diff --git a/AGENT_HANDOFF.md b/AGENT_HANDOFF.md
index 0c6580167..9146e1fcb 100644
--- a/AGENT_HANDOFF.md
+++ b/AGENT_HANDOFF.md
@@ -1,5 +1,43 @@
 # AGENT_HANDOFF — Tools
 
+## 2026-08-11 PR #4342 current-parent propagation candidate
+
+This no-publish candidate keeps PR `#4342` on
+`feat/4274-regional-plan-io` with base `feat/4274-regional-surface-ui`.
+Exact current child `c1f47f2ef68b3db102da5416aaac17a40f675207` is the first
+parent and exact reviewed local #4339 candidate
+`db335937afc4b587d235eb705e315f577519c5e6` is the second parent of a
+normal no-ff merge. Child-owned canonical request import/export, bounded UTF-8,
+native atomic save, browser-qualified download, tests, and limitations remain
+intact while inheriting current editor, wire, regional-physics, and complete
+ground-model ancestry.
+
+The merge also resolves one cross-runtime compatibility edge without weakening
+the ground contract. The default shared canonical encoder still rejects floats
+or integers outside JavaScript's safe range. A separately named extended
+finite-float policy reuses the same recursive encoder only through the
+capability-observation facade; integers remain safe-range bounded and integral
+finite doubles such as `1e20` and `1e21` emit exact exponent-free tokens that
+match the TypeScript capability serializer. Non-finite values still fail
+closed.
+
+Local qualification is green: all `909` Rate-of-Closure Python tests, all
+`110` React files / `686` tests, and the complete Cargo workspace pass. The
+focused compatibility/regional-I/O slices pass `47` Python and `12` React
+tests. Pinned Ruff 0.14.10 check/format passes `17` changed Python files; pinned
+MyPy 1.13 and Bandit pass `12` changed production files. TypeScript,
+zero-warning ESLint, the 199-module Vite build, Rust format and warning-denied
+clippy, both 400- and 500-LOC changed-file gates, manifest/docs/blocking-gate/
+assertion/minimum-test governance, marker scans, and diff checks pass. An
+untouched manual-delivery UI test timed out once during the first concurrent
+full run, then passed alone and in the single complete rerun.
+
+No branch has been rebased, retargeted, rewritten, force-pushed, or published.
+Execution and playback, result interchange, measured calibration, model-input
+persistence, changing geometry or velocity, TypeScript/compiled regional
+physics, downstream parity, protected exact-head evidence, approval,
+dependency integration, and release remain open.
+
 ## 2026-08-11 PR #4339 current-parent propagation candidate
 
 This no-publish candidate keeps PR `#4339` on
@@ -32,188 +70,16 @@ model-input persistence, changing geometry or surface velocity,
 TypeScript/compiled regional physics, downstream parity, protected exact-head
 evidence, approval, dependency integration, and release remain open.
 
-## 2026-08-11 PR #4339 structural helper repair candidate
-
-This local follow-up descends exact independently reviewed reconciliation
-candidate `eedad6d23b517eb4c99d4ba9000ff6555101099f`. Independent review found
-one publication blocker: the regional PyQt number-input helper accepted six
-parameters, above the campaign's mandatory four-parameter limit. RED-first
-tests now bind the helper's signature and all observable widget settings. A
-frozen `NumberInputSpec` validates finite positive step and finite ordered
-bounds before widget construction; the three-parameter helper preserves field
-names, values, units, ranges, increments, tooltips, and presentation order.
-
-Eight focused Python/PyQt tests and 25 focused React tests pass; the complete
-React suite passes all 108 files / 672 tests. Pinned Ruff check/format, MyPy on
-all six affected production modules, Bandit medium/high screening, TypeScript,
-zero-warning ESLint, documentation/manifest governance, changed-production,
-test-assertion, minimum-test, 500-LOC, and diff gates pass. Mandatory maxima
-are 400 module lines, 42 function lines, and three parameters. A broader
-Rate-of-Closure Python run reached the ten-minute workstation lifecycle bound
-without yielding an assertion failure; it is not counted as passing evidence.
-The monorepo-wide lane separately exposed unrelated baseline failures and
-xdist worker termination. No schema, digest, regional physics, UI behavior,
-PR base, protected evidence, or release boundary changes.
-
-## 2026-08-11 PR #4339 current-parent ancestry candidate
-
-The clean `feat/4274-regional-surface-ui` worktree normally merges exact
-published child `cbb9c0a6bdc6a50f59f7a661139b9d53e1892980` with exact newly
-published #4335 parent `9e01ccc3e891cc45907293751a192624195a77a5`,
-while preserving #4339's `feat/4271-regional-wire-contract` base. Production
-UI, contract, resolver, regional physics, and ground ancestry merge without
-conflict. Only SPEC, manifest, and handoffs are reconciled. Independent review,
-ordinary fast-forward publication, fresh protected CI/approval, dependency
-integration, and release remain required.
-
-## 2026-08-10 PR #4339 stale-validation invalidation follow-up
-
-Rendered exact-head browser QA found that a validated one-overlay readback
-remained visible after adding a second overlay. The broad PyQt6 suite also
-found missing hover hints on the new controls. PyQt6 and React now clear both
-canonical readback and prior validation state whenever any identity, interval,
-material, or overlay-row draft value changes; PyQt6 exposes the explicit
-"Changes not validated" pending state, and every PyQt6 interactive has a
-specific engineering-context tooltip. RED-first regression tests cover the
-dynamic-row path on both surfaces. Final local evidence is 870 Python/PyQt
-tests and 672 React tests passing; Ruff, format, MyPy, TypeScript, zero-warning
-ESLint, the 198-module production build, manifest/docs/file-size governance,
-and diff checks are clean. The known polynomial-generator empty-legend warning
-is unrelated. This changes no wire schema, provenance digest, physics, or
-persistence boundary; fresh protected CI is still required before merge.
-
-## 2026-08-11 Current regional-surface parent reconciled into PR #4335
-
-Open draft PR `#4335` retains branch `feat/4271-regional-wire-contract` and
-base `feat/4271-regional-surface-transitions`. Exact remote regional-wire child
-`9e01ccc3e891cc45907293751a192624195a77a5` is merged first and exact current
-regional-surface parent `5ac3c1f233de30dd42eeac57073cbdb149bde652`
-second by an ordinary no-ff merge. The child's Python/TypeScript wire schemas,
-canonical JSON/SHA-256 fixture, fail-closed parsing, Python resolver binding,
-tests, contracts, and limitations remain intact with both append-only handoff
-and specification histories.
-
-Focused local qualification is green: all `132` Python ground/regional-wire
-tests, `9` focused React regional/ground-contract tests, and all `137`
-`tools-core` Rust tests pass. TypeScript, zero-warning ESLint, the 190-module
-Vite build, Rust workspace format and warning-denied clippy, pinned Ruff 0.14.10
-across the six-file Python PR delta, pinned MyPy 1.13 and Bandit medium/high
-across four production files, the 500-LOC changed-file gate, manifest parsing
-plus all eight contracts, documentation governance, exact wire-tree
-preservation, history/marker, and diff checks pass.
-
-This candidate is unpublished and `not_released`. Changing geometry or
-velocity, internal transition-ledger export, regional UI, compiled or
-TypeScript regional physics, UpstreamDrift parity, protected exact-head CI,
-review, approval, dependency integration, and release remain open.
-
-## 2026-08-11 Current skid/roll parent reconciled into PR #4332
-
-Open draft PR `#4332` retains branch
-`feat/4271-regional-surface-transitions` and base
-`feat/4271-ground-skid-roll`. Exact remote regional-surface child
-`04ccf08dd990de1cd056a3420e67772773a4be2e` is merged first and exact current
-skid/roll parent `3f861c1faaa7455faee92eaa7f813174667208e0` second by an ordinary no-ff
-merge. The child's bounded coplanar overlays, exact transition splitting and
-evidence, regional production, tests, fixtures, contracts, and limitations
-remain intact with both append-only handoff and specification histories.
-
-Focused local qualification is green: all `121` Python ground/regional tests,
-`20` focused React ground-contract/transfer tests, and all `137` `tools-core`
-Rust tests pass. TypeScript, zero-warning ESLint, the 189-module Vite build,
-Rust workspace format and warning-denied clippy, pinned Ruff 0.14.10 across the
-14-file Python PR delta, pinned MyPy 1.13 and Bandit medium/high across 12
-production files, the 500-LOC changed-file gate, manifest parsing plus all
-eight contracts, documentation governance, exact regional-tree preservation,
-history/marker, and diff checks pass.
-
-This candidate is unpublished and `not_released`. Changing normals,
-discontinuous geometry or surface velocity, regional UI, compiled regional
-physics, UpstreamDrift parity, protected exact-head CI, review, approval,
-dependency integration, and release remain open.
-
-## 2026-08-11 Current impact/bounce parent reconciled into PR #4304
-
-Open draft PR `#4304` retains branch `feat/4271-ground-skid-roll` and base
-`feat/4270-ground-impact-bounce`. Exact remote skid/roll child
-`0ea6740965068542e9d8c7449e06ec07d88969e0` is merged first and exact current
-impact/bounce parent `62beb3e1aa951645fd556a53f9cbae4bb46c47e5`
-second by an ordinary no-ff merge. The child's arbitrary-plane skid, pure roll,
-rolling resistance, rest, edge localization, strict result composition,
-passive ledgers, fixtures, tests, and limitations remain intact with both
-append-only histories.
-
-Local verification is green: `154` focused ground/flight physics and contract
-tests plus all eight campaign-manifest tests pass, as do all `137` `tools-core`
-Rust tests. Pinned Ruff 0.14.10 check/format covers all 18 PR-delta Python files;
-pinned MyPy 1.13 and Bandit medium/high cover all 12 production files. Rust
-format and warning-denied clippy, manifest parsing/validation, documentation
-governance, exact skid/roll-tree preservation, marker/history, and diff checks
-pass.
-
-This local candidate is not published or released. Regional/changing-normal
-surfaces, deformation, torsional damping, roll-to-skid transitions, UI,
-compiled physics, protected exact-head CI, review, approval, downstream
-integration, and release remain open.
-
-## PR #4302 corrected-flight-transfer propagation
-
-Draft PR #4302 remains on `feat/4270-ground-impact-bounce` with unchanged base
-`feat/4269-flight-ground-transfer`. Exact corrected #4288 parent
-`247215422a6d4b677552955b4923bc609a553259` is incorporated by the normal merge
-containing this handoff. The child preserves its passive restitution/Coulomb
-sphere-plane impulse, deterministic repeated ballistic hops, exact physical
-contact, capture-to-skid handoff, cancellation/failure semantics, and airborne
-segment evidence while inheriting the corrected transfer ancestry, deterministic
-workspace timestamps, and canonical `swing_sim` import identity. No branch was
-rebased, retargeted, rewritten, or force-pushed.
-
-The slice remains partial and `not_released`: issue #4271 retains skid, roll,
-rest, total distance, and final `GroundSimulationResult`; terrain deformation,
-UI execution, TypeScript bounce physics, compiled Rust/PyO3/WASM bounce parity,
-and UpstreamDrift adapters remain excluded. Protected CI, independent review,
-and dependency-ordered collapse remain separate release gates.
-
-Merged-tree validation is `987` Python tests, `106` React files / `661` tests,
-and the complete `tools-core` Rust suite at `137` tests (`111` unit, `20`
-transfer, `6` wire). The affected cross-version suite is also `146` tests on
-real CPython 3.10.20. The 189-module Vite production build, TypeScript,
-zero-warning ESLint, Ruff check/format across 90 files, pinned mypy 1.13 across
-all 17 ground and nine transfer production modules, Rust workspace format plus
-warning-denied `tools-core` clippy, campaign-manifest validator plus eight
-contracts, documentation governance, 11-file 500-LOC budget, marker scan, and
-diff checks are clean. Hosted checks and review apply to the new exact merge
-head only.
-
-
-## PR #4288 corrected-ground propagation
-
-Draft PR #4288 remains on `feat/4269-flight-ground-transfer` with unchanged
-base `feat/4268-ground-contract`. Exact corrected #4285 parent
-`788aa547651a3685a363ea401824a5d81477bafb` is incorporated by the normal merge
-containing this handoff. The child preserves its signed terminal state,
-physical sphere/terrain contact brackets, strict provenance, and qualified
-Python/TypeScript/Rust/PyO3/WASM transfer while inheriting deterministic
-workspace timestamps and canonical `swing_sim` import identity. No branch was
-rebased, retargeted, rewritten, or force-pushed.
-
-This propagation adds no bounce, skid, roll, terrain response, total distance,
-or UI execution. Protected CI, independent review, and dependency-ordered
-collapse remain separate release gates.
-
-Merged-tree validation is `951` Python tests, `106` React files / `661` tests,
-and `26` focused Rust transfer/wire tests. The affected cross-version suite is
-also `110` tests on real CPython 3.10.20. The 189-module Vite production build,
-TypeScript, zero-warning ESLint, Ruff check/format across 82 files, pinned mypy
-1.13 across the six-file transfer and 13-file parent production namespaces,
-Rust workspace format plus warning-denied `tools-core` clippy, campaign-manifest
-validator plus eight contracts, documentation governance, 13-file 500-LOC
-budget, marker scan, and diff checks are clean. Hosted checks and review apply
-to the new exact merge head only.
 
+## 2026-08-11 PR #4342 append-only preservation repair
 
-> Update this file in every implementation commit and every push to `main`.
-> Last updated: 2026-08-11.
+Independent audit of local merge `e7fedfb18de1550eed3484ed2fc99d0baaecdca1`
+found that three older parent ancestry sections had been summarized instead of
+retained verbatim. This docs-only follow-up restores the full parent text below
+without changing production code, tests, schemas, manifest state, PR base, or
+release claims. The exact parent sections and SPEC rows are now governed by
+byte-for-byte comparison against
+`d21741e312b849a63f73cabf351a15d9de80fb94`.
 
 ## 2026-08-11 PR #4335 current-parent ancestry candidate
 
@@ -287,67 +153,6 @@ now records open PR #4332 at its still-published child head and open PR #4304
 at its exact newly published parent head; neither record claims protected or
 main-release evidence.
 
-## 2026-08-11 Current flight-transfer parent reconciled into PR #4302
-
-Open draft PR `#4302` retains branch `feat/4270-ground-impact-bounce` and base
-`feat/4269-flight-ground-transfer`. Exact remote bounce child
-`cf6e72bad98e5f36f782254942e6895b8b71e670` is merged first and exact current
-flight-transfer parent `c39297d37a17d5d8d3520ed62c5563cdcc609cab` second by an
-ordinary no-ff merge. Passive impact/restitution/friction, deterministic
-repeated bounce, exact contact, capture-to-skid handoff, bounded failure and
-cancellation, and all transfer production, tests, fixtures, and limitations
-remain intact. Both append-only histories are retained.
-
-The exact PR delta also receives the three AST-equivalent layouts required by
-the repository-pinned Ruff 0.14.10 formatter; no analysis, persistence, or
-contract assertion behavior changes.
-
-Local verification is green: `1129` Rate/ground/flight Python tests, `25`
-focused React transfer tests, and all `137` `tools-core` Rust tests pass.
-TypeScript, zero-warning ESLint, the 189-module Vite build, Rust format and
-warning-denied clippy, pinned Ruff 0.14.10 check/format across the 14-file PR
-delta, pinned MyPy 1.13 across eight production files, Bandit medium/high,
-campaign-manifest parsing, documentation preservation, and diff checks pass.
-
-This local candidate is not published or released. Skid, roll, rest, total
-distance, final ground results, UI, TypeScript bounce physics, and compiled
-bounce runtimes remain open, as do protected exact-head CI, independent review,
-approval, dependency integration, and release.
-
-## PR #4285 workspace timestamp propagation
-
-Draft PR #4285 remains on `feat/4268-ground-contract` with unchanged base
-`feat/4197-capability-observer`. Exact corrected #4282 carrier
-`5f77af4add23547a21cc3fabce98ae9ad4260427` is incorporated through the normal
-merge containing this handoff. No branch was rebased, retargeted, rewritten,
-or force-pushed.
-
-The child retains its strict, UI-neutral flight-to-ground schemas, canonical
-fixture, migrations, legacy-result adapter, explicit dependency, and pinned
-typing repairs while inheriting the deterministic Python 3.10-3.12 UTC parser
-and the corrected variation, scalar, wind, capability, and campaign-release
-ancestry. This merge does not claim bounce, skid, roll, terrain profiles,
-total distance, UI execution, or Rust/WASM delivery. Protected CI, review, and
-normal descendant propagation remain separate release gates.
-
-Exact-head CI on the preceding child head exposed an actionable collection
-defect after installing the package: embedded `src.shared.python.swing_sim`
-tests and canonical `shared.python.swing_sim` imports could load distinct
-package trees, making the ground and impact subpackages unavailable by import
-order. The shared alias registry now coalesces `swing_sim`; an isolated RED/GREEN
-identity contract and both affected public-API suites pin the correction. The
-separate file-size job was cancelled during checkout and did not execute its
-budget check, so it remains infrastructure evidence rather than a source
-failure.
-
-Merged-tree evidence is `915` Rate/ground/impact Python tests, the `12`-test
-shared-alias architecture suite, `28` focused ground/impact/alias tests on
-CPython 3.12, `28` direct compatibility tests on real CPython 3.10.20, and
-`104` React files / `642` tests. The 188-module Vite build, TypeScript,
-zero-warning ESLint, Ruff, format, pinned mypy 1.13, campaign-manifest
-validation plus eight contracts, documentation governance, the 44-file
-500-LOC budget, conflict-marker scan, and diff checks are clean.
-
 ## 2026-08-11 PR #4304 current-parent ancestry candidate
 
 The clean dedicated `feat/4271-ground-skid-roll` worktree first fast-forwarded
@@ -370,6 +175,180 @@ the official changed-file 500-LOC budget, and mandatory production limits of
 local merge is a PUBLISH candidate for independent review only; protected
 exact-head CI, approval, normal child propagation, and publication remain open.
 
