xeonvs · Oct 23, 2026 · Oct 23, 2026 · Oct 23, 2026 · Oct 23, 2026 · Oct 23, 2026
@@ -47,8 +47,8 @@ Read the matching owners and run their applicable guards. The general repository
 <!-- ew:route id="long-running-execution" triggers="long-running commands|builds|tests|local processes|polling|terminal sessions" owners="docs/engineering/project_principles.md|docs/development.md" guards="manual_review:verify process completion exit status bounded output and task-owned cleanup" -->
 | `long-running-execution` | Long-running command, waiter, or process cleanup | Project principles and development procedures | Completion evidence and bounded output |
 
-<!-- ew:route id="local-ocr-review" triggers="local OCR review|ocr-ci review --local|review artifacts" owners="docs/local.md|docs/development.md" guards="manual_review:verify selected auth mode qualified executable native model reachability supported local flags physical private paths terminal status report and admitted findings" -->
-| `local-ocr-review` | Authorized real local OCR review | Local review and development procedures | Mode, binary, flags, private paths, terminal status, report and findings |
+<!-- ew:route id="local-ocr-review" triggers="local OCR review|ocr-ci review --local|review artifacts" owners="docs/local.md|docs/development.md" guards="manual_review:verify selected auth mode qualified executable native model reachability supported local flags physical private paths terminal status report and admitted findings|manual_review:inspect failed auxiliary reads for material review gaps" -->
+| `local-ocr-review` | Authorized real local OCR review | Local review and development procedures | Mode, binary, flags, private paths, result, findings and failed reads |
 
 <!-- ew:route id="workflow-instructions" triggers="AGENTS.md|docs/engineering/project_principles.md|docs/codex/AGENT_EXECUTION_PITFALLS.md" owners="skill://engineering-workflow/references/instruction_lifecycle.md|docs/engineering/project_principles.md" guards="lint:instruction-contract" -->
 | `workflow-instructions` | Instruction owner, route, or incident change | Installed instruction lifecycle and project principles | `lint:instruction-contract` |
@@ -59,8 +59,8 @@ Read the matching owners and run their applicable guards. The general repository
 <!-- ew:route id="release-lifecycle" triggers="docs/release.md|.github/workflows/release.yml|release-required change|release-deferred change|stable publication|release authorization|stable delivery|release closure" owners="docs/release.md" guards="release_gate:.github/workflows/release.yml|manual_review:reconcile current archive and external receipts" -->
 | `release-lifecycle` | Release authorization, stable delivery, or closure | Release guide | Protected release gate and external receipt review |
 
-<!-- ew:route id="development-validation" triggers="docs/development.md|status-bearing documents|parser changes|provider changes|subprocess changes" owners="docs/development.md" guards="manual_review:compare status documents with current implementation|test:scripts/quality.sh check" -->
-| `development-validation` | Status or changed-boundary validation | Development guide | Status review and Python quality matrix |
+<!-- ew:route id="development-validation" triggers="docs/development.md|status-bearing documents|parser changes|provider changes|subprocess changes" owners="docs/development.md" guards="manual_review:compare status documents with current implementation|manual_review:map changed producer-consumer handoffs and failure cleanup|test:scripts/quality.sh check" -->
+| `development-validation` | Status or changed-boundary validation | Development guide | Status and handoff review; Python quality matrix |
 
 <!-- ew:route id="dependency-audit" triggers="pre-push|pyproject.toml|uv.lock|.github/workflows/security.yml|scripts/dependency-audit.sh" owners="docs/development.md" guards="test:scripts/dependency-audit.sh" -->
 | `dependency-audit` | Every push and dependency or audit-workflow change | Development guide | Shared local and hosted dependency audit |

@@ -90,6 +90,12 @@ On a successful rerun, the toolkit replaces untouched OCR-only notes instead of
 
 Suppression uses both the GitLab diff position and a stable finding fingerprint, so ordinary line shifts do not normally bring the same bug back. A materially changed finding can still receive a new discussion. See [GitLab review operations](docs/operations.md) for the complete lifecycle, posting modes, permissions, failure behavior, and Mermaid state diagram.
 
