | `source-inspection`, `static-contract`, `performance-quality`, `human-outcomes`, `independent-verification`, `docs-render-integrity` | planned |
| 18.1 | G1, G2, G3, G4, G5, G6, G7, G8, G9, G10, G11, G12, G13, G14, G15, G16, G17, G18 | open | 0.8, 1.9, 2.8, 3.8, 4.8, 5.11, 6.8, 7.8, 8.9, 9.7, 10.4, 11.8, 12.6, 12.8, 13.7, 14.7, 15.6, 16.6, 17.6, 17.21, 17.27 | `source-inspection`, `static-contract`, `automated-behavior`, `independent-verification` | planned |
| 18.2 | G1, G2, G3, G4, G5, G6, G7, G8, G9, G10, G11, G12, G13, G14, G15, G16, G17, G18 | open | 18.1 | `source-inspection`, `static-contract`, `automated-behavior`, `host-capability-observation` | planned |
| 18.3 | G1, G2, G3, G4, G5, G6, G7, G8, G9, G10, G11, G12, G13, G14, G15, G16, G17, G18 | open | 0.4 | `source-inspection`, `static-contract`, `real-agent-runtime`, `performance-quality`, `governance-decision` | planned |
| 18.4 | G1, G2, G3, G4, G5, G6, G7, G8, G9, G10, G11, G12, G13, G14, G15, G16, G17, G18 | open | 18.1, 18.2, 18.3 | `source-inspection`, `service-integration`, `persistence-recovery`, `real-model-provider`, `real-agent-runtime`, `real-database`, `real-vector-store`, `real-browser-ui`, `native-desktop-runtime`, `mobile-runtime`, `real-dcc-runtime`, `real-engine-runtime`, `real-multimodal-runtime`, `operational-load-soak`, `fault-recovery`, `security-adversarial`, `privacy-data-rights`, `performance-quality`, `human-outcomes`, `independent-verification` | planned |
| 18.5 | G1, G2, G3, G4, G5, G6, G7, G8, G9, G10, G11, G12, G13, G14, G15, G16, G17, G18 | open | 18.1, 18.2 | `source-inspection`, `automated-behavior`, `service-integration`, `persistence-recovery`, `authorization-isolation`, `real-model-provider`, `real-database`, `real-vector-store`, `real-browser-ui`, `native-desktop-runtime`, `real-dcc-runtime`, `external-channel-runtime`, `protocol-interop`, `operational-load-soak`, `fault-recovery`, `security-adversarial`, `privacy-data-rights`, `independent-verification` | planned |
| 18.6 | G1, G2, G3, G4, G5, G6, G7, G8, G9, G10, G11, G12, G13, G14, G15, G16, G17, G18 | open | 18.4, 18.5 | `source-inspection`, `static-contract`, `automated-behavior`, `service-integration`, `persistence-recovery`, `authorization-isolation`, `real-model-provider`, `real-agent-runtime`, `real-database`, `real-vector-store`, `real-browser-ui`, `native-desktop-runtime`, `mobile-runtime`, `real-dcc-runtime`, `real-engine-runtime`, `external-channel-runtime`, `real-multimodal-runtime`, `protocol-interop`, `host-capability-observation`, `operational-load-soak`, `fault-recovery`, `security-adversarial`, `privacy-data-rights`, `performance-quality`, `governance-decision`, `human-outcomes`, `manual-assistive-tech`, `independent-verification`, `docs-render-integrity` | planned |
| 18.7 | G1, G2, G3, G4, G5, G6, G7, G8, G9, G10, G11, G12, G13, G14, G15, G16, G17, G18 | open | 18.1, 18.2, 18.3, 18.4, 18.5, 18.6 | `source-inspection`, `static-contract`, `governance-decision`, `docs-render-integrity` | planned |
| 18.8 | G1, G2, G3, G4, G5, G6, G7, G8, G9, G10, G11, G12, G13, G14, G15, G16, G17, G18 | open | 18.1, 18.2, 18.3, 18.4, 18.5, 18.6, 18.7 | `source-inspection`, `static-contract`, `automated-behavior`, `independent-verification` | planned |
## Exit semantics
Each row exits only when every clause of its exact ledger requirement is
satisfied, its dependencies are complete, admitted evidence covers every named
operating boundary, an applicable negative control is observed red and then
green, and exact limitations remain explicit. A row marked `open` has no
artifact locators and cannot satisfy closure. A `completed` row must link
readable direct evidence.
Phase 18 must reclassify the evidence behind every row as proved, contradicted,
weak/indirect, missing, or externally blocked. Only proved work closes.
Unresolved blockers, including named external or accepted gaps and successor
work, keep the affected task, dependent phase, initiative, and charter
completion open. This matrix deliberately does not pre-classify planned work as
proved.
The charter workflow and runtime contract is
[`EVE_SOTA_CHARTER_COMPLETENESS_2026-09.md`](EVE_SOTA_CHARTER_COMPLETENESS_2026-09.md).
Task 0.8 owns full source coverage, tasks 5.10/5.11 own required runtime
delivery, task 11.8 owns the cross-release benchmark, and task 12.8 owns the
complete evidence gate. Phase 18 depends on those outcomes; registered-gap
coverage alone cannot establish charter completion.
## Explicit prerequisite rationale
Each dependency below is a required artifact or admission gate. Ledger order is
not a dependency. Work may be prepared independently, but completion still
requires every named prerequisite. New task IDs fail closed until reviewed edges
and proof scopes are added.
- **0.2 → 0.1:** Uses the corrected gap definitions rather than the superseded
audit claims.
- **0.3 → 0.2:** Uses the reproducible baseline and its actual capability
boundary.
- **0.4 → 0.2:** Uses the reproducible baseline and its actual capability
boundary.
- **0.5 → 0.3:** Consumes the preregistered outcome thresholds and decision
rules.
- **0.5 → 0.4:** Uses the versioned source/control crosswalk to define
applicable obligations.
- **0.6 → 0.5:** Uses the admitted evidence format and negative-control
requirements.
- **0.7 → 0.2:** Uses the reproducible baseline and its actual capability
boundary.
- **0.7 → 0.6:** Uses the task ownership and direct proof-boundary contract.
- **0.8 → 0.3:** Consumes the preregistered outcome thresholds and decision
rules.
- **0.8 → 0.4:** Uses the versioned source/control crosswalk to define
applicable obligations.
- **0.8 → 0.5:** Uses the admitted evidence format and negative-control
requirements.
- **0.8 → 0.6:** Uses the task ownership and direct proof-boundary contract.
- **0.8 → 0.7:** Requires fresh observed host and runtime availability.
- **1.1 → 0.6:** Uses the task ownership and direct proof-boundary contract.
- **1.2 → 1.1:** Consumes the surveyed authoritative Metis API and stable
surface inventory.
- **1.3 → 1.1:** Consumes the surveyed authoritative Metis API and stable
surface inventory.
- **1.3 → 1.2:** Uses the approved transport, identity and failure semantics.
- **1.4 → 1.3:** Exercises the admitted read views and their actual bindings.
- **1.5 → 1.1:** Consumes the surveyed authoritative Metis API and stable
surface inventory.