+
+## 2026-08-11 PR #4342 current-parent reconciliation candidate
+
+The clean `feat/4274-regional-plan-io` worktree normally merges exact
+published PR #4342 child `8e1c7ccd99a7c4886c5fb9ccc7e4d94a6d7e3833`
+with exact newly published PR #4339 parent
+`d21741e312b849a63f73cabf351a15d9de80fb94` as its second parent. PR #4342
+continues to target `feat/4274-regional-surface-ui`; neither branch is rebased,
+retargeted, rewritten, force-pushed, or published by this reconciliation.
+The merge preserves strict regional request import/export together with the
+parent's frozen engineering-number specification and bounded helper.
+
+The child widget module now delegates its canonical-precision controls to that
+shared three-parameter helper while retaining the inclusive cross-runtime safe
+number bounds and eleven-decimal presentation required by its I/O contract.
+Merged-tree qualification is 53 focused Python/PyQt/shared-ground tests and 14
+focused React tests passing. Pinned Ruff 0.14.10 check/format, MyPy 1.13 over
+all ten affected production modules, Bandit medium/high screening, TypeScript,
+zero-warning ESLint, and the 199-module production build pass. Campaign
+manifest validation and all eight manifest tests, documentation and tool
+manifest governance, changed-Python, minimum-test, exact-diff assertion,
+500-LOC, parent/child diff, and whitespace gates pass. Mandatory structural
+maxima are 396 module lines, 40 function lines, and three parameters.
+The local merge is not protected or release evidence. Independent review,
+ordinary fast-forward publication, fresh protected CI/approval, dependency
+integration, and release remain required.
+
+## 2026-08-11 PR #4339 structural helper repair candidate
+
+This parent follow-up descends exact independently reviewed reconciliation
+candidate `eedad6d23b517eb4c99d4ba9000ff6555101099f`. Independent review found
+one publication blocker: the regional PyQt number-input helper accepted six
+parameters, above the campaign's mandatory four-parameter limit. RED-first
+tests bind the helper's signature and all observable widget settings. A frozen
+`NumberInputSpec` validates finite positive step and finite ordered bounds
+before widget construction; the three-parameter helper preserves field names,
+values, units, ranges, increments, tooltips, and presentation order.
+
+Eight focused Python/PyQt tests and 25 focused React tests pass; the complete
+React suite passes all 108 files / 672 tests. Pinned Ruff check/format, MyPy,
+Bandit medium/high screening, TypeScript, zero-warning ESLint,
+documentation/manifest governance, changed-production, test-assertion,
+minimum-test, 500-LOC, and diff gates pass. Mandatory maxima are 400 module
+lines, 42 function lines, and three parameters. No schema, digest, regional
+physics, UI behavior, PR base, protected evidence, or release boundary changes.
+
+## 2026-08-11 PR #4339 current-parent ancestry candidate
+
+The clean `feat/4274-regional-surface-ui` worktree normally merged exact
+published child `cbb9c0a6bdc6a50f59f7a661139b9d53e1892980` with exact published
+#4335 parent `9e01ccc3e891cc45907293751a192624195a77a5`, while preserving
+#4339's `feat/4271-regional-wire-contract` base. Production UI, contract,
+resolver, regional physics, and ground ancestry merged without conflict; only
+SPEC, manifest, and handoffs were reconciled. Independent review, ordinary
+publication, protected exact-head CI/approval, dependency integration, and
+release remained open.
+
+## 2026-08-11 PR #4342 delta-MyPy follow-up
+
+Protected CI on exact PR #4342 head
+`cffe349ac0a8054f1d168cb36684fd00bc5f8a49` identified one redundant `bool`
+cast in the regional atomic-write adapter under the Linux delta-MyPy gate. The
+cast is removed; the typed helper's direct Boolean return is preserved. This
+changes no wire bytes, validation, file semantics, UI behavior, or physics.
+Focused persistence tests and the CI-equivalent pinned MyPy command pass
+locally; fresh protected CI/review and dependency ordering remain required.
+
+## 2026-08-11 regional request I/O protected publication
+
+Branch `feat/4274-regional-plan-io` is published normally as draft PR
+[#4342](https://github.com/D-sorganization/Tools/pull/4342), targeting exact
+PR #4339 branch `feat/4274-regional-surface-ui` at parent head
+`cbb9c0a6bdc6a50f59f7a661139b9d53e1892980`. Its reviewed implementation
+head is `d748e7a5ef3da5e6ce7737ff6829e0f14665fe97`; publication commits change
+no runtime behavior. Protected CI, independent review, issue #4274, parent
+ordering, integration, and release remain open.
+
+## 2026-08-11 bounded regional request read follow-up
+
+Independent review of local safe-number commit
+`10c394f6b1fd2927e7f3b1f96cc097cae6bfd380` found that native request import
+checked file size with `stat()` and then performed a separate unbounded text
+read. A concurrently grown or replaced file could therefore allocate beyond
+the 1 MiB wire cap before the strict parser rejected it.
+
+Native import now opens one binary handle, reads at most the wire cap plus one
+sentinel byte, rejects overflow, strictly decodes UTF-8, and only then delegates
+to the unchanged canonical v1 parser. Tests simulate content growth after the
+metadata check and reject invalid UTF-8 explicitly. The complete regional and
+atomic persistence set is 28 tests passing. This changes no schema, digest,
+physics, browser behavior, or write semantics. Static and governance evidence
+is recorded in this commit: Ruff, Ruff format, Black, focused MyPy, manifest,
+module-size, documentation-governance, and diff checks pass. The child remains
+local and unprotected.
+
+## 2026-08-11 regional request I/O safe-number follow-up
+
+Independent review of local commit `e39edf4b50b1fb9811b0032bec4758c7a08c9b74`
+found two cross-runtime representation gaps. The PyQt6 precedence spin box
+silently narrowed otherwise valid v1 integers above 1,000,000, and Python
+accepted integer-valued wire numbers above JavaScript's exact safe range.
+
+The native editor now uses an exact decimal integer entry for precedence and
+preserves every v1 value from zero through 9,007,199,254,740,991 without a
+floating-point conversion. Shared Python canonical-number and ground-record
+validation now reject magnitudes outside that same range, matching React while
+retaining all finite fractional values inside it. PyQt6 floating-point editors
+publish the same bounds. Regression coverage pins exact maximum-precedence
+open/save round-trip and matched native/browser refusal of unsafe material
+numbers. This changes no schema version, provenance algorithm, physics, or
+browser filesystem limitation. Final local evidence is 132 shared ground tests,
+21 regional editor/I/O tests, five atomic-workspace tests, 23 focused React
+tests, the whole-shell tooltip sweep, and eight manifest tests passing. Ruff,
+Ruff format, Black, focused MyPy, TypeScript, zero-warning ESLint, the 199-module
+production build, module and changed-file size budgets, documentation
+governance, and diff checks pass. The child remains local and unprotected.
+
+## 2026-08-10 PR #4339 stale-validation invalidation follow-up
+
+Rendered exact-head browser QA found that a validated one-overlay readback
+remained visible after adding a second overlay. The broad PyQt6 suite also
+found missing hover hints on the new controls. PyQt6 and React now clear both
+canonical readback and prior validation state whenever any identity, interval,
+material, or overlay-row draft value changes; PyQt6 exposes the explicit
+"Changes not validated" pending state, and every PyQt6 interactive has a
+specific engineering-context tooltip. RED-first regression tests cover the
+dynamic-row path on both surfaces. Final local evidence is 870 Python/PyQt
+tests and 672 React tests passing; Ruff, format, MyPy, TypeScript, zero-warning
+ESLint, the 198-module production build, manifest/docs/file-size governance,
+and diff checks are clean. The known polynomial-generator empty-legend warning
+is unrelated. This changes no wire schema, provenance digest, physics, or
+persistence boundary; fresh protected CI is still required before merge.
+
+> Update this file in every implementation commit and every push to `main`.
+> Last updated: 2026-08-11.
+
+## 2026-08-10 issue #4274 canonical regional request I/O local child
+
+Local branch `feat/4274-regional-plan-io` started directly from published PR
+#4339 head `a9ace5052bcd54b78f79b62d5d9ac26debedb4b1`, then fast-forwarded
+normally to corrected exact parent `cbb9c0a6bdc6a50f59f7a661139b9d53e1892980`.
+The parent's stale-validation invalidation and tooltip follow-up are preserved.
+This child has not been pushed, opened as a PR, reviewed, merged, or released.
+
+The matched Ground Surfaces editors now import and export the canonical
+`ground-regional-material-plan-request/v1` document. PyQt6 uses native Open
+and Save As dialogs plus a shared flush/fsync/atomic-replace UTF-8 writer;
+cancel is a no-op, failed reads and writes preserve the last-known-good editor
+or file, and the recent path advances only after success. React validates a
+bounded browser File before committing state and downloads exact canonical
+bytes while revoking its object URL. Browser filesystem limitations are
+explicit, and workspace persistence remains separate.
+
+Import accepts only editor producer/provider v1 evidence, the fixed target
+frame origin/downrange axis, and one to eight rows. It never relabels external
+evidence. An unchanged import round-trips the exact request and provenance;
+after any edit, validation binds a fresh draft digest. No measured calibration
+is synthesized and neither path executes physics.
+
+RED-first coverage now includes exact-byte round-trip, native cancel and
+atomic rollback, corruption, duplicate keys, oversize input, unsupported
+producer/axis qualification, stale-provenance rebinding, transactional PyQt6
+population, browser rollback, URL cleanup, and pre-allocation size rejection.
+Focused evidence is 25 Python/PyQt tests and 12 React tests passing. With
+adjacent wire and manifest coverage, 44 Python and 16 React tests pass.
+Focused MyPy, Ruff, TypeScript, zero-warning ESLint, the 199-module production
+build, release-manifest validation, and module budgets are clean.
+
+Issue #4274 remains open. Measured calibration, workspace model-input
+persistence, execution/playback, result interchange, visualization, changing
+geometry, compiled regional physics, downstream parity, protected CI/review,
+integration of this child into the protected stack, and release remain explicit
+gaps.
+
 ## 2026-08-10 issue #4274 matched regional surface editor local child
 