+For a safely separable changed Git submodule pointer under qualified OCR
+1.13.3, the toolkit reviews selected ordinary files and labels repository
+coverage partial. It leaves the pointer and submodule contents unreviewed,
+preserves prior finding threads and cannot automatically approve. A diff with
+no selected ordinary files starts no model review and posts an advisory status.
+
 After every current review note publishes, the GitLab adapter can add a conservative approval bound to receipt v11's exact reviewed diff, actual target, policy authority, merge-request author, DLP mode, and reconciled federation use. This write is enabled by default; set `OCR_AUTO_APPROVE=false` when the bot must remain comment-only. DLP-clean metadata, generic discussions, protected same-revision CI outcomes, unused governed services, and successful `review_read` tool use do not independently block approval. An unprotected target, disabled DLP, degraded metadata, a DLP rejection, required context degradation, admitted remediation history, older receipts, publication filtering, used advisory tools, uncertain federation accounting, author movement, or bot self-authorship prevents an approval write. A CI status is review context, never approval authority. GitLab approval rules, Code Owners, and protected-branch policy remain authoritative merge policy. They are not prerequisites for OCR execution. The toolkit only adds an eligible approval; it never removes an existing approval when a later review is ineligible or disabled.
 
 An intentionally absent toolkit receipt retains compatible direct posting outside

@@ -77,12 +77,13 @@ flowchart LR
   trusted base. Protected stable publication, external readback, and live site
   refresh completed on 2026-10-22.
 
-- The v0.13.0 release candidate separates the current GitLab MR diff and actual
+- The published v0.13.0 release separates the current GitLab MR diff and actual
   target from protected policy authority. The absent protection setting selects
   `auto`; an unprotected target can use its same-project protected default policy
   for single-MR review while remaining comment-only. Receipt v11 and publication
-  freshness checks bind these identities. Stable delivery and external readback
-  are pending. Optional related-MR review remains inactive planning in BL-025.
+  freshness checks bind these identities. The immutable GitHub Release was
+  published and #264 closed on 2026-10-23. Optional related-MR review remains
+  inactive planning in BL-025.
 
 - OCR compatibility and the established common evidence model converge at compact-bootstrap/evidence-MCP integration.
 - M3 preserves BL-011's historical characterization. M7 supersedes that runtime path with governed federation; the old direct composition and adapter execution contracts are not current fallback paths.

@@ -0,0 +1 @@
+GitLab and local reviews can now cover selected ordinary files alongside safely separable changed Git submodule pointers with qualified OCR 1.13.3. Reports identify the unreviewed pointers as partial repository coverage; these runs cannot automatically approve or clear earlier finding discussions. A change with no selected ordinary files stops before the model and posts an advisory GitLab status.
@@ -168,7 +168,7 @@ This is a diagnostic catalogue of recurring incident classes. Current requiremen
 - Owner: docs/engineering/project_principles.md#outcome-consistency
 - Route: trust-boundary
 - Guard: test:scripts/quality.sh check
-- Evidence: [0.4.7 final OCR correction](../engineering/execution_history/releases.md#plan-toolkit-0-4-7). Prior control: table-driven result and posting tests cover skipped, clean, warning, error, finding, partial, and zero-value cases through shared contracts. Historical cause: outcomes were assembled independently and tests asserted prose rather than one result invariant.
+- Evidence: [0.4.7 final OCR correction](../engineering/execution_history/releases.md#plan-toolkit-0-4-7) and the v0.13.1 gitlink review's safe-plus-filtered finding case in `tests/test_repository_coverage.py`. Prior control: table-driven result and posting tests cover skipped, clean, warning, error, finding, partial, and zero-value cases through shared contracts. Historical cause: outcomes were assembled independently and tests asserted prose rather than one result invariant; the v0.13.1 first review found that producer and consumer tests still missed a transformed handoff.
 - Status: guarded
 - Retirement: Retire when this historical class no longer improves diagnosis and its owner, route, and guard remain established.
 