- **1.5 → 1.2:** Uses the approved transport, identity and failure semantics.
- **1.5 → 1.4:** Requires the subject, tenant and scope guards around those
views.
- **1.6 → 1.3:** Exercises the admitted read views and their actual bindings.
- **1.6 → 1.4:** Requires the subject, tenant and scope guards around those
views.
- **1.6 → 1.5:** Uses the authoritative mutation or source-backed zero-write
boundary.
- **1.7 → 1.3:** Exercises the admitted read views and their actual bindings.
- **1.7 → 1.4:** Requires the subject, tenant and scope guards around those
views.
- **1.7 → 1.5:** Uses the authoritative mutation or source-backed zero-write
boundary.
- **1.7 → 1.6:** Depends on the restamped tool contract and measured Metis
battery.
- **1.8 → 1.1:** Consumes the surveyed authoritative Metis API and stable
surface inventory.
- **1.8 → 1.7:** Requires the observed Metis refusal, replay and outage
controls.
- **1.9 → 0.8:** Requires ratified source coverage so required workflows cannot
be omitted.
- **1.9 → 1.8:** Extends the existing workbench seam totality mechanism.
- **2.1 → 0.6:** Uses the task ownership and direct proof-boundary contract.
- **2.2 → 2.1:** Implements the adopted attributed delivery lifecycle.
- **2.3 → 2.2:** Consumes canonical readiness, lease, fencing and terminal-state
semantics.
- **2.4 → 2.2:** Consumes canonical readiness, lease, fencing and terminal-state
semantics.
- **2.4 → 2.3:** Exercises the real resumable drain and its process supervision.
- **2.5 → 2.2:** Consumes canonical readiness, lease, fencing and terminal-state
semantics.
- **2.5 → 2.3:** Exercises the real resumable drain and its process supervision.
- **2.5 → 2.4:** Requires evidence-linked retry and truthful triage behavior.
- **2.6 → 2.2:** Consumes canonical readiness, lease, fencing and terminal-state
semantics.
- **2.6 → 2.3:** Exercises the real resumable drain and its process supervision.
- **2.7 → 2.1:** Implements the adopted attributed delivery lifecycle.
- **2.7 → 2.2:** Consumes canonical readiness, lease, fencing and terminal-state
semantics.
- **2.7 → 2.3:** Exercises the real resumable drain and its process supervision.
- **2.7 → 2.4:** Requires evidence-linked retry and truthful triage behavior.
- **2.7 → 2.5:** Requires observable fleet state and operator kill controls.
- **2.7 → 2.6:** Requires isolated execution, preserved user work and
independent identities.
- **2.8 → 2.7:** Consumes the fault-tested queue and hand-lease/drain parity
contract.
- **2.8 → 4.8:** Live admission requires the security red-team gate; preparation
may proceed before it.
- **2.8 → 12.7:** Live admission requires the applicable-family evaluation gate,
not the future charter benchmark.
- **2.8 → 13.7:** Live admission requires operational detect, kill, recover and
verify readiness.
- **3.1 → 0.6:** Uses the task ownership and direct proof-boundary contract.
- **3.2 → 0.6:** Uses the task ownership and direct proof-boundary contract.
- **3.2 → 15.1:** Consumes priced, capability-correct, fail-loud model registry
bindings.
- **3.3 → 3.2:** Consumes the measured approved embedding route and provider
boundary.
- **3.4 → 3.3:** Uses the versioned ACL-bearing dense index and deletion
metadata.
- **3.5 → 3.1:** Requires independently human-labelled relevance cases for
quality claims.
- **3.5 → 3.4:** Exercises actual hybrid fusion, fallback and ranking behavior.
- **3.6 → 3.3:** Uses the versioned ACL-bearing dense index and deletion
metadata.
- **3.6 → 3.4:** Exercises actual hybrid fusion, fallback and ranking behavior.
- **3.7 → 3.5:** Requires paired independent retrieval quality and cost
measurements.
- **3.7 → 3.6:** Requires adversarial corpus, tenancy and poisoning evidence.
- **3.8 → 3.7:** Consumes the measured retrieval promotion or lexical-only
decision.
- **3.8 → 13.6:** Requires actual backup/restore, migration and
deletion-preserving recovery proof.
- **3.8 → 14.4:** Requires deletion/export propagation through actual data
stores and backups.
- **3.8 → 15.5:** Requires automated route canary, drift detection, quarantine
and rollback.
- **4.1 → 0.4:** Uses the versioned source/control crosswalk to define
applicable obligations.
- **4.2 → 4.1:** Consumes the threat model and concrete attack boundaries.
- **4.3 → 4.1:** Consumes the threat model and concrete attack boundaries.
- **4.3 → 4.2:** Requires trust labels and provenance to survive prompt
assembly.
- **4.4 → 4.1:** Consumes the threat model and concrete attack boundaries.
- **4.5 → 4.1:** Consumes the threat model and concrete attack boundaries.
- **4.5 → 4.4:** Uses deterministic effective-authority enforcement.
- **4.6 → 4.1:** Consumes the threat model and concrete attack boundaries.
- **4.6 → 4.4:** Uses deterministic effective-authority enforcement.
- **4.7 → 4.4:** Uses deterministic effective-authority enforcement.
- **4.7 → 4.5:** Requires execution isolation and scoped credentials.
- **4.8 → 4.1:** Consumes the threat model and concrete attack boundaries.
- **4.8 → 4.2:** Requires trust labels and provenance to survive prompt
assembly.
- **4.8 → 4.3:** Requires paired attack and benign-utility cases.
- **4.8 → 4.4:** Uses deterministic effective-authority enforcement.
- **4.8 → 4.5:** Requires execution isolation and scoped credentials.
- **4.8 → 4.6:** Consumes external-tool trust and credential boundary controls.
- **4.8 → 4.7:** Requires confirmation, cancellation and reversal at high-impact
boundaries.
- **5.1 → 0.6:** Uses the task ownership and direct proof-boundary contract.
- **5.2 → 5.1:** Consumes the audited real Bellona delegation path.
- **5.2 → 15.1:** Consumes priced, capability-correct, fail-loud model registry
bindings.
- **5.3 → 0.7:** Requires fresh observed host and runtime availability.
- **5.4 → 5.1:** Consumes the audited real Bellona delegation path.
- **5.4 → 5.3:** Requires a real Blender handshake and save/readback smoke.
- **5.5 → 4.8:** Live admission requires the security red-team gate; preparation
may proceed before it.
- **5.5 → 5.2:** Requires capability-correct registry model bindings.
- **5.5 → 5.4:** Uses the deterministic risk-classified Blender action catalog.
- **5.5 → 12.7:** Live admission requires the applicable-family evaluation gate,
not the future charter benchmark.
- **5.5 → 13.7:** Live admission requires operational detect, kill, recover and
verify readiness.
- **5.5 → 14.7:** Creative/live admission requires actual privacy, media rights
and consent governance.
- **5.6 → 5.5:** Requires governed leased Bellona admission.
- **5.7 → 5.6:** Extends the independently verified real Blender first-light
artifact.
- **5.8 → 5.5:** Requires governed leased Bellona admission.