 Local branch `feat/4274-regional-surface-ui` started from published PR #4335
@@ -544,751 +523,6 @@ All `115` ground tests, Ruff, formatting, a finding-free local scan of the
 affected file, documentation governance, the `370`-line source-size check,
 and diff gates pass before an ordinary guarded fast-forward publication.
 Fresh protected CI and review remain required.
-## 2026-08-11 Current ground parent propagated into PR #4288
-
-Open draft PR `#4288` retains branch `feat/4269-flight-ground-transfer` and base
-`feat/4268-ground-contract`. Exact remote transfer child
-`df716ac740beb9d47dced66f867dc3dde1fa5367` is merged first and exact current
-ground-contract parent `584e4e53a7990053a346938c4e45538786887ec7`
-second by an ordinary no-ff merge. Flight-to-ground production, tests, fixtures,
-cross-runtime contracts, Rust/PyO3/WASM integration, and explicit limitations
-remain intact. Only append-only handoff/SPEC aggregation conflicts require
-resolution; every dated record from both histories is retained. Focused gates
-pass 85 Python contract/transfer tests, 25 React tests, and 137 Rust library and
-transfer/wire tests. Pinned Ruff 0.14.10 covers all 13 PR-delta Python files,
-pinned MyPy 1.13 covers nine production files, and Bandit reports no medium/high
-findings. TypeScript, exact-delta ESLint, the 189-module web build, Rust format,
-docs, strict manifest, diff, exact transfer-tree, and both-parent preservation
-checks pass. Publication, protected exact-head CI, review, approval, and release
-remain open.
-
-## 2026-08-11 PR #4288 append-only history preservation repair
-
-This documentation-only follow-up to exact local candidate
-`1eb4188ff1a87a3d8e9a55aeca68194ca04e0bb4` restores the parent handoff
-sections and specification rows omitted during conflict resolution. It also
-repairs the interleaved rate-of-closure handoff paragraph without changing
-production code, tests, schemas, the campaign manifest, merge topology, or PR
-#4288's `feat/4268-ground-contract` base. Publication, protected exact-head CI,
-approval, dependency integration, and release remain open.
-
-## 2026-08-11 PR #4288 current-ground-parent ancestry candidate
-
-The clean `feat/4269-flight-ground-transfer` reconciliation starts from exact
-published child `afb1bdcfc6701caaf1f7bc3497a6a37dd9698c14` and normally merges
-exact newly published #4285 parent
-`e09ab96280cab363b36bd0e7db4a3cd064dc2527` second, while preserving #4288's
-`feat/4268-ground-contract` base. The child's bounded TypeScript integrator,
-signed terminal angular state, physical sphere/terrain contact, and cross-runtime
-transfer remain authoritative; the parent's split kinetics, plotting catalog,
-strict ground contract, and current wind/capability ancestry are retained.
-Seven surviving child-owned Python files are normalized with pinned Ruff
-0.14.10; this mechanical repair changes no AST. The TypeScript flight facade
-also preserves the exact child program with repository-standard line endings.
-Independent review, ordinary fast-forward publication, fresh protected CI,
-approval, dependency integration, and release remain open.
-
-> Update this append-only handoff in every implementation commit and every push
-> to `main`.
-> Last updated: 2026-08-11
-
-## 2026-08-11 Current capability carrier propagated into PR #4285
-
-- Open draft PR `#4285` retains branch `feat/4268-ground-contract` and base
-  `feat/4197-capability-observer`. Exact remote child
-  `e09ab96280cab363b36bd0e7db4a3cd064dc2527` is the first parent and exact
-  published capability-observer carrier
-  `d0b827667e61f0583e1dad0b3cbbca6819624d3c` is merged second by an ordinary
-  no-ff merge. No branch is rebased, retargeted, force-pushed, or rewritten.
-- Ground-contract production, tests, schemas, fixtures, dependencies, and
-  fail-closed limitations remain child-owned and intact. Only the four durable
-  handoff/SPEC files conflict; their dated records from both parents are
-  retained while the parent contributes the current wind/capability and release
-  ancestry.
-- Focused verification is green: 214 strict-ground, impact, PyQt layout/GUI,
-  kinetics, capability, wind, and manifest tests pass. Pinned Ruff 0.14.10
-  checks/formats all 27 PR-delta Python files, pinned MyPy 1.13 accepts all 20
-  production files, and Bandit reports no medium/high findings. Docs governance,
-  strict campaign-manifest validation, diff hygiene, exact child ground-tree
-  preservation, and both-parent handoff/SPEC preservation pass.
-- This is a bounded local audit. Publication, protected exact-head CI, review,
-  approval, and normal downstream propagation into #4288 remain open.
-
-## 2026-08-11 Current #4282 ancestry reconciled into PR #4285
-
-- Draft PR `#4285` retains branch `feat/4268-ground-contract` and base
-  `feat/4197-capability-observer`. Exact remote child
-  `7a38b9838d743e051a6900620c4d6e754582aa89` is the first parent; exact
-  published workflow/base head `b2baf6ef615b0d756a86ac4ca7eef2fc583210ee`
-  is merged second. The latter contains merged #4283 head
-  `9bbb98e16e435a0d4c74153b909f2ebfefbbce7a`, so dependency order is
-  preserved without rebase, retarget, force-push, or parent rewrite.
-- The strict flight-to-ground schemas, migrations, fixture, legacy adapter,
-  dependency declarations, and Python 3.10 compatibility remain intact. The
-  inherited parent adds its current wind/capability stack, Ground/Tee parity,
-  split kinetics facade, and pinned formatting repairs.
-- The only source conflict was the child's formatting-only pre-split kinetics
-  hunk; it resolves wholly to the validated parent facade. Four append-only
-  handoff/SPEC histories are retained and the campaign manifest records the
-  exact published heads that this unpublished candidate composes.
-- The child-owned scalar extractor's required `float` narrowing made the
-  459-line plotting registry part of the effective delta. Ordered catalog data
-  now lives in small private scalar/series collaborators; the public catalog is
-  61 lines and every changed production module/function/signature satisfies
-  the strict 400-line/50-line/four-parameter limits without changing keys,
-  order, extractors, units, or public objects.
-- Local verification passes the complete 1,539-test Rate/shared-swing Python
-  matrix with one explicit optional Rust-wheel skip, all 648 React tests across
-  106 files, TypeScript, zero-warning ESLint, the 188-module Vite production
-  build, and 12 `swing-core` Rust tests. Focused ground/parity/kinetics/wind/
-  manifest and catalog/plot suites pass 146 and 57 tests respectively.
-- This is a local publication candidate only. Independent review, ordinary
-  publication, fresh exact-head protected CI, required approval, and normal
-  downstream #4288 propagation remain mandatory.
-
-## 2026-08-11 merged #4283 carrier receives current #4282
-
-- PR `#4283` is already merged, but its preserved head branch
-  `feat/4197-capability-observer` remains the live base of open draft PR `#4285`.
-  Exact remote/local carrier child
-  `b2baf6ef615b0d756a86ac4ca7eef2fc583210ee` is merged first with exact newly
-  published #4282 head `3faf9e9493c9ba37168d9f65ab10cafeedc2a72f`
-  second by a normal no-ff merge. No PR is reopened, retargeted, or merged, and
-  no branch history is rewritten.
-- Because the carrier child is already the first parent of #4282, the source
-  merge is conflict-free and reproduces the exact #4282 implementation while
-  preserving the complete capability-observer history. This carrier record is
-  the only intentional tree delta from #4282.
-- Focused validation is green: 86 Python/PyQt capability, wind, and campaign-
-  manifest tests pass, as do all 86 tests in 12 focused React capability/wind
-  suites. React type-check, zero-warning lint, and the production build pass.
-  Docs governance, strict campaign-manifest validation, and exact-parent diff
-  hygiene pass. The exact tree delta from #4282 contains documentation only, so
-  changed-Python Ruff/MyPy/Bandit gates are not applicable.
-- Publication is explicitly outside this bounded audit;
-  protected current-head checks and ordinary dependency propagation into #4285
-  remain separate future actions.
-
-## 2026-08-11 current scalar-adapter propagation into wind workflow PR #4282
-
-- PR `#4282` remains on `feat/4199-wind-workflow`, based on
-  `feat/4199-wind-scalar-adapter`. Exact published child
-  `b2baf6ef615b0d756a86ac4ca7eef2fc583210ee` is merged first with exact newly
-  published parent `9321c1d2e091b8c7e5a4a83aa9ad726290e7fb5a` second by a
-  normal merge; no rebase, retarget, force-push, or parent rewrite is used.
-- All implementation and test paths merge automatically. The child responsive
-  PyQt6/React wind workflow, execution/progress/cancellation, persistence,
-  export, consolidated capability stack, and strict campaign manifest remain
-  authoritative while inheriting the parent's current wind scalar-adapter,
-  variation, workspace, launch-monitor/D-plane, split-kinetics, and Qt repairs.
-  Only append-only handoffs and SPEC conflict textually; both histories remain.
-- Focused pre-commit regression is green: all 94 focused Python/PyQt wind,
-  responsive-layout, navigation, playback, and manifest tests pass, and all 44
-  tests in eight focused React wind/responsive suites pass. Docs governance and
-  strict campaign-manifest generation/validation also pass.
-- Exact post-commit PR-delta gates are also green: pinned Ruff 0.14.10 checks
-  and formats all 57 changed Python files, pinned MyPy 1.13.0 accepts all 38
-  changed production files, and pinned Bandit 1.7.7 finds no medium/high issues
-  in 37 scanned source files. The 500-LOC budget, changed-Python policy,
-  minimum-test contract, changed-test assertions, diff hygiene, React
-  type-check, React lint, and production web build all pass.
-- This candidate remains local pending independent review, ordinary publication,
-  protected exact-head CI, unresolved-thread checks, and dependency order.
-
-## 2026-08-11 Exact #4281 parent propagation into PR #4282
-
-- Draft PR `#4282` remains on `feat/4199-wind-workflow`, based on
-  `feat/4199-wind-scalar-adapter`. Exact remote child
-  `29e15d6ff631f7f30afcc745be783f2e716d7dcf` normally merges exact published
-  parent `a7dce5f89b483303938f518b74b4028a1c68ba81` in child-first order without
-  rebase, reset, retarget, force-push, or parent rewrite.
-- The responsive PyQt6/React wind workflow and consolidated capability stack
-  remain intact while inheriting the parent's split kinetics facade, Ground/Tee
-  parity, scalar-adapter module/parameter repair, and complete prior stack.
-- The sole source conflict is the child's obsolete formatting-only edit to the
-  pre-split kinetics monolith; the validated parent facade is authoritative.
-  Four documentation conflicts retain both append-only histories. Seven
-  child-owned files require the pinned Ruff formatter before publication.
-- The exact skipped-import MyPy profile makes the facade cache suppression
-  redundant; an explicit `KineticsSeries | None` cast preserves the runtime
-  object while satisfying the static boundary without an ignore.
-- Local evidence is 74 focused Python/PyQt wind, kinetics, and Ground/Tee
-  tests; 29 focused React tests; and 8 campaign-manifest tests. TypeScript,
-  focused zero-warning ESLint, Ruff check/format, pinned MyPy 1.13, Bandit's
-  medium/high threshold, strict wind module/signature/function budgets,
-  manifest, documentation, policy, SPEC, and diff gates pass.
-- This is a local reconciliation candidate. Independent review, fresh
-  exact-head protected CI, unresolved-thread checks, dependency order, and
-  ordinary publication remain required.
-## 2026-08-11 current variation propagation into wind PR #4281
-
-- PR `#4281` remains on `feat/4199-wind-scalar-adapter`, based on
-  `feat/4144-variation-export-continuation`. Exact published child
-  `a7dce5f89b483303938f518b74b4028a1c68ba81` is merged first with exact newly
-  published parent `e6c7460a01082631565fb9ed48aa32538bd7772c` second by a
-  normal merge; no rebase, retarget, force-push, or parent rewrite is used.
-- All implementation and test paths merge automatically. The child
-  Python/TypeScript scalar-ensemble, deterministic-scenario, wind-strategy
-  adapter, and strict module-budget behavior remain authoritative while
-  inheriting the parent's current variation, workspace, launch-monitor/D-plane,
-  split-kinetics, and Qt typing repairs. Only append-only handoffs and SPEC
-  conflict textually; both histories remain below.
-- Focused pre-commit regression is green: all 77 focused wind/variation Python
-  tests pass, and all 20 tests in the five focused React wind/variation suites
-  pass.
-- Exact post-commit PR-delta gates are also green: pinned Ruff 0.14.10 checks
-  and formats all five changed Python files, pinned MyPy 1.13.0 accepts all
-  three changed production files, and pinned Bandit 1.7.7 finds no issues in
-  those three files. Both 500- and strict 400-LOC budgets, docs governance,
-  changed-Python policy, minimum-test contract, changed-test assertions, diff
-  hygiene, React type-check, React lint, and the production web build all pass.
-  The manifest artifact/checker/test do not exist on either side of this
-  pre-manifest stack, so that later release gate is structurally not applicable.
-- This candidate remains local pending independent review, ordinary
-  publication, protected
-  exact-head CI, unresolved-thread checks, and dependency order.
-## 2026-08-11 PR #4281 wind adapter module-budget repair
-
-- The 403-line `wind_strategy_plot_adapter.py` blocker is repaired by moving
-  request/analysis validation into the private
-  `_wind_strategy_plot_validation.py` collaborator. The public adapter is now
-  252 lines and the validator is 213 lines; both are below the mandatory
-  400-line ceiling.
-- The scenario validator accepts the scenario plus one frozen, slotted
-  expectation object. No production function exceeds four parameters or 50
-  lines.
-- `build_wind_strategy_plot_dataset`, `WIND_STRATEGY_*` exports, validation
-  messages, variable/stage/category/cohort order, row identity, scalar values,
-  availability, attributes, and the no-flight-rerun boundary are unchanged.
-- This is a local candidate only. Independent review, fresh exact-head
-  protected CI, unresolved-thread checks, dependency order, and ordinary
-  publication remain required before merge or release.
-
-## 2026-08-11 reviewed variation parent propagation into wind PR #4281
-
-- Exact remote wind child `247046d55afcad3e6cd4f8029f854856c427f59c`
-  normally merges exact reviewed variation parent
-  `3337945699966b63cb5cd8e52d7c3b194315e911` in child-first order. PR base,
-  stack topology, and both histories remain intact; no rebase, reset, retarget,
-  force-push, or parent rewrite is used.
-- Python/TypeScript scalar-ensemble and wind-strategy plot-adapter behavior is
-  unchanged while inheriting the parent's variation/export UI, kinetics split,
-  Ground/Tee parity, workspace/toolstrip, and Ruff-normalized repairs. The sole
-  feature-code conflict was obsolete monolithic kinetics and is resolved to the
-  validated parent `pendulum.sample(...)` facade.
-- This reconciliation is local only. Independent review, fresh exact-head CI,
-  unresolved-thread checks, dependency order, and ordinary publication remain
-  required before merge or release.
-## 2026-08-11 current workspace propagation into variation PR #4280
-
-- PR `#4280` remains on `feat/4144-variation-export-continuation`, based on
-  `feat/4218-toolstrip-workspace`. Exact published child
-  `3337945699966b63cb5cd8e52d7c3b194315e911` is merged first with exact newly
-  published parent `efbca84095b617b4018732f7802c2da3f0525387` second by a
-  normal merge; no rebase, retarget, force-push, or parent rewrite is used.
-- All implementation and test paths merge automatically. The child variation
-  export, typed unavailable outcomes, bounded tables, linked selection, and
-  all-trial arc behavior remain authoritative while inheriting the parent's
-  current workspace, launch-monitor/D-plane ancestry, split kinetics, and Qt
-  typing repairs. Only append-only handoffs and SPEC conflict textually; both
-  histories remain below.
-- Focused pre-commit regression is green: all 56 focused Python variation tests
-  pass, and all 10 tests in the three focused React variation suites pass.
-- Exact post-commit PR-delta gates against the new parent are also green:
-  pinned Ruff 0.14.10 checks and formats all five changed Python files, pinned
-  MyPy 1.13.0 accepts all four changed production files, and pinned Bandit
-  1.7.7 finds no issues in those four files. Docs governance, 500-LOC budget,
-  changed-Python policy, minimum-test contract, changed-test assertions, diff
-  hygiene, React type-check, React lint, and the production web build all pass.
-  The manifest artifact/checker/test do not exist on either side of this
-  pre-manifest stack, so that later release gate is structurally not applicable.
-- This candidate remains local pending independent review, ordinary
-  publication, protected
-  exact-head CI, unresolved-thread checks, and dependency order.
-## 2026-08-11 reviewed workspace parent propagation into variation PR #4280
-
-- Exact remote variation child
-  `668ba96746f79f7a12e8092161bd610054197f58` normally merges exact reviewed
-  workspace/toolstrip parent `ccd0e026c580c93038fdf5c59d5d452a85ba27a0`
-  in child-first order. PR base, stack order, and both histories remain intact;
-  no rebase, reset, retarget, force-push, or parent rewrite is used.
-- The variation/export feature remains unchanged while inheriting the current
-  kinetics split, Ground/Tee parity contracts, protected Ruff normalization,
-  and complete workspace/toolstrip behavior. The sole feature-code conflict
-  was obsolete monolithic kinetics source and is resolved to the validated
-  parent façade. Seven duplicate child automation edits fail protected Ruff
-  0.14.10 formatting and are normalized to the exact reviewed parent blobs;
-  the automation commit remains reachable in the child-first history.
-- This reconciliation is local only. Independent review, fresh exact-head
-  protected CI, unresolved-thread checks, dependency order, and ordinary
-  publication remain required before merge or release.
-## 2026-08-11 current #4203 propagation into workspace PR #4279
-
-- Draft PR `#4279` retains branch `feat/4218-toolstrip-workspace` and base
-  `feat/4181-launch-monitor-registry`. Exact published child
-  `ccd0e026c580c93038fdf5c59d5d452a85ba27a0` is merged first with exact
-  newly published parent `7abce9ad767fe8311da66a1e5998b892ea3ca9de`
-  second by a normal merge; no rebase, retarget, force-push, or parent rewrite
-  is used.
-- All implementation paths merge automatically. The child workspace,
-  toolstrip, visibility, navigation, playback, and independent-plot behavior
-  remain authoritative while inheriting the parent's split kinetics and four
-  behavior-preserving Qt primitive-return boundaries. Only append-only
-  handoffs and SPEC require textual reconciliation, and both histories remain
-  below.
-- Focused pre-commit regression is green: all 142 exact PR-delta Python tests
-  pass, and all 32 tests in the eight changed React suites pass. Exact
-  post-commit PR-delta gates against the new parent are also green: pinned Ruff
-  0.14.10 checks and formats all 27 changed Python files, pinned MyPy 1.13.0
-  accepts all 18 changed production files, and pinned Bandit 1.7.7 finds no
-  medium/high issues in those 18 files. Docs governance, 500-LOC budget,
-  changed-Python policy, minimum-test contract, changed-test assertions, diff
-  hygiene, React type-check, React lint, and the production web build all pass.
-  The manifest artifact/checker/test do not exist on either side of this
-  pre-manifest stack, so that later release gate is structurally not applicable
-  to this propagation.
-- This candidate remains local pending independent review, ordinary
-  publication, protected
-  exact-head CI, unresolved-thread checks, and dependency order.
-## 2026-08-11 remote automation reconciliation for workspace PR #4279
-
-- Exact local workspace head `0b22c401a26c31441a599d8d9b39de123706e7ea`
-  ordinarily merges exact remote head
-  `61fe2d556a5413e525d958612ccfd57e65b8d5a2`, preserving both histories and
-  the existing PR base `feat/4181-launch-monitor-registry` without rebase,
-  reset, retarget, force-push, or parent rewrite.
-- The remote commit is a broad formatting-only automation sweep. Fifteen of
-  its 23 paths, including six pre-existing `.codex-worktrees` gitlinks, were
-  already byte-identical in the current parent. Its seven unique formatting
-  edits did not match protected Ruff 0.14.10 output and are normalized back to
-  the pinned form. Its sole content conflict was obsolete pre-split kinetics
-  code; the current `pendulum.sample(...)` façade implementation remains
-  authoritative, preserving the parent split and runtime behavior.
-- Workspace/toolstrip, module visibility, navigation, playback, independent
-  plots, physics, frames, units, schemas, and public contracts remain intact.
-  This history-preserving local reconciliation still requires independent
-  review, exact-head protected CI, unresolved-thread checks, and dependency
-  gates before publication or merge.
-
-## 2026-08-11 hosted MyPy repair propagation into workspace PR #4279
-
-- Exact workspace child `7806a16f58e1c6999d32f0127a187fbb21f839a1`
-  normally merges exact published parent
-  `3796b49e40b677fbac4e05739f8be49f905df2cb`; PR base, stack order, and
-  both histories remain unchanged.
-- The inherited production delta is limited to four static
-  `numpy.ndarray` casts in the kinetics façade, series, and dynamics modules.
-  Workspace/toolstrip behavior and runtime arrays, physics, units, frames,
-  public identity, and UI behavior are unchanged.
-- This merge is local only. Fresh exact-head protected CI, required review,
-  unresolved-thread checks, and dependency gates remain release blockers.
-
-## 2026-08-11 latest #4203 propagation into workspace PR #4279
-
-- Exact parent `0216a547aa79727091a2939b96e779e8ddbd7304` is normally merged into
-  child `61b7f48b5aeb7d57246b4963da3df086e79cbe15` without changing PR base,
-  stack order, or either history.
-- No feature-code conflict exists. The workspace/toolstrip, visibility,
-  navigation, playback, and plot controls remain intact while the child
-  inherits the parent's pinned formatting and identity-preserving kinetics
-  size-budget repair.
-- This merge is local only. Fresh exact-head protected CI, review, unresolved
-  threads, and dependency gates remain required before publication or merge.
-
-## 2026-08-11 #4203 append-only SPEC preservation repair
-
-The first independent audit of local reconciliation candidate `e20b4f630...`