@@ -180,6 +180,6 @@ This is a diagnostic catalogue of recurring incident classes. Current requiremen
 - Owner: docs/local.md#review-preparation
 - Route: local-ocr-review
 - Guard: manual_review:verify selected auth mode qualified executable native model reachability supported local flags physical private paths terminal status report and admitted findings
-- Evidence: The 2026-10-23 v0.13.0 local review attempts rejected absent `OCR_LLM_MODEL` and `--background-file` before model execution; the corrected OCR 1.13.3 invocation completed with 29/29 selected items and an admitted finding. The earlier [v0.12.1 review](../engineering/execution_history/releases.md#plan-toolkit-0-12-1) exposed the symlinked-path report failure. The preparation route now names each check at the launch boundary.
+- Evidence: The 2026-10-23 v0.13.0 local review attempts rejected absent `OCR_LLM_MODEL` and `--background-file` before model execution; the corrected OCR 1.13.3 invocation completed with 29/29 selected items and an admitted finding. The [v0.12.1 review](../engineering/execution_history/releases.md#plan-toolkit-0-12-1) and v0.13.1 review both exposed symlinked-path report failures. The v0.13.1 runner now validates the report parent before model execution and the publisher checks again after execution; `docs/local.md` shows physical-path preparation.
 - Status: guarded
 - Retirement: Retire when the owner, route, and guard remain established and the historical class no longer improves launch diagnosis.
@@ -32,7 +32,7 @@ plan_archive_indexes:
   - docs/engineering/execution_history/README.md
   - docs/engineering/README.md
   - docs/README.md
-active_plan: null
+active_plan: PLANS.md
 runtime_agent_config_managed: true
 instruction_contract_version: 3
 planning_contract_version: 2

@@ -364,6 +364,13 @@ There is no toolkit CI adaptation beyond version and checksum pins. The hosted
 probe does not claim configured-model review quality. Native ChatGPT remains
 local-only; final release review uses the checksum-verified Darwin arm64 binary.
 
+Toolkit 0.13.1 uses this same qualified OCR 1.13.3 as a narrower changed-gitlink
+boundary. It verifies process-local Git projection, the full preview selection,
+and native manifest/source-artifact fingerprints against the original immutable
+range. This does not extend OCR review to submodule contents or qualify a later
+OCR version for partial repository coverage; other versions retain the
+no-model stop until separately proven.
+
 ## Promotion and rollback
 
 Promotion changes `recommended_version`, advances `monitoring_floor`, adds the tested release and evidence, and updates every machine-readable or executable version/checksum pin. Never edit only one copy. Human-qualified candidates must record the compatibility conclusion and release-note impact; an automatic-safe candidate may also record a reviewed conclusion when it is delivered with a human-reviewed chain instead of using the generic machine conclusion. Conclusions may name only versions present in that promotion. A reviewed promotion may cross only one adjacent semantic-version boundary at a time: the next patch, the next minor at `.0`, or the next major at `.0.0`; minor and major transitions always require an explicit human conclusion. Automatic-safe preparation remains limited to adjacent patches in the already-tested major/minor line and retains the same protected review boundary even though the patch itself is mechanical.
Original file line number	Diff line number	Diff line change
		@@ -0,0 +1 @@
		GitLab and local reviews can now cover selected ordinary files alongside safely separable changed Git submodule pointers with qualified OCR 1.13.3. Reports identify the unreviewed pointers as partial repository coverage; these runs cannot automatically approve or clear earlier finding discussions. A change with no selected ordinary files stops before the model and posts an advisory GitLab status.