- **5.8 → 5.7:** Requires the Blender breadth and reopen behavior under test.
- **5.9 → 0.7:** Requires fresh observed host and runtime availability.
- **5.10 → 0.8:** Requires ratified source coverage so required workflows cannot
be omitted.
- **5.10 → 4.8:** Live admission requires the security red-team gate;
preparation may proceed before it.
- **5.10 → 5.12:** Consumes source-owned runtime packages with exact adapter,
owner and acceptance boundaries.
- **5.10 → 5.13:** Uses the real governed Unreal adapter and editor state.
- **5.10 → 5.14:** Requires packaged/playable Unreal workflows, recovery and
independent quality evidence.
- **5.10 → 5.33:** Requires complete tool-schema delivery proven on the admitted
Unreal profile, which settles the optional native backend chain (5.30 to 5.32)
by adoption or measured rejection.
- **5.10 → 12.7:** Live admission requires the applicable-family evaluation
gate, not the future charter benchmark.
- **5.10 → 13.7:** Live admission requires operational detect, kill, recover and
verify readiness.
- **5.10 → 14.7:** Creative/live admission requires actual privacy, media rights
and consent governance.
- **5.11 → 0.8:** Requires ratified source coverage so required workflows cannot
be omitted.
- **5.11 → 5.8:** Requires observed Blender failure and false-success controls.
- **5.11 → 5.10:** Requires complete Unreal delivery evidence for required
engine workflows.
- **5.11 → 5.12:** Consumes source-owned runtime packages with exact adapter,
owner and acceptance boundaries.
- **5.11 → 5.15:** Requires internal Maya/UGC sandbox delivery and
hostile-content containment.
- **5.11 → 5.16:** Requires separate real solver and numerical-engine acceptance
receipts.
- **5.11 → 5.17:** Requires real voice/media and companion-simulation package
evidence.
- **5.11 → 5.18:** Requires actual declared platform/device packaging and
user-journey proof.
- **5.11 → 5.19:** Requires all additionally selected creative-application child
tasks to close.
- **5.12 → 0.8:** Requires ratified source coverage so required workflows cannot
be omitted.
- **5.12 → 5.9:** Uses the fresh host/runtime/licence matrix.
- **5.13 → 4.8:** Live admission requires the security red-team gate;
preparation may proceed before it.
- **5.13 → 5.12:** Consumes source-owned runtime packages with exact adapter,
owner and acceptance boundaries.
- **5.13 → 12.7:** Live admission requires the applicable-family evaluation
gate, not the future charter benchmark.
- **5.13 → 13.7:** Live admission requires operational detect, kill, recover and
verify readiness.
- **5.13 → 14.7:** Creative/live admission requires actual privacy, media rights
and consent governance.
- **5.13 → 15.1:** Consumes priced, capability-correct, fail-loud model registry
bindings.
- **5.14 → 5.13:** Uses the real governed Unreal adapter and editor state.
- **5.14 → 5.34:** Retains the target-resolution failure and persisted-state
receipts inside this battery for each admitted backend.
- **5.14 → 5.35:** Records the project authoring-convention results inside this
battery rather than in another release gate.
- **5.15 → 4.8:** Live admission requires the security red-team gate;
preparation may proceed before it.
- **5.15 → 5.12:** Consumes source-owned runtime packages with exact adapter,
owner and acceptance boundaries.
- **5.15 → 12.7:** Live admission requires the applicable-family evaluation
gate, not the future charter benchmark.
- **5.15 → 13.7:** Live admission requires operational detect, kill, recover and
verify readiness.
- **5.15 → 14.7:** Creative/live admission requires actual privacy, media rights
and consent governance.
- **5.16 → 4.8:** Live admission requires the security red-team gate;
preparation may proceed before it.
- **5.16 → 5.12:** Consumes source-owned runtime packages with exact adapter,
owner and acceptance boundaries.
- **5.16 → 12.7:** Live admission requires the applicable-family evaluation
gate, not the future charter benchmark.
- **5.16 → 13.7:** Live admission requires operational detect, kill, recover and
verify readiness.
- **5.16 → 14.7:** Creative/live admission requires actual privacy, media rights
and consent governance.
- **5.17 → 4.8:** Live admission requires the security red-team gate;
preparation may proceed before it.
- **5.17 → 5.12:** Consumes source-owned runtime packages with exact adapter,
owner and acceptance boundaries.
- **5.17 → 12.7:** Live admission requires the applicable-family evaluation
gate, not the future charter benchmark.
- **5.17 → 13.7:** Live admission requires operational detect, kill, recover and
verify readiness.
- **5.17 → 14.7:** Creative/live admission requires actual privacy, media rights
and consent governance.
- **5.17 → 17.4:** Uses the owning domain real generation toolchain and editable
artifacts.
- **5.17 → 17.5:** Requires separate modality quality, human and failure
evidence.
- **5.18 → 4.8:** Live admission requires the security red-team gate;
preparation may proceed before it.
- **5.18 → 5.12:** Consumes source-owned runtime packages with exact adapter,
owner and acceptance boundaries.
- **5.18 → 12.7:** Live admission requires the applicable-family evaluation
gate, not the future charter benchmark.
- **5.18 → 13.7:** Live admission requires operational detect, kill, recover and
verify readiness.
- **5.18 → 14.7:** Creative/live admission requires actual privacy, media rights
and consent governance.
- **5.19 → 5.12:** Consumes source-owned runtime packages with exact adapter,
owner and acceptance boundaries.
- **5.19 → 5.29:** Requires the selected After Effects application delivery and
breadth evidence.
- **5.20 → 5.3:** Corrects the observed compressed-save regression in the real
executable smoke.
- **5.21 → 5.9:** Refreshes current host and runtime availability rather than
borrowing another host receipt.
- **5.21 → 5.12:** Extends the source-owned runtime packages with usable
creative application hosts.
- **5.22 → 4.5:** Requires execution isolation before application mutation.
- **5.22 → 4.6:** Uses authenticated external-tool and transport trust.
- **5.22 → 4.7:** Requires deterministic preview, confirmation, and reversal
controls.
- **5.22 → 5.21:** Uses the provisioned and version-pinned After Effects host.
- **5.23 → 5.22:** Builds motion authoring on a verified live AE backend.
- **5.24 → 5.23:** Uses actual editable AE composition and keyframe operations.
- **5.24 → 14.7:** Requires source-media rights and artifact governance.
- **5.25 → 5.5:** Reuses attributed leased Bellona admission and its cross-phase
gates.
- **5.25 → 5.24:** Requires real AE save, reopen, and render operations.
- **5.25 → 12.10:** Uses registered creative-family evidence and rejection
rules.
- **5.25 → 14.7:** Requires actual media and project governance for live
execution.
- **5.26 → 5.5:** Requires governed leased Bellona execution.
- **5.26 → 5.13:** Uses the admitted real Unreal editor adapter and
authenticated transport.
- **5.26 → 14.7:** Requires cross-domain source and artifact rights governance.
- **5.26 → 17.4:** Uses live generation through the owning domain.
- **5.27 → 5.14:** Uses verified Unreal workflow and recovery behavior.