-found four exact D-plane parent rows omitted from the append-only SPEC history:
-2026-08-10 versions 1.13.11, 1.13.9, 1.13.7, and 1.13.6. They are restored
-verbatim in the current documentation-only follow-up. Production code, tests,
-merge parents, PR base, and local quality evidence are unchanged. Independent
-re-review, ordinary publication, protected CI, and downstream propagation
-remain open.
-
-## 2026-08-11 #4203 current D-plane parent reconciliation
-
-- Draft PR #4203 now requires a normal merge of exact current base
-  `f3363aa88868f6a5c7e9ccfc682a9eca014e86c1` after exact published child
-  `217e36dc93d30f79826847f958fbcd10805e58ed`; its base remains
-  `feat/4189-dplane` and no history is rewritten.
-- The parent changes nine files. Its sole textual conflict is an inherited
-  kinetics formatting edit. The already reviewed split kinetics facade stays
-  authoritative and uses the typed `DoublePendulumSwing` object directly;
-  the parent's explanatory geometry comment is retained.
-- Exact CI-pinned MyPy 1.13 validation then exposed four remaining Qt stub
-  boundaries in the child delta. Responsive-event handling, legend visibility,
-  ball-setup event filtering, and status text now narrow their unchanged Qt
-  return values to the declared primitive contracts.
-- Focused regression, full PR-delta quality gates, independent review, normal
-  fast-forward publication, and fresh protected exact-head CI remain required.
-
-## 2026-08-11 #4203 exact-head format completion
-
-- Protected CI on exact published head
-  `7d69a545ae555679f0318940e67c1786626d6794` failed only Ruff formatting.
-  Reproducing the exact pinned 0.14.10 check found eleven noncompliant changed
-  Python files: four inherited files plus seven altered by the automated
-  pre-commit repair.
-- The pending repair applies pinned Ruff formatting to exactly those eleven
-  files. AST equivalence and focused tests must pass before independent review
-  and ordinary fast-forward publication; no simulation, contract, or UI
-  behavior is intentionally changed.
-
-## 2026-08-11 #4203 hosted MyPy kinetics repair
-
-- Exact head `0216a547aa79727091a2939b96e779e8ddbd7304` failed CI Standard run
-  `31477542889`, job `93734652129`, at runtime merge ref `aede309`: NumPy's
-  typed API exposed four `no-any-return` findings in the newly extracted
-  kinetics modules.
-- The repair narrows only the results of `numpy.linalg.norm`,
-  `numpy.concatenate`, and the app-frame matrix projection to the already
-  declared `numpy.ndarray` return contract. Explicit casts do not allocate,
-  convert, or change any array, physics, units, frames, public identity, or UI.
-- RED evidence is the exact four hosted diagnostics at
-  `_kinetics_series.py:121/131`, `_kinetics_dynamics.py:194`, and
-  `kinetics.py:61`. GREEN evidence requires the complete PR-base changed-source
-  MyPy profile, not a three-file-only run, plus focused/full Rate regression and
-  normal protected CI after publication.
-- Local GREEN evidence is 102/102 complete-delta MyPy source files, 141/141
-  Ruff/format files, 101 Bandit source files with no medium/high finding, 28/28
-  focused tests, and 701/701 full Rate tests. Size, documentation, minimum-test,
-  and diff gates also pass.
-- This commit is local only. PR base, stack order, protected review, and the
-  paused #4279 parent propagation remain unchanged.
-
-## 2026-08-11 #4203 kinetics size-budget repair
-
-- Exact head `572bf525dd1ded26cbc3fbb4f228d1f6ca16e118` passes the PR-base
-  changed-file scan because `kinetics.py` is byte-identical to the inherited
-  parent, but the stricter `HEAD~1` scan selects the Ruff-formatted file and
-  fails it at 646 LOC against the ungrandfathered 500-LOC limit.
-- The behavior-preserving LoD/DRY split leaves
-  `simulation/kinetics.py` as the stable 222-LOC public façade, moves pure
-  dynamics to `_kinetics_dynamics.py` (205 LOC), and moves the immutable
-  series/DbC contract to `_kinetics_series.py` (131 LOC). Public constants,
-  class/function objects, and the private `_reaction_forces` compatibility
-  seam are identity-pinned by a RED-first contract test.
-- Physics, units, frames, numerical parity fixture, PyQt presentation, and
-  release-stack order are unchanged. This implementation is committed locally
-  only; no push, PR mutation, or release claim is authorized by this handoff.
-- Verification includes 28 focused kinetics/presentation/PyQt tests and the
-  complete 701-test Rate-of-Closure Python suite, all passing locally.
-
-## 2026-08-11 #4203 pinned-Ruff formatting repair
-
-- No material handoff change: this commit only applies the repository-pinned
-  Ruff 0.14.10 formatter to the eight Python files named by current-head CI
-  Standard run `31468208320`, job `93705508050`.
-- Physics, application behavior, public contracts, schemas, UI layout, stack
-  bases, and dependency order are unchanged. Fresh protected CI and review at
-  the resulting head remain required; queued runner jobs are not green
-  evidence.
-
-## 2026-08-10 #4143 child receives repaired #4203 parent
-
-- Ready PR `#4325` retains branch `feat/4143-tee-parity-fixture` and base
-  `feat/4181-launch-monitor-registry`; repaired parent head
-  `12dd76a8dbcc106c4683f2f2e53076f8dc6f1b76` is incorporated through a
-  normal merge commit without rebasing, retargeting, or rewriting history.
-- The merge tree has no production/test-code conflict. Both branches'
-  append-only handoff/SPEC evidence is retained, including the shared parity
-  fixture and deterministic web/PyQt visual evidence.
-- Fresh exact-head protected CI and review remain required. #4143 stays open
-  until the dependency line lands on `main`.
-
-## 2026-08-10 #4143 Python/React Golden Ball-Setup Parity
-
-- Branch `feat/4143-tee-parity-fixture` starts from exact draft PR #4203 head
-  `31cbc007d4c85b5479b7cd0fb0969124eab2af67`; it does not rewrite or retarget
-  the release stack.
-- `ball_setup_golden_v1.json` is the single Python/React source of truth for
-  schema `rate_of_closure.ball_setup_golden` version 1, SI metre units, and
-  reference `ground_plane_to_ball_bottom`. It pins Driver/Tee and iron/Ground
-  defaults, explicit overrides, Ground's zero effective tee height, derived
-  center geometry, serialization, negative and non-finite rejection, and
-  legacy-run migration to Ground.
-- New consumer tests plus the existing tee suites pass: 18 Python tests and 24
-  React tests. Web TypeScript, ESLint, and production Vite build gates pass;
-  scoped Ruff check and format pass. This is test/fixture-only and does not
-  change production physics or UI behavior.
-- Deterministic 1600 x 1200 Playwright captures cover the default Driver/Tee
-  and rerun explicit-Ground React states with checked/disabled controls,
-  present/absent tee geometry, and zero console/page errors. A hidden 1400 x
-  900 PyQt harness records the same states with canonical center and tee-artist
-  assertions; a headless regression keeps its temporary PNGs nonblank and
-  structurally distinct without pixel-perfect baselines. Artifact manifests
-  and SHA-256 digests are under
-  `C:\Users\diete\AppData\Local\Temp\rate-4143-visual-evidence-8050eeba`.
-- Issue #4143 remains open for exact-head protected CI/review and release to
-  `main`. The strict campaign release manifest is not present on #4203's exact
-  history (it was introduced later on a divergent campaign branch), so this
-  bounded child records that limitation instead of reconstructing it.
-## 2026-08-10 Repaired wind-scalar parent propagation
-
-## 2026-08-10 PR #4282 exact-head mypy repair
-
-Protected quality-gate run `31395741841`, job `93478091420`, checked merge ref
-`ae75c4b006072b20a713e45b177defcc16a380fe` for exact head
-`aa6eeffb0395f7ed7954f2315b1c625cada552d8`. Checkout, dependency install,
-Ruff, and formatting passed; pinned mypy 1.13 then failed only
-`scripts/rate_campaign_manifest.py:335:5 [no-any-return]`. Under the exact CI
-`--follow-imports=skip` profile, Pydantic's `model_validate_json` return is
-`Any`. The validated local now carries an explicit `CampaignManifest` type,
-preserving runtime behavior without a redundant cast or suppression. Do not
-retry or cite the obsolete failed head as passing.
-
-## 2026-08-10 Exact workspace-parent propagation into PR #4280
-
-Draft PR `#4280` remains on `feat/4144-variation-export-continuation` with
-base `feat/4218-toolstrip-workspace`. The normal merge starts from original
-child head `f90836e342efc8be624739802375af2876d11e5f` and incorporates exact
-published parent head `6717e9e09d507dbc24bedb36177f1cdf0b4fd90b` as
-the second parent. No rebase, retarget, force-push, parent rewrite, or draft-
-state change is permitted.
-
-The source merge is conflict-free: the child's selected-scatter CSV export,
-typed unavailable values, accessible raw tables, linked selection, and focused
-PyQt/React visualization modules remain intact alongside the parent's complete
-workspace/toolstrip, playback, plot, navigation, Python 3.10, and module-budget
-repairs. SPEC 1.14.12 records this combined child release above the unique
-parent 1.14.11 entry. Staged review found and the continuation corrected one
-release-blocking rerun defect: replacing a study after selecting a later trial
-now clears linked selection atomically. Every PyQt public setter validates
-against the current trial count, while React clears on result identity change
-and shares only bounded selections, so a smaller rerun cannot crash or leave
-all new points dimmed.
-
-Verification is green: 37 focused variation GUI/export tests and the complete
-Rate/shared-swing/golf-club matrix passes 1,528 tests with two explicit
-optional build123d skips; all
-546 React tests pass across 90 files, as do TypeScript, ESLint, and the Vite
-production build. Exact-parent Ruff, Ruff-format, Black, pinned MyPy 1.13 for
-all four changed Python production modules, Bandit, the four-file 500-line gate,
-docs, minimum-test, changed-test assertions, detect-secrets fingerprints, and
-diff checks pass. CPython 3.10.20 compiles all changed Python files, and 30
-compatibility/date-boundary regressions pass. There is no Rust delta from the
-exact parent, whose 12-test `swing-core` evidence remains applicable.
-Independent staged re-review found no actionable findings after verifying the
-atomic result-identity boundary and validated-value widget updates. Protected
-exact-head CI and required repository review remain release gates after
-guarded publication.
-
-## 2026-08-10 Second workspace-parent propagation into PR #4280
-
-## 2026-08-10 Exact wind-adapter propagation into PR #4282
-
-Draft PR `#4282` remains on `feat/4199-wind-workflow` with base
-`feat/4199-wind-scalar-adapter`. The normal merge starts from original child
-`5f77af4add23547a21cc3fabce98ae9ad4260427` and incorporates exact published
-parent `4a793c4c3f19aad43a3c215800b266be487ace49` second. Do not rebase,
-retarget, force-push, rewrite either parent, change draft state, or merge the PR.
-
-The consolidated responsive wind/capability workflow and campaign release
-authority remain additive with the parent's variation visualizations,
-result-local selection, workspace/toolstrip, Python 3.10 compatibility, and
-matching `scalar-ensemble/v1` wind adapter. The canonical navigation-state
-module now owns the child capability-tab ID, preserving the parent's safe
-legacy visibility migration without duplicating navigation constants. The
-parent's strict deterministic wind boundary also remains intact: exact
-per-trial provenance, zero shear/turbulence, seed zero, and no gusts.
-Static review replaces child-owned optimized-away runtime assertions with
-explicit invariant failures and identifies the fixed-argument executable
-builder subprocess boundary with Bandit's native annotations. The exact-head
-CI repair above handles the separate skipped-import Pydantic return boundary.
-
-SPEC 1.14.14 is the unique combined child entry above parent 1.14.13.
-Verification is green across 73 focused Python navigation/wind/compatibility
-tests, the complete Rate/shared-swing/golf-club matrix (`1,657 passed`, two
-explicit optional build123d skips), and all 643 React tests across 105 files.
-TypeScript, zero-warning ESLint, Vite production build, 12 Rust tests, 31 real
-CPython 3.10 compatibility tests, and 43 post-review invariant/GUI/manifest
-tests pass. Ruff/format, pinned MyPy 1.13 across 37 changed production files,
-Bandit, the exact-parent 500-line budget, campaign-manifest contracts,
-docs/minimum-test/assertion governance, CI-policy secrets scan, compilation,
-and diff gates pass. Independent re-review found no actionable findings after
-94 Python workflow/navigation tests, 20 manifest/GUI tests, and 61 React
-workflow tests plus independent static checks. Protected current-head CI,
-unresolved threads, and required repository review remain release gates.
-
-## Active Rate of Closure campaign
-
-The active checkout is
-`C:\Users\diete\Repositories\Tools-worktrees\toolstrip-workspace` on the
-existing PR #4282 carrier `feat/4199-wind-workflow`. The local continuation
-starts from original published head
-`5f77af4add23547a21cc3fabce98ae9ad4260427` and normally incorporates exact
-published #4281 parent `4a793c4c3f19aad43a3c215800b266be487ace49`
-without changing the branch base `feat/4199-wind-scalar-adapter`. It composes
-four reviewed slices:
-strict capability parsing and signed decimal entry; complete capability
-diagnostics/result exports and quantitative React scatter scales; package-safe
-static-web release entrypoints; and the strict `rate-of-closure-campaign/v1`
-release-evidence authority. The normal merge also carries the parent's
-Python 3.10 compatibility, variation-export, and scalar-ensemble corrections.
-No branch was rebased, retargeted, rewritten, or force-pushed. Protected CI
-and review remain due after normal publication.
-
-Canonical files:
-
-- `docs/release/rate_of_closure_campaign.v1.json` — normalized campaign state;
-- `scripts/rate_campaign_manifest.py` — schema and contradiction validator;
-- `docs/release/RATE_OF_CLOSURE_CAMPAIGN_MANIFEST.md` — maintenance procedure;
-- `src/rate_of_closure/AGENT_HANDOFF.md` — current Rate-specific handoff;
-- `docs/development/RATE_OF_CLOSURE_CAMPAIGN_HANDOFF.md` — historical detail.
-
-The manifest deliberately reports the campaign as **not released**. It
-distinguishes specified scope, feature-stack implementation, protected parent
-merge, and exact `main` release. A local pass, a feature-branch merge, and a
-default-branch release are never interchangeable.
-
-## Current carrier and release state
-
-Capability PRs #4294, #4289, and #4283 were merged top-down into
-`feat/4199-wind-workflow`; their feature parents were not protected. PR #4282
-is the current open carrier on base `feat/4199-wind-scalar-adapter`. Exact
-published parent `4a793c4c3f19aad43a3c215800b266be487ace49` is incorporated
-through the normal local merge containing this handoff.
-
-The previous exact #4282 head `3186a265b1` built and loaded `swing_core`, but
-its Python 3.11 parity job failed before collection because a cached
-`pytest-qt` plugin loaded without PyQt6. Commit `18fe89201` disables third-party
-pytest plugin auto-loading only in that Rust parity step. Its focused workflow
-contract test passes; a fresh hosted parity run is still required.
-
-Combined local evidence on the composed continuation is 1,657 Python tests
-with two explicit optional skips and 105 React files / 643 tests, plus the
-TypeScript, zero-warning ESLint, 188-module Vite build, Rust, Python 3.10,
-manifest, Ruff, pinned mypy, Bandit, governance, secrets, compilation, and
-diff evidence recorded above.
-
-Hosted quality-gate run `31340032608` reached mypy 1.13 and exposed
-CI-context-only `no-any-return` findings in the Pydantic manifest loader and
-Qt elapsed-timer adapter because the delta lane uses `--follow-imports=skip`.
-Both boundaries now narrow their return values explicitly. The exact Python
-3.12/mypy 1.13 delta is clean across 54 files; Ruff passes and 62 focused
-regression tests plus eight campaign-manifest tests pass (only pre-existing
-optional-plugin config warnings).
-
-Corrected-parent propagation evidence is 62 focused wind/scalar/variation and
-compatibility tests on both Python 3.11 and real CPython 3.10.20, plus 8 React
-files / 35 tests, TypeScript, and focused zero-warning ESLint. The Python 3.10
-run exposed one child-owned direct `enum.StrEnum` import in capability
-observations; it now uses the shared runtime compatibility contract and is
-included in the source-level regression. Ruff check/format passes 15 focused
-files, pinned mypy 1.13 passes 10 production modules, and all nine campaign
-manifest/parity contracts pass.
-
-The direct web launcher dynamically loads the root bootstrap through
-`importlib` instead of mutating `sys.path` in the changed entrypoint. Its real
-child-process delegation test and the changed-Python policy guard cover that
-release path.
-
-Only `main` is the release boundary. It requires `quality-gate` and
-`tests (3.11)`. Outer PR #4119 remains the main-targeting platform carrier and
-requires current-main reconciliation. Impact-interval PR #4133 merged after
-its parent stack had already propagated and is not proven in #4119.
-
-## Critical open programs
-
-- #4142/#4144: variation graphics are substantial, but global sensitivity,
-  bounded performance, downstream pinning, and release remain.
-- #4146: club-builder core exists; manufacturing export, image fitting, full
-  client workflow, and downstream qualification remain.
-- #4158: wedge kinematics, clearance, turf, visualization, and forgiveness are
-  feature-stack implementations without a protected release.
-- #4191/#4201: wind, inverse flight, playback, and capability work exist on the
-  stack; scientific, installed-package, and cross-surface release gates remain.
-- #4218/#4234: toolstrip and selected layout fixes exist; camera issue #4284,
-  complete persistence, high-DPI, keyboard, and visual baselines remain.
-- #4260/#4267: four-surface parity is specified; ground contracts/transfer are
-  partial and bounce, skid, roll, total distance, UI, and parity remain open.
-
-## Required validation
-
-```powershell
-python scripts/rate_campaign_manifest.py
-python -m pytest tests/rate_of_closure/test_campaign_release_manifest.py -q
-python -m pytest tests/ops/test_maturin_swing_core_workflow.py -q
-python -m ruff check scripts/rate_campaign_manifest.py `
-  tests/rate_of_closure/test_campaign_release_manifest.py
-python -m ruff format --check scripts/rate_campaign_manifest.py `
-  tests/rate_of_closure/test_campaign_release_manifest.py
-python -m mypy --ignore-missing-imports scripts/rate_campaign_manifest.py
-```
-
-For a feature change, also run the complete affected Rate Python/PyQt/shared
-swing and React suites recorded in the per-tool handoff.
-
-## Do not
-
-- Do not mark a program released without a 40-character `main` merge SHA.
-- Do not convert queued, cancelled, skipped, or failed checks into passing
-  evidence.
-- Do not retarget or rewrite the protected stack, force-push, or admin-merge.
-- Do not duplicate carrier or test metadata in individual program records.
-- Do not treat an UpstreamDrift launcher tile as four-surface parity.
-
-Fleet handoff policy is tracked by Repository_Management #1393/#1397. Any
-`src/**` handoff change also requires the same-commit `SPEC.md` update.
-
-## 2026-08-10 Final repaired #4282 propagation into PR #4285
-
-Draft PR `#4285` retains branch `feat/4268-ground-contract` and base
-`feat/4197-capability-observer`. Original child head
-`e5bcbd1096d3be1f621a805c9d9f3fd321e375a5` now normally incorporates exact
-quality-green workflow head `1e82f15026786ea0b08f78f4c001590ddce9ff39`
-second. No ground-contract production or test code conflicted, and no rebase,
-retarget, force-push, parent rewrite, or draft-state change was used. Fresh
-protected CI/review is required before downstream #4288 propagation.
-
-## 2026-08-10 Final repaired ground-contract propagation into PR #4288
-
-Draft PR `#4288` retains branch `feat/4269-flight-ground-transfer` and base
-`feat/4268-ground-contract`. Original child head
-`108a841b1378c992defd3c7b7ee263d41a6c8b24` normally incorporates exact
-quality-green parent `a93edb4bfd6a8dc9334122cd2ae660983d5bf424`
-second. No transfer production or test code conflicted. No rebase, retarget,
-force-push, parent rewrite, or draft-state change was used; fresh protected CI
-and review remain required.
-
-## PR #4288 exact repaired-ground propagation
-
-Draft PR #4288 remains on `feat/4269-flight-ground-transfer` with unchanged
-base `feat/4268-ground-contract`. Original child
-`247215422a6d4b677552955b4923bc609a553259` normally incorporates exact repaired
-#4285 parent `e5bcbd1096d3be1f621a805c9d9f3fd321e375a5` second in the
-merge containing this handoff. The child preserves its signed terminal state,
-physical sphere/terrain contact brackets, strict provenance, and qualified
-Python/TypeScript/Rust/PyO3/WASM transfer while inheriting deterministic
-workspace timestamps, canonical `swing_sim` import identity, the hosted-mypy
-manifest repair, and the complete variation, wind, capability, and campaign
-ancestry. No branch was rebased, retargeted, rewritten, or force-pushed.
-
-This propagation adds no bounce, skid, roll, terrain response, total distance,
-or UI execution. Protected CI, independent review, and dependency-ordered
-collapse remain separate release gates.
-
-Exact composed-tree verification is 1,080 Python tests passed with six explicit
-optional installed-`tools_core` wheel skips, 107 React files / 662 tests,
-26 direct Rust transfer/wire tests, TypeScript, zero-warning ESLint, and the
-189-module production build. Ruff check/format passes all 28 changed Python
-files; pinned mypy 1.13 passes all 21 changed production files. Campaign
-manifest validation, documentation governance, and diff checks are clean. The
-missing local wheel is not accelerated installed-package evidence and remains
-an explicit release boundary; the direct Rust suites executed rather than
-being relabeled as wheel parity.
-
-## 2026-08-10 Exact repaired #4282 propagation into PR #4285
-
-Draft PR #4285 remains on `feat/4268-ground-contract` with unchanged base
-`feat/4197-capability-observer`. The normal merge containing this handoff keeps
-original child `788aa547651a3685a363ea401824a5d81477bafb` first and incorporates
-exact repaired #4282 carrier `686016196a2f895058b8a566dff103a0fd32cd10`
-second. That carrier contains merged capability PR #4283 commit
-`c1827bbdc50a6e11cc475db2636b4e47a4c15416`, exact observer head
-`9bbb98e16e435a0d4c74153b909f2ebfefbbce7a`, and the hosted-mypy manifest
-repair following predecessor `aa6eeffb0395f7ed7954f2315b1c625cada552d8`.
-No branch was rebased, retargeted, rewritten, or force-pushed.
 
 ## 2026-08-09 PR #4304 corrected implementation evidence
 
@@ -1391,54 +625,6 @@ UI, TypeScript/Rust/PyO3/WASM physics, and downstream parity remain open.
 Protected CI, independent review, exact-head publication, parent integration,
 and consumer delivery are still required.
 
-## 2026-08-10 Exact repaired #4282 propagation into PR #4285 (continued)
-
-The inherited ground descendant passes 1,703 Python tests with two explicit optional
-`build123d` skips, 643 React tests across 105 files plus type-check,
-zero-warning lint and production build, 12 Rust tests, and 77 ground plus
-compatibility tests on real CPython 3.10. Ruff/format cover 78 changed Python
-files; pinned mypy and Bandit cover 52 changed production files. Campaign
-manifest, documentation, minimum-test, changed-assertion, 500-LOC, changed-file
-secrets, Python 3.10 compilation, and diff gates are clean. Protected
-current-head CI and required review remain separate release gates. Its exact
-repaired head is now incorporated into #4288; normal propagation of the
-reviewed #4288 head into #4298 is the next ancestry gate.
-
-The composed pinned-mypy profile additionally exposed integration-only `Any`
-returns at the Pydantic manifest and scalar plotting boundaries. The repaired
-#4282 parent owns the manifest annotation; scalar extraction normalizes the
-already-validated value with `float`. Bandit also replaced an optimized-away
-command-state assertion with an explicit invariant error. The alias finder
-retains its intentional optional-probe catch-and-continue boundary under a
-narrow explanatory `B112` annotation.
-
-## 2026-08-09 Ground parent reconciled with corrected wind carrier
-
-Draft PR #4285 remains on `feat/4268-ground-contract` with unchanged base
-`feat/4197-capability-observer`. Exact corrected wind carrier
-`bb101cedd555d07d493aae998b46050c68660cdd` is incorporated by the normal
-local merge containing this handoff; no branch was rebased, retargeted,
-force-pushed, or published. The result is the carrier-reconciled ground parent:
-it retains the strict flight-to-ground schemas, canonical fixtures, migrations,
-legacy-result adapter, dependency declaration, and pinned-mypy repairs while
-carrying the complete corrected wind/scalar/variation/Python-3.10 and
-consolidated capability ancestry.
-
-No qualified bounce, skid, roll, terrain-profile, total-distance, UI, or
-Rust/WASM ground physics is added by this propagation. PR #4288 must normally
-merge the resulting exact #4285 head before its flight-to-ground transfer can
-be evaluated as a descendant of the corrected carrier. Because #4285 still
-targets the observer branch while the corrected carrier now contains that
-observer through wind workflow #4282, this merge is the explicit
-observer-to-wind ancestry reconciliation; the GitHub base remains unchanged.
-
-Focused evidence is 89 ground-contract, compatibility, scalar-adapter, and
-responsive-wind tests on Python 3.11 and the same 89 on real CPython 3.10.20.
-Ruff check/format passes 34 focused Python files; pinned mypy 1.13 passes 23
-production modules. Ground production modules remain below 400 lines and
-functions below 50 lines, with no placeholder markers. The inherited campaign
-manifest validates and its nine manifest/parity contracts pass.
-
 ## 2026-08-09 PR #4302 pinned-MyPy current-head correction
 