- **5.27 → 5.20:** Uses corrected compressed and uncompressed Blender
readback/export verification.
- **5.27 → 5.26:** Requires actual cross-app provider bindings and transfer.
- **5.28 → 5.24:** Uses the real native project and render lifecycle.
- **5.28 → 5.26:** Exercises failure across actual cross-application boundaries.
- **5.28 → 13.3:** Reuses canonical retry, deadline, idempotency, and
cancellation semantics.
- **5.29 → 5.25:** Exercises real attributed Eve-to-AE execution.
- **5.29 → 5.28:** Requires durable application failure and recovery behavior.
- **5.29 → 12.9:** Uses preregistered creative measurement and failure criteria.
- **5.30 → 5.13:** Compares a required workflow against the current governed
Bellona backend.
- **5.30 → 5.21:** Extends the provisioned, version-pinned host records;
provisioning stays with 5.21.
- **5.31 → 5.5:** Production admission stays behind the attributed leased
Bellona registration.
- **5.31 → 5.13:** Reuses the governed adapter authority contract and its
live-admission gates.
- **5.31 → 5.30:** Exposes the native transport only for profiles the
qualification admitted.
- **5.32 → 5.31:** Authorizes operations dispatched over the serialized native
transport.
- **5.33 → 5.32:** Native integration and rollout use the authorized discovery
profile; generic fixture work may start before it.
- **5.33 → 15.3:** Extends the long-context instruction and schema loss handling
rather than duplicating it.
- **5.34 → 5.13:** Extends the governed adapter actor, asset and planning
contracts with target identity.
- **5.35 → 5.13:** Carries the project policy identity through the governed
adapter plans and result evidence.
- **6.1 → 0.6:** Uses the task ownership and direct proof-boundary contract.
- **6.2 → 6.1:** Uses the native-versus-browser ownership decision.
- **6.3 → 6.1:** Uses the native-versus-browser ownership decision.
- **6.3 → 6.2:** Requires the shared planning/vision registry and governed
screenshot-input contract.
- **6.4 → 0.7:** Requires fresh observed host and runtime availability.
- **6.4 → 6.1:** Uses the native-versus-browser ownership decision.
- **6.5 → 6.3:** Requires the actual native execution chain to pass
fabrication/security review.
- **6.5 → 6.4:** Uses a real OS-level desktop fixture and permission preflight.
- **6.6 → 4.8:** Live admission requires the security red-team gate; preparation
may proceed before it.
- **6.6 → 6.2:** Requires native planning/vision registry bindings.
- **6.6 → 6.3:** Requires the actual native execution chain to pass
fabrication/security review.
- **6.6 → 6.4:** Uses a real OS-level desktop fixture and permission preflight.
- **6.6 → 6.5:** Requires per-run authority, focus, budget and interrupt
enforcement.
- **6.6 → 12.7:** Live admission requires the applicable-family evaluation gate,
not the future charter benchmark.
- **6.6 → 13.7:** Live admission requires operational detect, kill, recover and
verify readiness.
- **6.6 → 14.7:** Creative/live admission requires actual privacy, media rights
and consent governance.
- **6.7 → 6.6:** Requires governed native leased-work admission.
- **6.8 → 6.7:** Requires measured native task breadth and recovery behavior.
- **7.1 → 0.6:** Uses the task ownership and direct proof-boundary contract.
- **7.2 → 7.1:** Uses the corrected existing Hermes/fusion boundary.
- **7.2 → 12.2:** Consumes independent cohort, grader and statistically feasible
sample contracts.
- **7.3 → 7.1:** Uses the corrected existing Hermes/fusion boundary.
- **7.4 → 4.8:** Live admission requires the security red-team gate; preparation
may proceed before it.
- **7.4 → 7.2:** Consumes independently graded watcher, channel and
assistant-runtime cases.
- **7.4 → 7.3:** Implements the specified watcher event, freshness and lifecycle
semantics.
- **7.4 → 12.7:** Live admission requires the applicable-family evaluation gate,
not the future charter benchmark.
- **7.4 → 13.7:** Live admission requires operational detect, kill, recover and
verify readiness.
- **7.4 → 14.7:** Creative/live admission requires actual privacy, media rights
and consent governance.
- **7.5 → 7.4:** Uses admitted confirmed watcher tools and scheduler state.
- **7.6 → 7.1:** Uses the corrected existing Hermes/fusion boundary.
- **7.7 → 3.7:** Consumes the measured retrieval promotion or lexical-only
decision.
- **7.7 → 7.6:** Uses one governed skill ownership, retrieval and poisoning
policy.
- **7.7 → 9.4:** Uses provenance, supersession, expiry and reauthorization
policy.
- **7.8 → 4.8:** Live admission requires the security red-team gate; preparation
may proceed before it.
- **7.8 → 7.2:** Consumes independently graded watcher, channel and
assistant-runtime cases.
- **7.8 → 7.5:** Requires live watcher breadth, dedupe and recovery evidence.
- **7.8 → 7.6:** Uses one governed skill ownership, retrieval and poisoning
policy.
- **7.8 → 7.7:** Uses the shared semantic recall path and deletion/tenant
boundaries.
- **7.8 → 12.7:** Live admission requires the applicable-family evaluation gate,
not the future charter benchmark.
- **7.8 → 13.7:** Live admission requires operational detect, kill, recover and
verify readiness.
- **7.8 → 14.7:** Creative/live admission requires actual privacy, media rights
and consent governance.
- **7.9 → 7.6:** Reuses the ratified skill ownership and lifecycle.
- **7.9 → 14.7:** Requires retained examples and source assets to carry rights
governance.
- **7.9 → 17.7:** Uses explicit creative workflow and capability requirements.
- **7.10 → 7.9:** Uses versioned and authority-clamped creative recipes.
- **7.10 → 12.9:** Uses held-out briefs and preregistered human-grounded quality
rules.
- **8.1 → 0.6:** Uses the task ownership and direct proof-boundary contract.
- **8.2 → 8.1:** Uses the actual shipped surface and invocation inventory.
- **8.3 → 8.1:** Uses the actual shipped surface and invocation inventory.
- **8.3 → 8.2:** Consumes the approved restrained interaction and design brief.
- **8.4 → 8.1:** Uses the actual shipped surface and invocation inventory.
- **8.4 → 8.2:** Consumes the approved restrained interaction and design brief.
- **8.5 → 8.2:** Consumes the approved restrained interaction and design brief.
- **8.6 → 8.2:** Consumes the approved restrained interaction and design brief.
- **8.7 → 8.1:** Uses the actual shipped surface and invocation inventory.
- **8.7 → 8.2:** Consumes the approved restrained interaction and design brief.
- **8.8 → 8.3:** Exercises contextual invocation totality and sanitized context.
- **8.8 → 8.4:** Requires the measured mobile streaming and lifecycle contract.
- **8.8 → 8.5:** Exercises long-turn steering, cancel and recovery behavior.
- **8.8 → 8.6:** Requires the evaluated typed UI or deterministic fallback
decision.
- **8.8 → 8.7:** Requires the measured voice depth and accessible fallback
boundary.