 Hosted quality-gate run `31350134551` exposed four deterministic MyPy 1.13
@@ -1740,6 +926,7 @@ npm run type-check
 npm run lint
 npm run build
 ```
+
 ## 2026-08-09 PR #4302 deterministic-digest scanner repair
 
 Protected CI correctly remained blocking at head
@@ -1753,567 +940,3 @@ Commit this narrow repair with all three canonical handoffs and push normally
 to `feat/4270-ground-impact-bounce`. Do not retry the unchanged failed run,
 amend history, or force-push. Descendant PRs #4304 and #4305 inherit this file
 and must later receive the parent by ordinary merge commits.
-## 2026-08-11 parent handoff sections preserved during #4282 reconciliation
-
-## 2026-08-10 Second variation-parent propagation into PR #4281
-
-- Draft PR `#4281` retains branch `feat/4199-wind-scalar-adapter` and base
-  `feat/4144-variation-export-continuation`. Exact repaired parent head
-  `b90e5021a59e2081415b51ef29fbed06377bc201` is incorporated through a
-  normal merge commit without rebase, retarget, force-push, or history rewrite.
-- There is no scalar-adapter production/test conflict. Both append-only
-  handoff/SPEC histories are retained and the adapter contract is unchanged.
-- Parent quality success is not child release evidence. Fresh exact-head CI,
-  review, and all earlier dependencies remain open.
-- Reconciled-child evidence is 25 focused D-plane/impact tests plus docs
-  governance, changed-file size budget, and whitespace checks.
-
-## 2026-08-10 Parent Repair Propagation (#4202 → #4203)
-
-- Draft child PR `#4203` remains on `feat/4181-launch-monitor-registry`, based
-  on `feat/4189-dplane`; neither branch nor PR base is rewritten.
-- Original child head `08a2fdd8ce6bbc8fbb8f121927a677d4addb6b11` is
-  normally merged with exact parent head
-  `b443fdbed7064c5db0320106013c8413e3e24356`, in that parent order.
-- The semantic reconciliation preserves #4203's responsive
-  `SimulationViewControlsMixin` architecture while delegating persisted
-  D-plane checkbox state to the parent's focused `ImpactLayerControls`
-  helper. The existing `_impact_layer_checks` automation seam aliases the
-  helper's single checkbox mapping, so no duplicate UI state is introduced.
-- The inherited Python 3.10 compatibility repairs, frame-explicit D-plane
-  geometry, launch-monitor registry, responsive layout, and exports remain
-  additive. `simulation_view.py` and its controls mixin both remain below the
-  protected 500-line limit.
-- The exact original child also carried three ungrandfathered module-budget
-  blockers before this propagation: `simulation/sources.py` (540 LOC),
-  `plotting/catalog.py` (533 LOC), and `ui/pyqt6/main_window.py` (528 LOC).
-  Behavior-preserving extractions now isolate triple-pendulum dynamics,
-  plotting metadata, and versioned primary-navigation persistence. The legacy
-  source, catalog, and main-window imports are identity-preserving re-exports;
-  the split modules are 282/282, 459/98, and 494/85 lines respectively.
-- Focused reconciliation and extraction evidence is green: 36 PyQt
-  simulation/layout tests, 38 plotting/navigation tests, and 21 simulation
-  source/export tests. Final combined-stack evidence is 1,249 passing Python
-  tests with six explicit optional-dependency/Rust-wheel skips; 521 passing
-  React tests plus TypeScript, ESLint, and Vite production gates; 12
-  `swing-core` tests; real CPython 3.10 compilation/import checks; scoped
-  Ruff/Black/pinned MyPy; docs, minimum-test, assertion, changed-file size,
-  detect-secrets, and diff gates. A full-tree audit separately retains two
-  untouched non-candidate size findings (`kinetics.py`, 646 LOC;
-  `torque_profile_panel.py`, 612 LOC). Independent staged review found no
-  actionable findings after 95 additional focused tests. Exact-head protected
-  CI and required repository review remain release gates.
-
-## 2026-08-09 PR #4203 Python 3.10 string-enum compatibility
-
-Current-head child CI exposed a collection boundary that was already present
-in PR #4203's parent surface: seven Rate/shared swing modules imported
-`enum.StrEnum`, which does not exist on Python 3.10. Runtime imports now use
-the repository's existing `shared.python.compatibility.StrEnum`; type checking
-retains the native stdlib symbol behind `TYPE_CHECKING` so pinned mypy 1.13
-keeps enum-member types rather than weakening them to strings. No enum values,
-serialized contracts, physics, or UI behavior changed.
-
-This repair is published at exact #4203 head
-`ab7de5a47977417e02926c3fbc7476002e82b690`. Evidence:
-64 focused convention, D-plane, manual-delivery, flight-result, inverse,
-impact-family, capability, and compatibility tests pass; Ruff and format pass;
-pinned mypy 1.13 passes all eight changed Python files; and a real CPython
-3.10.20 probe verifies the shared fallback and all seven runtime import paths.
-Propagate the new parent normally through #4279, #4280, #4281, and #4282.
-
-The subsequent full Rate scan also found one parent-owned direct
-`datetime.UTC` import in the torque-profile controller. It now uses the same
-shared Python 3.10 compatibility module. The focused torque-profile UI suite
-and the real-3.10 source/runtime probe are required before publishing this
-follow-up; serialized timestamps remain UTC and the workspace schema is
-unchanged.
-
-## 2026-08-09 PR #4203 Linux collection repair
-
-Draft PR #4203 remains on `feat/4181-launch-monitor-registry`, based on
-`feat/4189-dplane`; no branch was rebased, retargeted, force-pushed, or merged
-on GitHub. Exact-head CI run `31199764932` passed the quality gate but all
-three Python lanes failed while collecting the in-package flight and solver
-facade contract tests. Pytest loaded those tests through its `src.shared...`
-package namespace, while their absolute dotted aliases requested the editable
-`shared...` namespace; Python then reported that `flight`/`solver` could not be
-imported from `src.shared.python.swing_sim` before any assertion ran.
-
-The bounded repair uses package-relative facade imports in those two tests, so
-collection and the public API assertions stay in one namespace. It does not
-change production code or widen either facade. Verify both contract modules
-with `--import-mode=importlib`, Ruff/format, and pinned mypy before a normal
-push. The run's Rust `-lpython3.11` link error is missing runner toolchain
-state, not simulation evidence; do not modify the model to hide it. After the
-new #4203 head passes, propagate it through #4279, #4280, #4281, and #4282 in
-normal stack order.
-
-Local evidence is now `12 passed` on Windows and `12 passed` under WSL Python
-3.11 with importlib collection. Ruff check/format and pinned mypy 1.13 pass
-for both changed test modules. The frozen-dataclass assertion casts only its
-introspection target to `Any`, matching the later carrier boundary while
-retaining the runtime assertion. The minimal WSL environment reports only
-unknown-option warnings for intentionally omitted optional pytest plugins.
-
-## 2026-08-10 PR #4202 D-plane ndarray typing repair
-
-- Draft PR `#4202` remains on `feat/4189-dplane`, based on
-  `feat/4162-wedge-impact-visualization`; the verified published repair base is
-  `b443fdbed7064c5db0320106013c8413e3e24356`.
-- CI Standard run `31384810375`, job `93442745760`, exposed two exact pinned
-  MyPy 1.13 `no-any-return` findings in the private D-plane ndarray helpers.
-  Explicit local ndarray result boundaries now preserve those helpers' return
-  contracts without changing validation, arithmetic, frames, schemas, or any
-  public API.
-- TDD evidence: the exact two-error MyPy failure was reproduced before the
-  repair; the same command is green afterward. Twenty-four focused D-plane,
-  impact-contract, impact-kinematics, and impact-scene tests pass, together
-  with seven metadata/pre-push contract tests, scoped Ruff, Ruff format, Black,
-  docs governance, minimum-test, module-size, changed-file-size, and diff
-  checks. An exploratory CI-workflow contract slice retains three unrelated
-  failures for later toolcache/environment steps absent from this older branch;
-  no workflow file is changed by this repair.
-- This is a bounded quality-gate repair, not completion of D-plane issue
-  `#4189`. Protected current-head CI, dependency order, and required review
-  remain release gates; no push, retarget, or draft-state change is included.
-
-## 2026-08-10 Parent Repair Propagation (#4179 → #4202)
-
-- Child PR `#4202` remains on `feat/4189-dplane`, based on
-  `feat/4162-wedge-impact-visualization`; neither branch nor PR base is
-  rewritten.
-- Original child head `b4abec03bccfbdd87ddf91427159c5c2332c21dd` is
-  normally merged with exact parent head
-  `6704a3e541a3e74c28b4a284530d1a21269dd340`, in that parent order.
-- The Python 3.10 UTC repair and source-wide AST guard remain additive to the
-  frame-explicit 3D D-plane geometry, desktop/web overlays, and export
-  contracts.
-- The persisted D-plane layer controls are extracted into a focused helper,
-  restoring `simulation_view.py` to the protected 500-line module budget
-  without changing its compatibility seam or behavior.
-- Combined-stack verification is green: 93 focused and 825 scoped Python tests
-  with two optional `build123d` skips; 360 React tests plus TypeScript, ESLint,
-  and Vite production gates; real CPython 3.10.20 compilation/UTC; Ruff/Black;
-  focused pinned MyPy 1.13; docs, minimum-test, file-size, detect-secrets, and
-  diff checks. The exact parent's 12 unchanged `swing-core` tests remain
-  applicable because this child has no Rust delta. The inherited broad MyPy
-  baseline remains 17 Qt/NumPy typing findings in 11 untouched files.
-  Protected CI and required review remain release gates.
-
-## 2026-08-10 Parent Repair Propagation (#4178 → #4179)
-
-- Child PR `#4179` remains on `feat/4162-wedge-impact-visualization`, based on
-  `feat/4166-wedge-turf-physics`; neither branch nor PR base is rewritten.
-- Original child head `0eb804e70887c788421332369e42792411aff55a` is
-  normally merged with exact parent head
-  `bfa83aedc88ead380babc73a699377d98b971006`, in that parent order.
-- The Python 3.10 UTC repair and source-wide AST guard remain additive to the
-  exact-event, locked-scale wedge impact visualization contracts.
-- Combined-stack verification is green: 58 focused and 739 scoped Python tests
-  with two optional `build123d` skips; 347 React tests plus TypeScript, ESLint,
-  and Vite production gates; real CPython 3.10.20 compilation/UTC; Ruff/Black;
-  focused pinned MyPy 1.13; docs, minimum-test, file-size, detect-secrets, and
-  diff checks. The exact parent's 12 unchanged `swing-core` tests remain
-  applicable because this child has no Rust delta. The inherited broad MyPy
-  baseline remains 17 Qt/NumPy typing findings in 11 untouched files.
-  Protected CI and required review remain release gates.
-
-## 2026-08-10 Parent Repair Propagation (#4174 → #4178)
-
-- Child PR `#4178` remains on `feat/4166-wedge-turf-physics`, based on
-  `feat/4161-wedge-ground-clearance`; neither branch nor PR base is rewritten.
-- Original child head `aaae3f73e17dbfaad5cca1dc6f49559b3aebe9d5` is
-  normally merged with exact parent head
-  `9ea93e92563280ec34bca682ad44d7409edd7a02`, in that parent order.
-- The Python 3.10 UTC repair and AST guard remain additive to the validated
-  turf-contact contracts and their explicit scientific/calibration boundary.
-- Combined-stack verification is green: 56 focused and 732 scoped Python tests
-  with two optional `build123d` skips; real CPython 3.10.20 compilation/UTC;
-  Ruff/Black; focused pinned MyPy 1.13; docs, minimum-test, file-size,
-  detect-secrets, and diff checks. With no TypeScript or Rust delta, the exact
-  parent evidence of 345 React tests/all web gates and 12 `swing-core` tests is
-  unchanged. The inherited broad MyPy baseline remains 17 Qt/NumPy typing
-  findings in 11 untouched files. Protected CI and review remain release gates.
-
-## 2026-08-10 Parent Repair Propagation (#4173 → #4174)
-
-- Child PR `#4174` remains on `feat/4161-wedge-ground-clearance`, based on
-  `feat/4163-impact-inspector`; neither branch nor PR base is rewritten.
-- Original child head `880a6465fc872cf3d6650283db154ddc41793a31` is
-  normally merged with exact parent head
-  `9ddaff3b6bca542fd7a2befc7d7b0ae53910a60a`, in that parent order.
-- The inherited Python 3.10 UTC repair and source-wide AST guard remain
-  additive to the swept wedge ground-clearance contracts.
-- Combined-stack verification is green: 56 focused Python tests; 703 scoped
-  Python tests with two optional `build123d` skips; 345 React tests plus
-  TypeScript, ESLint, and Vite production gates; 12 `swing-core` tests; real
-  CPython 3.10.20 compilation and UTC checks; Ruff/Black; focused pinned MyPy
-  1.13; docs, minimum-test, file-size, detect-secrets, and diff checks. The
-  inherited broad MyPy baseline is 17 Qt/NumPy typing findings in 11 untouched
-  files and remains outside this propagation scope. Protected CI and review
-  remain release gates.
-
-## 2026-08-10 Parent Repair Propagation (#4167 → #4173)
-
-- Child PR `#4173` remains on `feat/4163-impact-inspector`, based on
-  `feat/4144-variation-visualizations`; neither branch nor PR base is rewritten.
-- The original child head `3c43955aaeb3964ff8c3ef2748d626baae518b76`
-  is being merged normally with exact parent head
-  `22b66b560652b78de84141344c4ddd9a92a83b26`, in that parent order.
-- The inherited repair replaces Python 3.11-only `datetime.UTC` use in Rate
-  torque-profile persistence with `shared.python.compatibility.UTC` and adds a
-  Rate-source AST regression guard for direct, aliased, and module-attribute
-  access.
-- Combined local verification is green: 63 focused Python tests, all 562 Rate
-  tests, all 334 React tests, TypeScript type-check, ESLint, Vite production
-  build, 12 `swing-core` Rust tests, real CPython 3.10.20 compile/UTC checks,
-  Ruff check/format, Black, focused pinned MyPy 1.13, docs governance,
-  minimum-test, file-size, detect-secrets, and diff checks. A broader MyPy
-  sweep still reports 17 pre-existing Qt/NumPy typing errors in 11 untouched
-  files. Protected CI and required review remain release gates; queued,
-  cancelled, missing-toolcache, and dependency-download jobs are not green
-  evidence.
-- Every implementation commit must update this file, the Rate handoff, the
-  campaign handoff, and `SPEC.md`, or explicitly record why there is no
-  material handoff change.
-
-## Where This Repo Is Headed
-
-Tools is the D-sorganization fleet's shared engineering-tools monorepo (45+
-tools: PyQt6 GUIs, FastAPI/React web mirrors, Rust kernels). The current
-center of gravity is `src/rate_of_closure`, being grown from a single
-closure-rate calculator into a full swing → impact → ball-flight simulation
-platform under **Repository_Management#1390** (this handoff rollout) and a
-stack of golf-simulation epics:
-
-| Epic                                                                          | Status (one line)                                                                                                                                                                                                   |
-| ----------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
-| #4103 — Swing–Impact–Ball-Flight Simulation Platform                          | Phases 0-6 implemented on branch `feat/impact-simulation-platform`, consolidated into PR **#4119** (open, auto-merge armed, awaiting review). Phase 7 (WASM web parity swap, Pages CI) still open.                  |
-| #4120 — Investigation & Variation Suite (plotting/viewers/Monte Carlo/help)   | V1-V4 implemented, stacked on #4119, consolidated into PR **#4124** (open, draft-for-review, no auto-merge yet — targets `feat/investigation-suite`, itself stacked on #4119).                                      |
-| #4125 — Realistic Clubs/Kinetics/Putting/Public Release Mgmt/Showcase Styling | H1-H7 implemented, stacked on #4124, consolidated into PR **#4129** (open, draft-for-review, targets `feat/course-showcase`, stacked on #4124). H5 (public release-management repo) is cross-repo, not yet started. |
-| #4130 — Impact-Interval Club Dynamics (contact-interval rigid-body model)     | Foundation epic only (F1 formulation doc not yet started); no PR yet. Next major physics wave after #4125 lands.                                                                                                    |
-
-The separate shared Club Builder epic #4146 is active. Its first dependency
-slice, #4147, lives on `feat/4147-club-builder-core` and establishes the
-UI-independent assembly mass/CG/inertia, frame, length-datum, and persistence
-contracts that the later shaft, CAD, export, fitting, and UI issues consume.
-
-Active infrastructure repair: #4155 hardens the Rust/PyO3 job against
-incomplete setup-python cache entries whose interpreter works but whose
-declared link library is missing. The repair is isolated on
-`fix/4155-rust-libpython-cache` and does not change simulation code.
-
-See `src/rate_of_closure/AGENT_HANDOFF.md` for the detailed stack breakdown
-and architecture pointers for this tool specifically.
-
-## Must-Read Architecture Pointers
-
-1. `CLAUDE.md` — repo-wide conventions, CI gate list, cross-repo dependency
-   rules (Tools is a leaf dependency; UpstreamDrift and Gasification_Model
-   consume it).
-2. `docs/architecture/CANONICAL_TOPOLOGY.md` — canonical repo topology policy.
-3. `SPEC.md` — living specification; §12 Change Log requires a dated row for
-   every PR touching `src/` (enforced by `spec-check.yml`, see gates below).
-4. `src/rate_of_closure/AGENT_HANDOFF.md`, `src/pendulum_simulator/AGENT_HANDOFF.md`,
-   `src/rotation_converter/AGENT_HANDOFF.md` — per-tool handoff docs.
-5. `docs/AGENT_HANDOFF_TEMPLATE.md` — template for adding a handoff doc to a
-   new tool.
-
-## In-Flight Branches (what stacks on what)
-
-```
-main
- └─ feat/impact-simulation-platform   (PR #4119, epic #4103, auto-merge armed)
-     └─ feat/investigation-suite      (PR #4124, epic #4120, stacked on #4119)
-         └─ feat/course-showcase      (PR #4129, epic #4125, stacked on #4124)
-docs/agent-handoff-1390               (this branch, off origin/main, Repository_Management#1390)
-```
-
-Other active non-golf branches worth knowing about: `fix/file-size-budget-bounded-checkout`
-(#4096, CI checkout-scope fix), `agent/scada-phase-a-foundation` (#4091, SCADA
-epic #4085), several `scada/pr*` branches (SCADA epics #4085-#4089), and a
-handful of Bolt/Palette/Sentinel micro-PRs (#4070-#4102) unrelated to the
-golf-sim stack.
-
-## Gate Commands (repo-wide)
-
-```bash
-python3 -m ruff check .                          # lint
-python3 -m ruff format --check .                  # format check
-python3 -m pytest -n auto --timeout=60            # full test suite
-python3 -m pytest -m contract                     # API contract tests (downstream-facing)
-python3 -m pytest -m integration --timeout=60     # cross-repo integration
-```
-
-SPEC freshness (CI job `spec-freshness` in `.github/workflows/spec-check.yml`):
-any PR touching `src/**`, `tests/**`, `config/**`, `pyproject.toml`,
-`Cargo.toml`, `package.json`, or `requirements.txt` must also modify
-`SPEC.md` in the same PR, or carry the `spec-exempt` label. Runs on the
-`d-sorg-fleet` self-hosted runner.
-
-## Do-Not List
-
-- Do not modify public function signatures in `src/shared/python/**` without
-  opening coordinated migration issues in UpstreamDrift and Gasification_Model
-  (see `CLAUDE.md` Cross-Repo Dependencies).
-- Do not import across package boundaries (e.g. `signal_processing_studio`
-  importing from `sidekick.process_calculators`) — LoD is enforced.
-- Do not exceed the 500-LOC file budget on new/modified files in the golf-sim
-  packages (`rate_of_closure`, `swing_sim`, `swing-core`) — sub-package,
-  don't grow monoliths.
-- Do not use `git commit --no-verify` / `--push --no-verify` to bypass hooks;
-  see `CLAUDE.md` Hook bypass policy.
-- Do not regenerate the sidekick API baseline (`tests/sidekick_api_baseline.json`)
-  without coordinating a breaking-change migration.
-- Do not merge #4124 or #4129 ahead of their base (#4119, #4124 respectively)
-  — they are stacked and will conflict/duplicate SPEC.md sections if merged
-  out of order.
-- Do not hand-roll a GitHub Pages deploy workflow for `rate_of_closure/web`
-  yet — Phase 7 of #4103 owns this; today Pages hosting elsewhere in the repo
-  (e.g. `unit_converter`) is done via manual branch-folder publish, not CI.
-
-## Short-Term Roadmap (ordered)
-
-1. Land PR #4119 (base platform) — currently the long pole; everything else
-   stacks on it.
-2. Get #4124 out of draft-for-review and merge into `feat/investigation-suite`
-   → cascades onto #4119.
-3. Get #4129 out of draft-for-review and merge into `feat/course-showcase`
-   → cascades onto #4124.
-4. Start #4130 Phase F1 (formulation document) once #4125's stack is in.
-5. Phase 7 of #4103: WASM swap for the web mirror + real Pages CI deploy for
-   `rate_of_closure/web`.
-6. #4125 H5: stand up the public release-management repo (cross-repo, not
-   started).
-
-## 2026-08-11 additional parent sections preserved during #4282 reconciliation
-
-## 2026-08-10 Exact variation-parent propagation into PR #4281
-
-Draft PR `#4281` remains on `feat/4199-wind-scalar-adapter` with base
-`feat/4144-variation-export-continuation`. The normal merge starts from
-original child `cf52529b1e68479321bb93b1be3d59c77f782008` and incorporates exact
-published parent `8bcc49fc4e16e5e43be0b7f0f03c3017d5b79d0c` as its second parent. Do
-not rebase, retarget, force-push, rewrite either parent, or change draft state.
-
-The child's matching Python/TypeScript `scalar-ensemble/v1` contract and wind
-strategy plot adapter remain additive with the parent's selected-scatter
-exports, accessible raw tables, result-local linked selection, complete
-workspace/toolstrip, playback, navigation, compatibility, and module-budget
-repairs. Both adapters now enforce the same deterministic wind-scenario
-boundary: exact per-trial provenance, zero shear and turbulence, seed zero, and
-no gusts. SPEC 1.14.13 is the unique combined child entry above parent 1.14.12.
-
-Verification is green across 27 focused Python contract/adapter/selection
-tests, the complete Rate/shared-swing/golf-club matrix (`1,549 passed`, two
-explicit optional build123d skips), and all 556 React tests across 92 files.
-TypeScript, zero-warning ESLint, Vite production build, Ruff, Ruff-format,
-pinned MyPy 1.13, Bandit, the exact-parent two-file 500-line budget, docs,
-minimum-test, assertions, detect-secrets, diff, and CPython 3.10 compilation
-pass. Independent re-review found no actionable findings after checking the
-exact provenance derivation and all five deterministic-scenario regressions;
-protected current-head CI, unresolved threads, and required repository review
-remain release gates.
-
-## 2026-08-10 PR #4280 workspace timestamp propagation
-
-Draft PR #4280 remains on `feat/4144-variation-export-continuation` with base
-`feat/4218-toolstrip-workspace`. Exact corrected parent
-`05383d333b6fd87eaf5e37305476f50b505c2c2e` is incorporated through the normal
-merge containing this handoff, without rebasing, retargeting, force-pushing,
-or rewriting either branch. The child retains its selected-scatter CSV export,
-typed unavailable values, accessible raw tables, and focused visualization
-modules while inheriting the deterministic Python 3.10-3.12 UTC parser.
-
-The monotonic specification assigns the parent repair version 1.14.10 and the
-variation child version 1.14.11. The reconciled tree passes all `778` Rate
-tests, the real-CPython-3.10.20 compatibility suite (`27 passed`), the focused
-React variation suite (`1 file / 8 tests`), TypeScript, zero-warning focused
-ESLint, Ruff, format, and pinned mypy 1.13. Documentation, size, and diff gates
-must remain clean in the merge commit. Protected CI, review, and propagation
-to #4281 and later descendants remain separate release gates.
-
-## 2026-08-10 Second parent propagation into PR #4279
-
-- Draft PR `#4279` retains branch `feat/4218-toolstrip-workspace` and base
-  `feat/4181-launch-monitor-registry`; exact repaired parent head
-  `12dd76a8dbcc106c4683f2f2e53076f8dc6f1b76` is incorporated through a
-  normal merge commit with no rebase, retarget, or history rewrite.
-- The trees have no feature-code conflict. Reconciliation retains both
-  branches' append-only handoff/SPEC evidence, while the inherited D-plane
-  ndarray typing repair remains numerically and behaviorally neutral.
-- Parent quality-gate success does not authorize this child. Fresh exact-head
-  protected CI, review, and all dependency gates remain open.
-- Reconciled-child evidence is 25 focused D-plane/impact tests plus docs
-  governance, changed-file size budget, and whitespace checks.
-
-## 2026-08-10 Exact parent propagation into PR #4279
-
-Draft PR `#4279` remains on `feat/4218-toolstrip-workspace` with base
-`feat/4181-launch-monitor-registry`. This merge starts from original child
-head `05383d333b6fd87eaf5e37305476f50b505c2c2e` and incorporates exact
-published parent head `31cbc007d4c85b5479b7cd0fb0969124eab2af67` through a
-normal two-parent merge; no rebase, retarget, force-push, or draft-state change
-is permitted.
-
-The semantic reconciliation preserves the child's File/View/Tools workspace,
-module visibility, granular playback, trail, plot, and launcher integration.
-It also keeps the parent's focused `ImpactLayerControls`, plotting catalog,
-triple-pendulum, and primary-navigation modules. Workspace navigation now
-aliases the parent's canonical stable-ID/settings constants, while retaining
-the child's visibility and required-module state. Impact-layer automation and
-the rendered controls share the parent's single checkbox mapping.
-
-Combined verification is green: 1,339 Python tests pass with six explicit
-optional build123d/Rust-wheel skips; 545 React tests pass across 89 files, and
-TypeScript, ESLint, and the Vite production build pass. The post-format semantic
-rerun is 40 passing navigation/simulation GUI tests. Ruff, Ruff-format, Black,
-pinned MyPy 1.13 across all 18 changed production modules, Bandit, the exact
-18-file 500-line gate, docs, minimum-test, changed-test assertions,
-detect-secrets fingerprints, and staged/unstaged diff checks pass. CPython
-3.10.20 compiles every changed Python file and its dependency-free navigation
-state round-trip passes; 30 compatibility/date-boundary regressions also pass.
-There is no Rust delta from the exact parent, whose 12-test `swing-core`
-evidence remains applicable. Independent staged review found no actionable
-findings after 76 additional focused PyQt/navigation/workspace tests; protected
-exact-head CI and required repository review remain release gates after an
-ordinary guarded push.
-
-## 2026-08-10 PR #4279 fractional-timestamp Python 3.10 repair
-
-Exact-head CI on descendant PR #4281 exposed one remaining workspace parser
-difference: CPython 3.10 accepts only three or six fractional-second digits in
-`datetime.fromisoformat`, while newer supported interpreters accept the
-one-digit value already used by the workspace ordering contract. The earliest
-owner is PR #4279. Its shared workspace validator now enforces one anchored UTC
-grammar and parses zero through six fractional digits consistently, rejecting
-greater precision instead of silently truncating it on newer Python versions.
-UTC-only validation, serialized values, schema fields, and chronological
-comparison are unchanged.
-
-Evidence is `778 passed` for the full Rate suite and `45 passed` for the
-compatibility plus complete workspace document suites on the local supported
-interpreter, plus `27 passed` for the source-level compatibility suite on real
-CPython 3.10.20. The latter has only expected warnings for plugins absent from
-that intentionally minimal runtime. Ruff, formatting, pinned mypy 1.13,
-documentation governance, and the 400-line budget pass. Exact remote-head
-identity and normal publication/descendant propagation remain release gates.
-
-## 2026-08-09 PR #4281 scalar-ensemble parent propagation
-
-Draft PR #4281 remains on `feat/4199-wind-scalar-adapter` with unchanged base
-`feat/4144-variation-export-continuation`. Exact corrected #4280 parent
-`38ed58cab96842a3007e76a855db83ee2452b8fd` is incorporated by the normal
-merge containing this handoff; no branch was rebased, retargeted, force-pushed,
-or published by this continuation. The result preserves the complete parent
-workspace, Python 3.10 compatibility, and variation-export history together
-with #4281's independently owned scalar-ensemble and wind-strategy adapters.
-
-The child supplies matching UI-neutral Python/TypeScript
-`scalar-ensemble/v1` contracts with stable RFC3986 composite row identity,
-immutable nullable rows, and overall/per-cohort availability. Its wind adapter
-retains every actual and perfect-information status and exposes true/estimated
-wind, launch/aim, target, landing, miss, cost, and information delta. Impact
-variables remain explicitly unavailable because the analysis starts at
-prescribed launch. SPEC 1.14.12 records this child source delta separately
-from parent versions 1.14.11 and earlier.
-
-Current focused evidence is `49 passed` on Python 3.11 and `49 passed` on real
-CPython 3.10.20 with PyQt6/scientific dependencies present. React passes `2
-files / 10 tests`, TypeScript, and focused zero-warning ESLint. Ruff
-check/format passes the four child Python files; pinned mypy 1.13 passes both
-production modules and the corrected wind-adapter test. Documentation
-governance, ancestry/SPEC assertions, and final diff checks are required in
-this same local merge. Protected CI, review, publication, and descendant
-propagation remain separate release gates. The background runner, progress and
-cancellation, PyQt6/React strategy/scatter workflow, persistence, and export
-wiring remain incomplete and must not be represented as delivered.
-
-## 2026-08-09 PR #4280 parent propagation and SPEC restoration
-
-Draft PR #4280 remains on `feat/4144-variation-export-continuation` with base
-`feat/4218-toolstrip-workspace`. Exact corrected parent
-`3f67ed466fefc8991db9c4409f921f25e1c37142` is incorporated by the normal
-merge containing this handoff; no branch was rebased, retargeted, force-pushed,
-or published by this continuation. The result retains the complete workspace
-and Python 3.10 compatibility history together with #4280's independently
-owned variation scatter/export changes.
-
-The child exports every selected scatter axis to CSV with stable trial index,
-typed outcome, and explicit unavailable values in both PyQt6 and React. PyQt6
-also exposes the selected raw rows in bounded accessible tables, with shared
-table population and focused scatter/matrix modules. SPEC 1.14.11 now records
-this child source delta separately from parent versions 1.14.10 and earlier.
-
-Current evidence is `46 passed` on Python 3.11 and `19 passed` on real CPython
-3.10.20 with PyQt6 present. React passes `1 file / 8 tests`, TypeScript, and
-focused zero-warning ESLint. Ruff check/format passes five child Python files;
-pinned mypy 1.13 passes the four child production modules. Documentation
-governance, ancestry/SPEC assertions, and final diff checks pass locally.
-Protected CI, review, publication, and descendant propagation remain separate
-release gates.
-
-## 2026-08-09 PR #4279 Python 3.10 compatibility completion
-
-Draft PR #4279 remains on `feat/4218-toolstrip-workspace` with unchanged base
-`feat/4181-launch-monitor-registry`. Exact corrected parent
-`08a2fdd8ce6bbc8fbb8f121927a677d4addb6b11` was incorporated by the normal
-local merge `a340fabefa443d47325c5538f342683b38c01ade`; no branch was rebased,
-retargeted, force-pushed, or published by this continuation.
-
-The child-owned command registry and view-workspace enums now obtain
-`StrEnum` from `shared.python.compatibility` at runtime while retaining the
-native type behind `TYPE_CHECKING`. Workspace timestamp validation likewise
-uses the shared `UTC` value. This removes all three Python 3.11-only imports
-introduced by #4279 without changing command IDs, enum values, timestamp
-serialization, workspace schemas, or UI behavior. The merged regression
-guards nine string-enum modules and both UTC modules by inspecting the actual
-runtime import branch, then executes the three child workspace modules.
-
-Evidence is `126 passed` on Python 3.11, plus `14 passed` and a successful
-10-module dotted-import probe on real CPython 3.10.20. Ruff check/format passes
-the 11 production modules and compatibility test; pinned mypy 1.13 passes all
-11 production modules with the changed-file CI settings. Documentation
-governance and final diff checks must pass in this same local commit. Protected
-CI, review, publication, and descendant propagation remain separate gates.
-
-## 2026-08-09 PR #4279 launch-registry propagation
-
-Draft PR #4279 remains on `feat/4218-toolstrip-workspace` with unchanged base
-`feat/4181-launch-monitor-registry`. Exact parent head
-`08a2fdd8ce6bbc8fbb8f121927a677d4addb6b11` is incorporated through the normal
-merge commit containing this handoff; neither branch was rebased, retargeted,
-force-pushed, or pushed by this continuation. Source changes applied cleanly;
-the overlapping handoff and specification histories were reconciled
-monotonically. The result preserves the parent's package-relative facade and
-Python 3.10 enum repairs plus the child's workspace, toolstrip, playback, plot,
-and module-navigation implementation.
-
-Focused validation is `126 passed` across both facade contract modules and all
-Python/PyQt test files changed by #4279. React workspace validation is `8 files
-/ 32 tests passed`. Ruff check and format pass for all 28 relevant Python
-files. CI-pinned mypy 1.13 passes the 18 changed production files plus the two
-facade contract tests. The type gate exposed one real child-boundary defect:
-`legend_visible()` returned an untyped Qt value; it now converts that value at
-the widget boundary with `bool(...)`. The affected simulation GUI rerun is `29
-passed`; documentation governance and staged/unstaged diff checks also pass.
-
-## 2026-08-10 Second Parent Repair Propagation (#4202 -> #4203)
-
-- Draft child PR `#4203` retains base `feat/4189-dplane` and receives exact
-  repaired parent head `7d8d2f06dc797021d01939691e58f8425b652b33`
-  through a normal merge commit; no base, parent history, or draft state is
-  rewritten.
-- The propagated parent repair adds explicit NumPy ndarray result boundaries
-  to the two D-plane helpers that failed the hosted pinned MyPy gate. Numerical
-  behavior, frames, schemas, and UI behavior remain unchanged.
-- The exact parent quality gate is green. Protected child checks, review, and
-  all earlier dependency gates remain open; this propagation does not
-  authorize a merge or close #4189.
-- Child-tree verification after conflict reconciliation is 25 focused D-plane,
-  impact-contract, impact-kinematics, and impact-scene tests; docs governance,
-  changed-file size budget, and whitespace checks also pass. The Windows
-  unpinned MyPy 1.15 environment cannot parse the installed Python-3.12-only
-  NumPy stub syntax under this branch's Python 3.11 target, and WSL currently
-  fails to start with `E_FAIL`; neither attempt is reported as a passing gate.
diff --git a/SPEC.md b/SPEC.md
index 071cdeb96..4c64c7e88 100644
--- a/SPEC.md
+++ b/SPEC.md
@@ -26,162 +26,14 @@
 | **Owner**               | D-sorganization                            |
 | **Primary Language(s)** | Python 3.11+, Rust, JavaScript, TypeScript |
 | **License**             | MIT                                        |
-| **Current Version**     | 1.14.31                                    |
-| **Spec Version**        | 1.14.31                                    |
+| **Current Version**     | 1.14.35                                    |
+| **Spec Version**        | 1.14.35                                    |
 | **Last Spec Update**    | 2026-08-11                                 |
 
 ## 2. Purpose & Mission
 
 Comprehensive monorepo housing 45+ utility tools for data processing, scientific computing, process engineering, and automation. This is the central tooling hub for the D-sorganization fleet, providing modular engineering calculation tools with PyQt6 GUIs, FastAPI web services, Rust numerical kernels, and a unified launcher with plugin architecture for extensibility.
 
-### 2026-08-11 PR #4288 append-only history preservation repair
-
-The documentation-only follow-up to exact local candidate
-`1eb4188ff1a87a3d8e9a55aeca68194ca04e0bb4` restores parent handoff sections
-and change-log rows omitted during reconciliation and repairs one interleaved
-handoff paragraph. Production code, tests, schemas, the campaign manifest,
-child-first merge topology, and PR #4288's `feat/4268-ground-contract` base are
-unchanged. Publication and protected exact-head release evidence remain open.
-
-### 2026-08-11 Ground contract receives current workflow/base ancestry
-
-- Exact remote ground-contract child
-  `7a38b9838d743e051a6900620c4d6e754582aa89` is the first parent and exact
-  published workflow/base head
-  `b2baf6ef615b0d756a86ac4ca7eef2fc583210ee` is merged second, preserving
-  PR #4285's `feat/4197-capability-observer` base and child-first topology.
-- The strict `flight-to-ground/v1` contract remains behaviorally unchanged
-  while inheriting the current wind/capability stack, Ground/Tee parity, split
-  kinetics facade, and pinned formatting repairs. The ground-side kinetics
-  conflict was an obsolete formatting-only edit and contributes no behavior.
-- The retained scalar extractor narrowing is housed behind a 61-line public
-  plotting catalog facade. Ordered scalar and series entries are separated in
-  private collaborators, keeping every changed production module/function/
-  signature within 400 lines, 50 lines, and four non-receiver parameters while
-  preserving catalog keys, order, categories, units, and extractor behavior.
-- Verification passes the complete 1,539-test Rate/shared-swing Python matrix
-  with one optional Rust-wheel skip, all 648 React tests in 106 files,
-  TypeScript, zero-warning ESLint, the 188-module production build, 12 Rust
-  tests, 146 focused contract/parity tests, and 57 catalog/plot regressions.
-- This merge creates a local publication candidate only. Bounce, skid, roll,
-  terrain profiles, total distance, ground UI, Rust/WASM ground parity,
-  protected CI, approval, release, and downstream #4288 propagation remain
-  explicitly open.
-
-### 2026-08-11 Wind workflow receives current scalar-adapter parent
-
-- The responsive PyQt6/React wind-strategy workflow and consolidated
-  capability stack inherit exact scalar-adapter parent
-  `a7dce5f89b483303938f518b74b4028a1c68ba81` through a normal child-first
-  merge, preserving branch topology and both histories.
-- The validated split kinetics facade replaces one obsolete child formatting
-  hunk. Seven remaining child-owned paths receive pinned Ruff formatting only;
-  physics, public APIs, schemas, frames, units, and UI behavior are unchanged.
-- The exact skipped-import MyPy profile replaces one unused cache-return ignore
-  with an explicit `KineticsSeries | None` cast. The cached object and runtime
-  behavior are unchanged.
-
-### 2026-08-11 Wind strategy plot-adapter module-budget repair
-
-- `rate_of_closure.variation.wind_strategy_plot_adapter` remains the public
-  plot-data facade. Its request/analysis validation is delegated to the private
-  `_wind_strategy_plot_validation` collaborator so both production modules are
-  below the mandatory 400-line ceiling.
-- The public API, `scalar-ensemble/v1` wire shape and ordering, row identity,
-  values, typed availability, attributes, validation messages, and contract
-  prohibiting a flight-physics rerun are unchanged.
-- Private wind-scenario validation receives a scenario and one frozen, slotted
-  expectation object, keeping every production function at or below four
-  parameters and 50 lines without weakening any precondition.
-
-### 2026-08-11 Kinetics module-budget repair
-
-- The established `rate_of_closure.simulation.kinetics` module remains the
-  public façade for swing kinetics. Immutable result validation now lives in
-  `_kinetics_series.py`, and pure double-pendulum inverse/forward dynamics and
-  reaction-force helpers live in `_kinetics_dynamics.py`.
-- Public constants, classes, and functions retain identity-preserving imports;
-  the historical private `_reaction_forces` test/consumer seam remains an
-  alias to the extracted implementation. Physics, SI units, frames, numerical
-  fixtures, UI behavior, and serialization contracts are unchanged.
-- This extraction repairs the proactively reproduced changed-file failure:
-  exact head `572bf525d` versus `HEAD~1` selected the Ruff-formatted
-  `kinetics.py` at 646 LOC, above the ungrandfathered 500-LOC ceiling. The
-  façade and its focused implementation modules are now 222, 205, and 131 LOC.
-
-### 2026-08-09 Capability workflow contract hardening
-
-- Python and TypeScript capability workflow readers reject JSON primitive type
-  mismatches throughout the nested v1 document instead of coercing numeric
-  strings, booleans, or numeric text fields.
-- A shared golden parser fixture pins identical acceptance and rejection,
-  including portable acceptance of mathematically integral JSON numbers for
-  integer fields.
-- React Shot Optimizer numeric controls retain decimal draft text, allowing
-  ordinary leading-minus keyboard entry before validated numeric commit.
-
-### 2026-08-09 Capability result diagnostics
-
-- PyQt6 and React ranked-alternative views expose the full v1 optimization
-  diagnostics, explicit parameter units, evidence-envelope state, and Pareto
-  state rather than a reduced carry/dispersion summary.
-- Ranked results export through spreadsheet-safe CSV and a deterministic
-  `capability-result-export/v1` JSON envelope. Lossless raw observation exports
-  remain separate so aggregate recommendations cannot obscure failed or
-  no-impact trials.
-- React scalar-ensemble scatter plots render numeric grid/tick scales, visible
-  axis ranges, and cohort legends with paired-finite counts.
-
-### 2026-08-09 Rate campaign release-evidence authority
-
-- `docs/release/rate_of_closure_campaign.v1.json` is the normalized authority
-  for primary campaign issues, specifications, carriers, four-surface support,
-  commit-bound tests, limitations, dependencies, and release status.
-- `scripts/rate_campaign_manifest.py` validates the strict version-1 contract
-  and emits its deterministic JSON Schema. Feature-stack implementation,
-  protected parent integration, and exact `main` release are mutually
-  exclusive states with fail-closed cross-field requirements.
-- The current manifest truthfully records no complete campaign release SHA;
-  local tests and unprotected feature-parent merges cannot satisfy that claim.
-- Composed local implementation evidence is commit-bound in the manifest;
-  maintainers must still refresh live mutable PR heads before any write,
-  merge, release, or issue-closure decision.
-- The manifest loader explicitly narrows Pydantic's validated value for the
-  CI mypy 1.13 `--follow-imports=skip` context without weakening runtime
-  validation or the generated schema.
-- The PyQt playback adapter similarly converts Qt's elapsed-time result to a
-  concrete `float`, preserving the typed seconds contract when Qt imports are
-  deliberately skipped by the changed-file CI lane.
-
-### 2026-08-09 Swing-core parity CI isolation
-
-- The dedicated `swing_core` Rust-to-Python parity lane disables third-party
-  pytest plugin auto-loading so cached GUI plugins on self-hosted runners cannot
-  introduce undeclared PyQt dependencies into the Rust-only qualification job.
-- The lane continues to install the built wheel and run the same non-skippable
-  parity suite; numerical behavior and production dependencies are unchanged.
-
-### 2026-08-09 Rate web-release stability
-
-- The documented direct Rate web launcher imports its registration through the
-  package-qualified path and dynamically loads the repository bootstrap without
-  direct `sys.path` mutation, so `python src/rate_of_closure/launch_web.py`
-  works outside package-module execution and satisfies changed-file policy.
-- Rate supports a static Vite web release and a PyQt6/PyInstaller desktop
-  release. It does not claim Tauri support without a `src-tauri` project or an