- **8.9 → 8.1:** Uses the actual shipped surface and invocation inventory.
- **8.9 → 8.2:** Consumes the approved restrained interaction and design brief.
- **8.9 → 8.3:** Exercises contextual invocation totality and sanitized context.
- **8.9 → 8.4:** Requires the measured mobile streaming and lifecycle contract.
- **8.9 → 8.5:** Exercises long-turn steering, cancel and recovery behavior.
- **8.9 → 8.6:** Requires the evaluated typed UI or deterministic fallback
decision.
- **8.9 → 8.7:** Requires the measured voice depth and accessible fallback
boundary.
- **8.9 → 8.8:** Requires assistive-tech and WCAG acceptance for the affected
states.
- **8.9 → 8.20:** Requires the proven chat-surface parity journeys and register.
- **8.10 → 8.2:** Uses the required interface and interaction design brief.
- **8.10 → 8.5:** Reuses user steering and long-turn controls.
- **8.10 → 17.9:** Presents actual versioned storyboard alternatives and
selection.
- **8.10 → 17.12:** Binds annotations to a working native revision loop.
- **8.11 → 8.10:** Exercises the real creative review controls and service
state.
- **8.11 → 17.13:** Requires actual portable, downloadable native project
packages.
- **8.12 → 8.1:** Uses the actual shipped surface and invocation inventory.
- **8.12 → 8.2:** Consumes the approved restrained interaction and design brief.
- **8.13 → 8.2:** Consumes the approved restrained interaction and design brief.
- **8.13 → 8.3:** Reuses the shared privacy boundary and untrusted
context-handoff block.
- **8.13 → 8.12:** Consumes the ratified per-surface parity register.
- **8.13 → 14.8:** Requires attachment classification, scanning, retention and
deletion policy.
- **8.13 → 17.1:** Uses the real modality/toolchain capability and limitation
matrix.
- **8.14 → 8.2:** Consumes the approved restrained interaction and design brief.
- **8.14 → 8.12:** Consumes the ratified per-surface parity register.
- **8.15 → 8.5:** Exercises long-turn steering, cancel and recovery behavior.
- **8.15 → 8.6:** Extends the adopted typed generative-UI allowlist and
validators.
- **8.15 → 8.14:** Uses the rich renderer for artifact previews and media.
- **8.16 → 8.5:** Exercises long-turn steering, cancel and recovery behavior.
- **8.16 → 8.12:** Consumes the ratified per-surface parity register.
- **8.17 → 8.12:** Consumes the ratified per-surface parity register.
- **8.17 → 9.2:** Uses the operator-visible memory inspect/correct/forget
controls.
- **8.17 → 14.8:** Requires deletion propagation and permalink access policy.
- **8.18 → 8.12:** Consumes the ratified per-surface parity register.
- **8.18 → 15.7:** Uses the admitted generation legs and priced route choices.
- **8.19 → 8.5:** Exercises long-turn steering, cancel and recovery behavior.
- **8.19 → 8.15:** Delivers results into the artifact workspace.
- **8.19 → 13.3:** Uses standardized deadline, retry, idempotency and
cancellation semantics.
- **8.20 → 8.8:** Requires assistive-tech and WCAG acceptance for the affected
states.
- **8.20 → 8.13:** Requires the shipped multimodal composer.
- **8.20 → 8.14:** Requires rich rendering on every surface.
- **8.20 → 8.15:** Requires the delivered, versioned artifact workspace on every
surface.
- **8.20 → 8.16:** Requires turn-level edit, regenerate, copy, feedback and
share parity.
- **8.20 → 8.17:** Requires conversation list, rename, archive, delete and
project parity.
- **8.20 → 8.18:** Requires composer mode, tool and route selection.
- **8.20 → 8.19:** Requires long-running work progress and completion delivery
in chat.
- **9.1 → 0.6:** Uses the task ownership and direct proof-boundary contract.
- **9.2 → 9.1:** Uses the actual persisted-memory and session-only posture.
- **9.3 → 9.1:** Uses the actual persisted-memory and session-only posture.
- **9.4 → 9.1:** Uses the actual persisted-memory and session-only posture.
- **9.4 → 9.2:** Requires operator data inspection, correction and deletion
controls.
- **9.5 → 9.2:** Requires operator data inspection, correction and deletion
controls.
- **9.5 → 9.4:** Uses provenance, supersession, expiry and reauthorization
policy.
- **9.6 → 3.7:** Consumes the measured retrieval promotion or lexical-only
decision.
- **9.6 → 7.7:** Uses the shared semantic recall path and deletion/tenant
boundaries.
- **9.6 → 9.4:** Uses provenance, supersession, expiry and reauthorization
policy.
- **9.7 → 9.2:** Requires operator data inspection, correction and deletion
controls.
- **9.7 → 9.3:** Requires independently human-labelled multi-session memory
cases.
- **9.7 → 9.4:** Uses provenance, supersession, expiry and reauthorization
policy.
- **9.7 → 9.5:** Requires observed poisoning, isolation and
deletion-resurrection controls.
- **9.7 → 9.6:** Requires the admitted semantic-memory path or evidenced
non-adoption boundary.
- **10.1 → 0.6:** Uses the task ownership and direct proof-boundary contract.
- **10.2 → 10.1:** Uses the authoritative existing release-blocked room
expectations.
- **10.3 → 10.2:** Exercises runtime release-scope expectation selection.
- **10.4 → 10.1:** Uses the authoritative existing release-blocked room
expectations.
- **10.4 → 10.2:** Exercises runtime release-scope expectation selection.
- **10.4 → 10.3:** Requires provider-free dual-scope and inverted-mapping
controls.
- **11.1 → 0.6:** Uses the task ownership and direct proof-boundary contract.
- **11.2 → 2.1:** Implements the adopted attributed delivery lifecycle.
- **11.3 → 0.6:** Uses the task ownership and direct proof-boundary contract.
- **11.4 → 4.8:** Live admission requires the security red-team gate;
preparation may proceed before it.
- **11.4 → 11.2:** Implements or grades the explicit plan-quality and scope
contract.
- **11.4 → 11.3:** Uses the repository execution, isolation and evidence
contract.
- **11.5 → 11.3:** Uses the repository execution, isolation and evidence
contract.
- **11.5 → 11.11:** Exercises revision-bound fleet execution, invalidation and
recovery.
- **11.6 → 11.1:** Uses unseen versioned engineering cases with hidden
acceptance evidence.
- **11.6 → 11.4:** Requires independent behavioral review and anti-fabrication
evidence.
- **11.6 → 11.5:** Requires real delivery conflict, crash, rollback and resume
behavior.
- **11.6 → 11.12:** Requires complete operator labor accounting, including
approvals and failed work.
- **11.6 → 12.2:** Consumes independent cohort, grader and statistically
feasible sample contracts.
- **11.7 → 2.8:** Requires the actual governed fleet drain and soak evidence.
- **11.7 → 11.6:** Requires measured delivery competence against paired
baselines.
- **11.7 → 11.11:** Exercises revision-bound fleet execution, invalidation and
recovery.
- **11.8 → 0.8:** Requires ratified source coverage so required workflows cannot
be omitted.
- **11.8 → 5.11:** Requires every selected runtime package, including newly
discovered children, to be delivered.
- **11.8 → 11.7:** Requires an actual operator-goal-to-verified-delivery
trajectory.
- **11.8 → 12.8:** Requires the full evidence-joining charter gate
implementation, independently of later production results.
- **11.8 → 17.6:** Requires real governed cross-modal delivery and verification.
- **11.9 → 11.2:** Implements or grades the explicit plan-quality and scope
contract.
- **11.10 → 11.9:** Consumes persisted attributed goal and requirement
revisions.
- **11.11 → 2.7:** Consumes the fault-tested queue and hand-lease/drain parity
contract.
- **11.11 → 11.10:** Consumes the persisted acyclic dependency graph and proof
plan.
- **11.12 → 11.2:** Implements or grades the explicit plan-quality and scope
contract.
- **11.12 → 13.2:** Consumes attributed end-to-end trace/correlation records.
- **12.1 → 0.8:** Requires ratified source coverage so required workflows cannot
be omitted.
- **12.2 → 0.3:** Consumes the preregistered outcome thresholds and decision
rules.
- **12.2 → 12.1:** Uses the complete required-capability and risk taxonomy.
- **12.3 → 0.3:** Consumes the preregistered outcome thresholds and decision
rules.
- **12.4 → 12.2:** Consumes independent cohort, grader and statistically
feasible sample contracts.
- **12.5 → 0.4:** Uses the versioned source/control crosswalk to define
applicable obligations.
- **12.6 → 12.2:** Consumes independent cohort, grader and statistically
feasible sample contracts.
- **12.6 → 12.4:** Requires trajectory-level grading of state, efficiency and
truthfulness.
- **12.6 → 12.5:** Uses held-out-data governance, provenance and contamination
controls.
- **12.7 → 4.8:** Live admission requires the security red-team gate;
preparation may proceed before it.
- **12.7 → 12.1:** Uses the complete required-capability and risk taxonomy.
- **12.7 → 12.2:** Consumes independent cohort, grader and statistically
feasible sample contracts.
- **12.7 → 12.3:** Requires real blinded human labels and validated judge
classes.
- **12.7 → 12.4:** Requires trajectory-level grading of state, efficiency and
truthfulness.
- **12.7 → 12.5:** Uses held-out-data governance, provenance and contamination
controls.
- **12.8 → 0.8:** Requires ratified source coverage so required workflows cannot
be omitted.
- **12.8 → 12.2:** Consumes independent cohort, grader and statistically
feasible sample contracts.
- **12.8 → 12.7:** Live admission requires the applicable-family evaluation
gate, not the future charter benchmark.
- **12.9 → 0.3:** Uses the north-star outcome and decision contract.
- **12.9 → 12.1:** Extends the existing task and evaluation family taxonomy.
- **12.9 → 12.2:** Reuses independent case sampling and stochastic decision
rules.
- **12.9 → 17.7:** Measures the required creative workflow outcomes.
- **12.10 → 12.7:** Extends the existing applicable-family release gate.
- **12.10 → 12.8:** Extends the charter completeness evidence join.
- **12.10 → 12.9:** Uses explicit creative family thresholds and evidence
boundaries.
- **12.11 → 8.12:** Consumes the ratified per-surface parity register.
- **12.11 → 12.1:** Uses the expanded task/eval taxonomy covering UX, multimodal
and cost.
- **12.11 → 12.2:** Uses the case diversity and stochastic-run rules.
- **13.1 → 0.3:** Consumes the preregistered outcome thresholds and decision
rules.
- **13.2 → 13.1:** Uses the fixed per-plane SLO, RTO, budget and alert contract.
- **13.3 → 13.1:** Uses the fixed per-plane SLO, RTO, budget and alert contract.
- **13.4 → 13.2:** Consumes attributed end-to-end trace/correlation records.
- **13.4 → 13.3:** Exercises bounded retry, fencing, backpressure and
cancellation semantics.
- **13.5 → 13.2:** Consumes attributed end-to-end trace/correlation records.
- **13.5 → 13.3:** Exercises bounded retry, fencing, backpressure and
cancellation semantics.
- **13.6 → 13.1:** Uses the fixed per-plane SLO, RTO, budget and alert contract.
- **13.7 → 13.1:** Uses the fixed per-plane SLO, RTO, budget and alert contract.
- **13.7 → 13.2:** Consumes attributed end-to-end trace/correlation records.
- **13.7 → 13.3:** Exercises bounded retry, fencing, backpressure and
cancellation semantics.
- **13.7 → 13.4:** Requires resource-safe realistic load and soak evidence.
- **13.7 → 13.5:** Requires observed dependency/crash/resource fault recovery.
- **13.7 → 13.6:** Requires actual backup/restore, migration and
deletion-preserving recovery proof.
- **14.1 → 0.4:** Uses the versioned source/control crosswalk to define
applicable obligations.
- **14.2 → 14.1:** Uses the complete source-to-destination data inventory.
- **14.3 → 14.1:** Uses the complete source-to-destination data inventory.
- **14.4 → 3.3:** Uses the versioned ACL-bearing dense index and deletion
metadata.
- **14.4 → 9.2:** Requires operator data inspection, correction and deletion
controls.
- **14.4 → 14.1:** Uses the complete source-to-destination data inventory.
- **14.5 → 14.1:** Uses the complete source-to-destination data inventory.
- **14.6 → 14.1:** Uses the complete source-to-destination data inventory.
- **14.6 → 14.2:** Requires the reviewed provider transfer, retention and
subprocess posture.
- **14.7 → 14.1:** Uses the complete source-to-destination data inventory.
- **14.7 → 14.2:** Requires the reviewed provider transfer, retention and
subprocess posture.
- **14.7 → 14.3:** Requires minimization and sensitive-data/exfiltration
controls.
- **14.7 → 14.4:** Requires deletion/export propagation through actual data
stores and backups.
- **14.7 → 14.5:** Requires retention, disclosure and audited access policy.
- **14.7 → 14.6:** Requires applicable human-owned legal and oversight
decisions.
- **14.8 → 14.1:** Uses the complete source-to-destination data inventory.
- **14.8 → 14.3:** Requires minimization and sensitive-data/exfiltration
controls.
- **14.8 → 14.4:** Requires deletion/export propagation through actual data
stores and backups.
- **14.8 → 14.5:** Requires retention, disclosure and audited access policy.
- **15.1 → 0.6:** Uses the task ownership and direct proof-boundary contract.
- **15.2 → 15.1:** Consumes priced, capability-correct, fail-loud model registry
bindings.
- **15.3 → 15.1:** Consumes priced, capability-correct, fail-loud model registry
bindings.
- **15.4 → 15.1:** Consumes priced, capability-correct, fail-loud model registry
bindings.
- **15.5 → 15.2:** Requires measured fallback and endpoint-failover contract
preservation.
- **15.5 → 15.3:** Requires long-context authority, provenance and
state-retention evidence.
- **15.5 → 15.4:** Uses billed cost and latency per independently verified
outcome.
- **15.6 → 15.1:** Consumes priced, capability-correct, fail-loud model registry
bindings.
- **15.6 → 15.2:** Requires measured fallback and endpoint-failover contract
preservation.
- **15.6 → 15.3:** Requires long-context authority, provenance and
state-retention evidence.
- **15.6 → 15.4:** Uses billed cost and latency per independently verified
outcome.
- **15.7 → 15.1:** Consumes priced, capability-correct, fail-loud model registry
bindings.
- **15.7 → 15.2:** Uses the provider/route resilience policy.
- **16.1 → 0.4:** Uses the versioned source/control crosswalk to define
applicable obligations.
- **16.2 → 0.4:** Uses the versioned source/control crosswalk to define
applicable obligations.
- **16.3 → 16.2:** Consumes the pinned MCP lifecycle and transport conformance
contract.
- **16.4 → 0.4:** Uses the versioned source/control crosswalk to define
applicable obligations.
- **16.5 → 4.8:** Live admission requires the security red-team gate;
preparation may proceed before it.
- **16.5 → 12.7:** Live admission requires the applicable-family evaluation
gate, not the future charter benchmark.
- **16.5 → 13.7:** Live admission requires operational detect, kill, recover and
verify readiness.
- **16.5 → 14.7:** Creative/live admission requires actual privacy, media rights
and consent governance.
- **16.5 → 16.2:** Consumes the pinned MCP lifecycle and transport conformance
contract.
- **16.5 → 16.3:** Requires protected HTTP MCP authorization and token-boundary
proof.
- **16.5 → 16.4:** Uses the evidenced A2A adoption or source-backed non-adoption
decision.
- **16.6 → 4.8:** Live admission requires the security red-team gate;
preparation may proceed before it.
- **16.6 → 12.7:** Live admission requires the applicable-family evaluation
gate, not the future charter benchmark.
- **16.6 → 13.7:** Live admission requires operational detect, kill, recover and
verify readiness.
- **16.6 → 14.7:** Creative/live admission requires actual privacy, media rights
and consent governance.
- **16.6 → 16.1:** Uses the current event interoperability decision.
- **16.6 → 16.2:** Consumes the pinned MCP lifecycle and transport conformance
contract.
- **16.6 → 16.3:** Requires protected HTTP MCP authorization and token-boundary
proof.
- **16.6 → 16.4:** Uses the evidenced A2A adoption or source-backed non-adoption
decision.
- **16.6 → 16.5:** Requires admitted external integration provenance, health and
quarantine controls.
- **17.1 → 0.6:** Uses the task ownership and direct proof-boundary contract.
- **17.2 → 17.1:** Uses the real modality/toolchain capability and limitation
matrix.
- **17.3 → 17.1:** Uses the real modality/toolchain capability and limitation
matrix.
- **17.4 → 15.1:** Consumes priced, capability-correct, fail-loud model registry
bindings.
- **17.4 → 17.1:** Uses the real modality/toolchain capability and limitation
matrix.
- **17.5 → 17.2:** Exercises actual structured document extraction and source
anchors.
- **17.5 → 17.3:** Exercises grounded multimodal understanding and uncertainty
handling.
- **17.5 → 17.4:** Uses the owning domain real generation toolchain and editable
artifacts.
- **17.6 → 4.8:** Live admission requires the security red-team gate;
preparation may proceed before it.
- **17.6 → 5.8:** Requires observed Blender failure and false-success controls.
- **17.6 → 12.7:** Live admission requires the applicable-family evaluation
gate, not the future charter benchmark.
- **17.6 → 13.7:** Live admission requires operational detect, kill, recover and
verify readiness.
- **17.6 → 14.7:** Creative/live admission requires actual privacy, media rights
and consent governance.
- **17.6 → 17.2:** Exercises actual structured document extraction and source
anchors.
- **17.6 → 17.3:** Exercises grounded multimodal understanding and uncertainty
handling.
- **17.6 → 17.4:** Uses the owning domain real generation toolchain and editable
artifacts.
- **17.6 → 17.5:** Requires separate modality quality, human and failure
evidence.
- **17.7 → 0.8:** Extends the ratified source-traceable charter coverage
inventory.
- **17.7 → 17.1:** Uses the actual modality and toolchain capability matrix.
- **17.8 → 14.7:** Requires rights, retention, and source-media governance.
- **17.8 → 17.2:** Uses structured documents, OCR, and source anchors.
- **17.8 → 17.3:** Uses grounded visual, audio, and video understanding.
- **17.8 → 17.7:** Uses the creative input and output contracts.
- **17.9 → 17.7:** Uses explicit brief, output, and checkpoint requirements.
- **17.9 → 17.8:** Consumes grounded and attributed references.
- **17.10 → 15.1:** Uses approved capability-correct and priced model bindings.
- **17.10 → 17.4:** Uses actual owning-domain generation tooling.
- **17.10 → 17.9:** Consumes approved shot plans and asset slots.
- **17.11 → 5.25:** Executes through the attributed live AE path.
- **17.11 → 17.9:** Compiles approved versioned storyboard semantics.
- **17.11 → 17.10:** Uses selected and validated generated asset bytes.
- **17.12 → 5.28:** Uses durable creative execution and cancellation/recovery.
- **17.12 → 17.9:** Binds feedback to immutable storyboard revisions.
- **17.12 → 17.11:** Applies changes to actual editable AE projects.
- **17.13 → 5.24:** Packages real reopenable AE project and render artifacts.
- **17.13 → 5.26:** Uses actual cross-domain artifact storage and transfer.
- **17.13 → 14.7:** Requires media rights, retention, and deletion governance.
- **17.14 → 17.7:** Uses the architectural accuracy and output requirements.
- **17.14 → 17.8:** Consumes measured and source-anchored floor-plan evidence.
- **17.15 → 5.6:** Uses governed and independently verified Blender execution.
- **17.15 → 5.20:** Requires correct native save/reopen and export verification.
- **17.15 → 17.14:** Uses the dimensioned architectural scene contract.
- **17.16 → 12.9:** Uses calibrated architectural visual and dimensional quality
criteria.
- **17.16 → 17.10:** Uses verified generated/imported assets and provenance.
- **17.16 → 17.15:** Preserves independently verified architectural geometry.
- **17.17 → 5.27:** Uses semantically verified Blender-to-Unreal interchange.
- **17.17 → 17.15:** Uses editable dimensionally checked geometry.
- **17.17 → 17.16:** Uses reviewed reference-based materials, lighting, and
furnishings.
- **17.18 → 5.29:** Requires real AE breadth and failure evidence.
- **17.18 → 7.10:** Uses creative skills proven on held-out briefs.
- **17.18 → 8.11:** Requires the actual review and delivery user journey.
- **17.18 → 12.10:** Uses the admitted creative-family evidence and quality
release gate.
- **17.18 → 17.11:** Uses real native motion compilation and artifacts.
- **17.19 → 5.29:** Requires real AE breadth and failure evidence.
- **17.19 → 7.10:** Uses transferable creative recipes rather than memorized
clips.
- **17.19 → 8.11:** Requires the actual review and delivery user journey.
- **17.19 → 12.10:** Uses the original-motion family quality gate.
- **17.19 → 17.11:** Produces editable native compositions from approved plans.
- **17.20 → 8.11:** Requires the actual review and delivery user journey.
- **17.20 → 12.10:** Uses the architectural and walkthrough family gates.
- **17.20 → 17.13:** Delivers verified portable native project packages.
- **17.20 → 17.17:** Requires a delivered and tested interactive walkthrough.
- **17.21 → 17.18:** Requires accepted held-out reference-motion delivery
evidence.
- **17.21 → 17.19:** Requires accepted original-brief motion delivery evidence.
- **17.21 → 17.20:** Requires accepted dimensional reconstruction and
walkthrough evidence.
- **17.22 → 4.8:** Live admission requires the security red-team gate;
preparation may proceed before it.
- **17.22 → 14.7:** Requires generated-media licence, disclosure and provenance
governance.
- **17.22 → 15.7:** Uses the admitted generation legs and priced route choices.
- **17.22 → 17.4:** Uses the owning domain real generation toolchain and
editable artifacts.
- **17.23 → 8.14:** Uses the shared rich renderer for inline images, players and
viewers.
- **17.23 → 8.15:** Uses the artifact workspace for variants, edits and saves.
- **17.23 → 8.19:** Uses progress and completion delivery for generation jobs.
- **17.23 → 17.22:** Requires admitted, validated, provenance-bearing generation
tools.
- **17.24 → 8.13:** Requires the shipped multimodal composer and attachment
envelope.
- **17.24 → 15.7:** Uses the admitted vision and speech legs through the
registry control.
- **17.24 → 17.2:** Exercises actual structured document extraction and source
anchors.
- **17.24 → 17.3:** Exercises grounded multimodal understanding and uncertainty
handling.
- **17.25 → 4.5:** Requires execution isolation for code with scoped resources.
- **17.25 → 4.8:** Live admission requires the security red-team gate;
preparation may proceed before it.
- **17.25 → 7.1:** Uses the recorded Hermes fusion scope for the code-execution
tool.
- **17.25 → 8.15:** Delivers files and charts into the artifact workspace.
- **17.26 → 4.8:** Live admission requires the security red-team gate;
preparation may proceed before it.
- **17.26 → 7.1:** Uses the recorded Hermes fusion scope for web search and
fetch tools.
- **17.26 → 8.15:** Delivers cited research reports into the artifact workspace.
- **17.26 → 8.18:** Uses the composer mode picker for search and research modes.
- **17.26 → 8.19:** Uses progress and completion delivery for research runs.
- **17.27 → 12.11:** Uses the admitted parity and generated-media evaluation
family.
- **17.27 → 17.23:** Requires delivered generated media in the conversation.
- **17.27 → 17.24:** Requires grounded answers from attachments.
- **17.27 → 17.25:** Requires admitted sandboxed data analysis.
- **17.27 → 17.26:** Requires admitted web search and deep research.
- **18.1 → 0.8:** Requires ratified source coverage so required workflows cannot
be omitted.
- **18.1 → 1.9:** Requires required-workbench completion to reject deferred or
missing seams.
- **18.1 → 2.8:** Requires the actual governed fleet drain and soak evidence.
- **18.1 → 3.8:** Requires operational retrieval freshness, recovery and canary
evidence.
- **18.1 → 4.8:** Live admission requires the security red-team gate;
preparation may proceed before it.
- **18.1 → 5.11:** Requires every selected runtime package, including newly
discovered children, to be delivered.
- **18.1 → 6.8:** Requires native injection and benign continuation proof.
- **18.1 → 7.8:** Requires explicitly chosen, live-gated channel operation.
- **18.1 → 8.9:** Requires actual cross-client journeys and server/tool
round-trip proof.
- **18.1 → 9.7:** Requires measured memory rollout, usefulness and zero-harm
evidence.
- **18.1 → 10.4:** Requires actual V1.2 availability and restored-semantics live
evidence.
- **18.1 → 11.8:** Requires the full charter benchmark and complete observation
windows.
- **18.1 → 12.6:** Requires online feedback and drift evidence linked to offline
cases.
- **18.1 → 12.8:** Requires the full evidence-joining charter gate
implementation, independently of later production results.
- **18.1 → 13.7:** Live admission requires operational detect, kill, recover and
verify readiness.
- **18.1 → 14.7:** Creative/live admission requires actual privacy, media rights
and consent governance.
- **18.1 → 15.6:** Requires bounded capacity and denial-of-wallet evidence.
- **18.1 → 16.6:** Requires conformance against independent counterpart
implementations.
- **18.1 → 17.6:** Requires real governed cross-modal delivery and verification.
- **18.1 → 17.21:** Requires all added creative workflows, dependencies, and
truthful capability claims to reconcile before closure.
- **18.1 → 17.27:** Requires generated-media and agentic chat claims to
reconcile with the parity register before closure.
- **18.2 → 18.1:** Requires every gap and required charter workflow to reconcile
to direct evidence.
- **18.3 → 0.4:** Uses the versioned source/control crosswalk to define
applicable obligations.
- **18.4 → 18.1:** Requires every gap and required charter workflow to reconcile
to direct evidence.
- **18.4 → 18.2:** Requires the fresh source/runtime audit with new gaps owned.
- **18.4 → 18.3:** Requires current source and comparative capability gaps to be
resolved.
- **18.5 → 18.1:** Requires every gap and required charter workflow to reconcile
to direct evidence.
- **18.5 → 18.2:** Requires the fresh source/runtime audit with new gaps owned.
- **18.6 → 18.4:** Requires the clean-state full charter and operator benchmark
evidence.
- **18.6 → 18.5:** Requires the final security/reliability game day without
hard-lock events.
- **18.7 → 18.1:** Requires every gap and required charter workflow to reconcile
to direct evidence.
- **18.7 → 18.2:** Requires the fresh source/runtime audit with new gaps owned.
- **18.7 → 18.3:** Requires current source and comparative capability gaps to be
resolved.
- **18.7 → 18.4:** Requires the clean-state full charter and operator benchmark
evidence.
- **18.7 → 18.5:** Requires the final security/reliability game day without
hard-lock events.
- **18.7 → 18.6:** Requires all applicable final verification and derived
artifacts.
- **18.8 → 18.1:** Requires every gap and required charter workflow to reconcile
to direct evidence.
- **18.8 → 18.2:** Requires the fresh source/runtime audit with new gaps owned.
- **18.8 → 18.3:** Requires current source and comparative capability gaps to be
resolved.
- **18.8 → 18.4:** Requires the clean-state full charter and operator benchmark
evidence.
- **18.8 → 18.5:** Requires the final security/reliability game day without
hard-lock events.
- **18.8 → 18.6:** Requires all applicable final verification and derived
artifacts.
- **18.8 → 18.7:** Requires accurate open-blocker reporting and current evidence
documentation.
## Limitations
- Open rows contain evidence plans only; a planned scope or manifest path is not
an executed result.
- Completed Phase-0 rows link direct repository evidence created before or
during schema admission; later phase closure still requires task-specific
evidence manifests.
- A matrix verifier proves ownership, dependency, and proof-boundary integrity,
not the truth of a future runtime claim.