-  artifact qualification lane.
-- The Flight Explorer parent-panel test explicitly settles its lazy wind-panel
-  import inside React `act`, removing suite-load timing from the assertion.
-
-### 2026-08-10 D-plane ndarray typing boundary repair
-
-- The private D-plane vector conversion and horizontal-projection helpers bind
-  NumPy expression results to explicit ndarray locals before returning them.
-- This preserves the existing numerical calculations and DbC validation while
-  satisfying the changed-file MyPy `no-any-return` contract on Python 3.12.
-- No public API, reference frame, serialized schema, physics assumption, or UI
-  behavior changes in this repair.
-
 ### 2026-08-06 GUI module-budget repair
 
 - Movement Optimizer motion palette and chain-length helpers live in a small
@@ -194,7 +46,7 @@ unchanged. Publication and protected exact-head release evidence remain open.
 
 ## 3. Goals & Non-Goals
 
-### 2026-08-10 Regional Surface-Plan Editor First Slice
+### 2026-08-10 Regional Surface-Plan Canonical Request I/O
 
 - The standalone PyQt6 and React shells expose matched `Ground Surfaces`
   modules for editing one SI base domain and one to eight coplanar static
@@ -208,9 +60,28 @@ unchanged. Publication and protected exact-head release evidence remain open.
   errors preserve the draft and provide accessible error state; successful
   validation exposes schema, SI-unit, provenance, and canonical request
   readback.
-- This first slice is session-only. Ground execution/playback, result import,
-  export, measured calibration workflows, model-input persistence, compiled
-  regional physics, and UpstreamDrift consumers remain explicit non-goals.
+- Native PyQt6 Open/Save As commands use strict complete-read validation and
+  atomic UTF-8 replacement. Browser import/download uses the same bounded
+  canonical request and discloses that browser downloads cannot guarantee a
+  path, atomic replacement, or recent-file access. Workspace persistence is a
+  separate contract.
+- Imports accept only editor-v1 provenance, providers, geometry/axis
+  qualification, and row capacity. Unchanged imports preserve the exact
+  request and provenance; edits create a fresh draft digest. External or
+  differently qualified requests are rejected without coercion.
+- Both runtimes accept wire numbers only inside JavaScript's exact safe range.
+  Native precedence entry remains integer-exact through the inclusive v1
+  maximum of 9,007,199,254,740,991 and cannot silently narrow imported data.
+- Native request import reads a single binary-handle snapshot bounded to the
+  1 MiB wire limit plus one sentinel byte, then performs strict UTF-8 decoding.
+  Concurrent file growth cannot trigger an unbounded allocation.
+- PyQt6 canonical-precision inputs share one frozen `NumberInputSpec` and
+  three-parameter construction helper. The regional I/O editor retains its
+  eleven-decimal presentation and inclusive cross-runtime safe bounds without
+  restoring the superseded six-parameter helper.
+- Ground execution/playback, result interchange, measured calibration
+  workflows, model-input workspace persistence, compiled regional physics,
+  and UpstreamDrift consumers remain explicit non-goals.
 
 ### 2026-08-06 Impact-to-Flight Solution-Family Foundation
 
@@ -457,15 +328,6 @@ unchanged. Publication and protected exact-head release evidence remain open.
   overrides survive club changes, legacy runs migrate to Ground, and the
   derived ball center drives contact, alignment, impact records, flight origin,
   and both standalone renderers.
-- Python and React consume the same version-1 golden ball-setup fixture. Its
-  strict metadata fixes SI metre units and the
-  `ground_plane_to_ball_bottom` reference; its cases pin club defaults,
-  explicit overrides, Ground's zero effective tee height, derived geometry,
-  serialization, invalid-height rejection, and legacy migration.
-- Visual verification uses semantic state and structural image contracts, not
-  pixel-perfect baselines: Playwright records default Tee and rerun Ground web
-  states with zero browser errors, while the hidden-window PyQt harness pins
-  canonical center/artist state and nonblank, distinct Ground/Tee captures.
 - Variation plans retain their complete v2 schema and can be saved, loaded,
   duplicated, and deleted from a versioned local library. Users can select
   simultaneous, one-at-a-time, or combined analyses, while paired common-
@@ -3073,32 +2935,24 @@ Active development with stable core, continuous tool expansion, and web API in p
 
 | Date | Version | Changes |
 | ---- | ------- | ------- |
+| 2026-08-11 | 1.14.35 | fix(rate-of-closure, #4342): retain the default canonical JSON encoder's JavaScript-safe numeric boundary for ground contracts while adding an explicit capability-observation-only extended finite-float policy that reuses the same recursive encoder; keep integers safe-range bounded, emit integral finite doubles such as `1e20` and `1e21` as exact exponent-free decimal tokens matching the TypeScript capability serializer, and preserve fail-closed non-finite handling with Python, TypeScript, ground-boundary, full-suite, static-analysis, build, and governance evidence. |
+| 2026-08-11 | 1.14.34 | merge(rate-of-closure, #4274 #4339 #4342): normally merge exact current regional request-I/O child `c1f47f2ef68b3db102da5416aaac17a40f675207` first with exact reviewed local regional-editor parent candidate `db335937afc4b587d235eb705e315f577519c5e6` second while preserving PR #4342's `feat/4274-regional-surface-ui` base, child-owned canonical request import/export, bounded UTF-8 and safe-number validation, native atomic save, browser-qualified download, tests, limitations, and all append-only handoff/SPEC history; inherit current editor, wire, regional-physics, and complete ground ancestry without rebase, retarget, rewrite, force-push, or publication, while keeping execution/playback, result interchange, measured calibration, model-input persistence, changing geometry or velocity, TypeScript/compiled regional physics, downstream parity, protected exact-head evidence, review, approval, dependency integration, and release open. |
 | 2026-08-11 | 1.14.31 | merge(rate-of-closure, #4274 #4335 #4339): normally merge exact current regional-editor child `d21741e312b849a63f73cabf351a15d9de80fb94` first with exact published regional-wire parent `8f933ed8dcb29e55ece4ec6bb1e60813f6794d57` second while preserving PR #4339's `feat/4271-regional-wire-contract` base, matched PyQt6/React editors, strict validation invalidation and readback, parent wire/resolver/regional-physics/ground ancestry, and both append-only handoff/SPEC histories; retain the extracted navigation-state contract with `regional_surfaces` in default and legacy migration order, without rebase, retarget, rewrite, force-push, or publication, while keeping execution/playback, result interchange, measured calibration, model-input persistence, changing geometry or velocity, TypeScript/compiled regional physics, downstream parity, protected exact-head evidence, review, approval, dependency integration, and release open. |
+| 2026-08-11 | 1.14.33 | docs(rate-of-closure, #4342): restore verbatim the omitted PR #4335, #4332, and #4304 current-parent ancestry sections in all three canonical handoffs and the parent's exact parallel-branch SPEC 1.14.29 and 1.14.28 reconciliation rows; record the historical version collision without rewriting either branch's append-only record; change no production code, tests, schemas, manifest state, PR base, protected evidence, or release status. |
+| 2026-08-11 | 1.14.32 | chore(rate-of-closure, #4274 #4342): normally reconcile exact published regional request-I/O child `8e1c7ccd99a7c4886c5fb9ccc7e4d94a6d7e3833` with exact newly published regional-editor parent `d21741e312b849a63f73cabf351a15d9de80fb94` without changing PR #4342's base; preserve native atomic and browser-qualified canonical request I/O, safe-number and bounded UTF-8 behavior, and the parent's frozen validated engineering-number specification by delegating child controls to its three-parameter helper while retaining eleven-decimal presentation; keep protected CI, approval, integration, execution, calibration, downstream parity, and release open. |
+| 2026-08-11 | 1.14.31 | fix(rate-of-closure, #4274 #4342): remove one redundant Boolean cast rejected by protected delta-MyPy without changing wire bytes, validation, file semantics, UI behavior, or physics. |
 | 2026-08-11 | 1.14.30 | fix(rate-of-closure, #4274 #4339): replace the six-parameter PyQt engineering-number helper with a three-parameter helper and frozen validated specification while preserving names, values, units, ranges, increments, tooltips, and field order; add RED-first helper/configuration and structural-bound tests, retain byte-stable wire/digest/physics behavior, and keep protected CI, approval, integration, and release open. |
 | 2026-08-11 | 1.14.29 | chore(rate-of-closure, #4274 #4339): normally reconcile exact published regional-surface editor child `cbb9c0a6bdc6a50f59f7a661139b9d53e1892980` with exact newly published regional-plan wire parent `9e01ccc3e891cc45907293751a192624195a77a5` without changing the PR base or rewriting either branch; retain matched PyQt6/React request editors, invalidation and engineering hints together with the strict Python/TypeScript plan contract, resolver binding, bounded coplanar regional physics, and complete ground ancestry, while preserving execution/playback, result interchange, measured calibration, workspace persistence, changing geometry, compiled-runtime physics, downstream parity, protected evidence, and release as open work. |
-| 2026-08-11 | 1.14.29 | merge(ground-regional-wire, #4332 #4335): normally merge exact remote regional-wire child `9e01ccc3e891cc45907293751a192624195a77a5` first with exact current regional-surface parent `5ac3c1f233de30dd42eeac57073cbdb149bde652` second while preserving PR #4335's `feat/4271-regional-surface-transitions` base, strict Python/TypeScript regional-plan request/result schemas, canonical JSON/SHA-256 fixture evidence, fail-closed parsing, Python resolver binding, tests, limitations, and both append-only handoff/SPEC histories; inherit current regional, skid/roll, impact/bounce, and flight-transfer ancestry without rebase, retarget, force-push, parent rewrite, or remote publication, while keeping changing geometry or velocity, internal transition-ledger export, regional UI, compiled or TypeScript regional physics, downstream parity, protected exact-head CI, review, approval, and release open. |
 | 2026-08-11 | 1.14.28 | chore(ground-regions, #4271 #4335): normally reconcile exact live regional-wire child `74a053d2d544da9f44a88007660ad28c0127f285` with exact newly published regional-surface parent `04ccf08dd990de1cd056a3420e67772773a4be2e` without changing the PR base or rewriting either branch; preserve the strict Python/TypeScript plan wire boundary and Python resolver binding together with the parent's reconciled coplanar regional physics and complete skid/roll ancestry, while retaining protected CI/review, UI, compiled or TypeScript regional physics, changing geometry/velocity, internal transition-ledger export, downstream parity, and release as open work. |
+| 2026-08-10 | 1.14.29 | feat(rate-of-closure, #4274): add strict matched PyQt6 native atomic Open/Save As and React bounded import/canonical download for editor-qualified regional requests; preserve exact unchanged provenance, rebind edited drafts, reject unsupported qualifications without coercion, keep workspace persistence separate, and retain execution, calibration, protected evidence, and release as open work. |
+| 2026-08-10 | 1.14.28 | feat(rate-of-closure, #4274): add matched PyQt6 and React regional surface-plan editors with explicit SI material inputs, one-to-eight bounded overlays, illustrative/unvalidated and persistence/execution qualifications, accessible strict validation/readback, provenance bound to the actual draft, and delegation to the separate regional-plan v1 wire contract; retain playback, execution, result interchange, measured calibration, workspace model-input persistence, other runtimes, protected evidence, and release as open work. |
 | 2026-08-10 | 1.14.27 | feat(ground-roll, #4271): add separate strict Python/TypeScript `ground-regional-material-plan-request/v1` and result/v1 wire contracts with exact keys, SI units, explicit coplanar/static limitations, finite bounded base/overlay intervals, bounded region/document counts, unique region/precedence/surface identities, canonical cross-runtime JSON/SHA-256 evidence, fail-closed no-fabrication result binding, and a Python-only adapter into the existing regional resolver; preserve the frozen flight-to-ground v1 request/result and retain TypeScript/compiled regional physics, UI, changing geometry, internal transition-ledger export, downstream parity, protected CI, review, and release as open work. |
 | 2026-08-10 | 1.14.26 | feat(ground-roll, #4271): add bounded Python-reference support for finite coplanar material overlays with unique explicit precedence, exact quadratic boundary splitting, base-edge precedence, state/energy continuity, strict Python/TypeScript `surface_transition` events, internal from/to identity evidence, typed transition limits, and randomized piecewise-analytic qualification; reject changing normals, height/velocity discontinuities, and overlays owning the impact handoff rather than fabricating impulses or geometry transitions, while retaining UI, compiled regional physics, and downstream parity as open work. |
-| 2026-08-11 | 1.14.28 | merge(ground-regions, #4304 #4332): normally merge exact remote regional-surface child `04ccf08dd990de1cd056a3420e67772773a4be2e` first with exact current skid/roll parent `3f861c1faaa7455faee92eaa7f813174667208e0` second while preserving PR #4332's `feat/4271-ground-skid-roll` base, bounded coplanar material overlays, exact transition splitting and evidence, regional production, tests, fixtures, contracts, limitations, and both append-only handoff/SPEC histories; inherit current impact, bounce, skid, roll, and flight-transfer ancestry without rebase, retarget, force-push, parent rewrite, or remote publication, while keeping changing normals, discontinuous geometry or velocity, regional UI, compiled regional physics, downstream parity, protected exact-head CI, review, approval, and release open. |
-| 2026-08-11 | 1.14.27 | chore(ground-regions, #4271 #4332): normally reconcile exact live regional-surface child `1a48d749af508843fac2a5102f4dd56294429bda` with exact newly published skid/roll parent `0ea6740965068542e9d8c7449e06ec07d88969e0` without changing the PR base or rewriting either branch; preserve bounded coplanar material overlays, exact transition splitting and evidence, strict contracts, and all current parent physics while retaining changing normals, discontinuous geometry/velocity, UI, compiled regional physics, and downstream parity as open work. |
-| 2026-08-11 | 1.14.27 | merge(ground-roll, #4302 #4304): normally merge exact remote skid/roll child `0ea6740965068542e9d8c7449e06ec07d88969e0` first with exact current impact/bounce parent `62beb3e1aa951645fd556a53f9cbae4bb46c47e5` second while preserving PR #4304's `feat/4270-ground-impact-bounce` base, arbitrary-plane skid, static-feasible pure roll, rolling resistance, qualified rest, edge localization, strict result composition, passive ledgers, fixtures, tests, limitations, and both append-only histories; inherit current bounce and flight-transfer ancestry without rebase, retarget, force-push, parent rewrite, or remote publication, while keeping regional/changing-normal surfaces, deformation, torsional damping, roll-to-skid transitions, UI, compiled physics, protected exact-head CI, review, approval, downstream integration, and release open. |
-| 2026-08-11 | 1.14.26 | chore(ground-roll, #4271 #4304): fast-forward the local skid/roll carrier to exact live child `52d9a6a978d8e6b8b19ef92f02f265c9058b00ad`, then normally merge exact live impact/bounce parent `cf6e72bad98e5f36f782254942e6895b8b71e670` without changing the PR base or rewriting either branch; retain the reviewed skid/roll physics and strict contracts while reconciling current ancestry and deterministic formatting evidence. |
 | 2026-08-10 | 1.14.25 | fix(ground-roll, #4271 #4304): normally propagate exact corrected ground-impact/bounce parent `846653c21bd61a40aab99ab838c29915d0728e70` into the skid/roll descendant without changing its base; preserve the reviewed arbitrary-plane skid, pure-roll, resistance, rest, edge, composition, and passive-ledger behavior while inheriting corrected flight-transfer ancestry, deterministic workspace timestamp parsing, and canonical `swing_sim` import identity; retain explicit exclusions for regional/changing-normal surfaces, terrain deformation, torsional spin damping, roll-to-skid transitions, UI, compiled physics, and downstream parity. |
 | 2026-08-10 | 1.14.24 | fix(ground-roll, #4304): explicitly allowlist the immutable SHA-256 golden-fixture digest as non-secret test evidence so protected secret scanning distinguishes a scientific integrity fingerprint from a credential; no physics, numerical, schema, API, or fixture bytes change. |
 | 2026-08-09 | 1.14.23 | feat(ground-roll, #4271): continue exact #4270 capture through arbitrary-plane kinetic skid, static-feasible pure roll, rolling resistance, qualified rest, finite-axis edge localization, bounded typed limits, relative surface-path and passive energy ledgers; compose representable prefix/suffix evidence into strict v1 results without duplicate or epsilon-time samples; pin an analytic shared fixture and retain explicit exclusions for regions, changing normals, torsional spin damping, roll-to-skid, UI, compiled runtimes, and downstream parity. |
-| 2026-08-11 | 1.14.26 | merge(ground-impact, #4288 #4302): normally merge exact remote ground-impact/bounce child `cf6e72bad98e5f36f782254942e6895b8b71e670` first with exact current flight-ground-transfer parent `c39297d37a17d5d8d3520ed62c5563cdcc609cab` second while preserving PR #4302's `feat/4269-flight-ground-transfer` base, passive restitution/Coulomb impact, deterministic repeated bounce, exact-contact and capture-to-skid behavior, cancellation/failure evidence, and both append-only histories; apply AST-equivalent pinned Ruff 0.14.10 layout to three inherited PR-delta files; inherit the current transfer and strict-ground ancestry without rebase, retarget, force-push, parent rewrite, or remote publication, while keeping skid, roll, rest, total distance, final ground results, UI, compiled bounce runtimes, protected exact-head CI, review, approval, and release open. |
 | 2026-08-10 | 1.14.22 | fix(ground-impact, #4270 #4302): normally propagate exact corrected flight-transfer parent `247215422a6d4b677552955b4923bc609a553259` into the ground-impact/bounce descendant without changing its base; preserve the passive restitution/Coulomb impulse, deterministic repeated hops, exact-contact and capture-to-skid behavior while inheriting deterministic workspace timestamp parsing and canonical `swing_sim` import identity; retain explicit non-delivery boundaries for skid, roll, rest, total distance, final ground results, UI, TypeScript physics, and compiled bounce runtimes. |
 | 2026-08-09 | 1.14.21 | feat(ground-impact, #4270): add a typed passive restitution/Coulomb sphere-plane impulse with full spin and moving-boundary accounting, deterministic repeated ballistic hops, exact bracket contact, capture-to-skid handoff, cancellation and bounded failure states, airborne-segment evidence, a shared golden fixture, and analytic/property/convergence tests; retain #4271 ownership of skid, roll, rest, total distance, and final `GroundSimulationResult`, with UI and compiled runtimes explicitly excluded. |
-| 2026-08-11 | 1.14.25 | merge(flight-ground, #4285 #4288): normally merge exact remote flight-ground-transfer child `df716ac740beb9d47dced66f867dc3dde1fa5367` first with exact current strict-ground parent `584e4e53a7990053a346938c4e45538786887ec7` second while preserving #4288's `feat/4268-ground-contract` base, complete Python/TypeScript/Rust/PyO3/WASM transfer production/tests/fixtures/contact evidence, cross-runtime contracts, and explicit incomplete ground-physics/UI/release limitations; retain both append-only histories without rebase, retarget, force-push, parent rewrite, or remote publication, and keep protected exact-head CI, review, approval, and release open. |
-| 2026-08-11 | 1.14.24 | docs(flight-ground, #4285 #4288): restore every parent handoff section and specification history row omitted during the local #4288 reconciliation, repair the interleaved rate-of-closure handoff paragraph, and preserve the exact candidate's implementation, tests, schemas, campaign manifest, child-first topology, and `feat/4268-ground-contract` base; keep publication, protected exact-head CI, approval, dependency integration, and release open. |
-| 2026-08-11 | 1.14.23 | chore(flight-ground, #4285 #4288): normally reconcile exact published flight-transfer child `afb1bdcfc6701caaf1f7bc3497a6a37dd9698c14` with exact newly published strict-ground parent `e09ab96280cab363b36bd0e7db4a3cd064dc2527` without changing #4288's base or rewriting either branch; preserve the child's bounded TypeScript flight integrator, signed terminal angular state, physical sphere/terrain contact, and Python/TypeScript/Rust/PyO3/WASM transfer while inheriting the parent's current wind/capability, strict-ground, split-kinetics, plotting-catalog, and formatting ancestry; apply AST-equivalent pinned Ruff normalization to seven child-owned Python files and standard line endings to the preserved TypeScript facade; retain bounce, skid, roll, terrain response, total distance, UI, protected evidence, and release as open work. |
-| 2026-08-11 | 1.14.24 | merge(rate-of-closure, #4283 #4285): normally merge exact remote ground-contract child `e09ab96280cab363b36bd0e7db4a3cd064dc2527` first with exact current capability-observer carrier `d0b827667e61f0583e1dad0b3cbbca6819624d3c` second while preserving #4285's `feat/4197-capability-observer` base, complete strict `flight-to-ground/v1` production/tests/schemas/fixtures/dependencies, fail-closed unimplemented-physics boundary, and both append-only histories; inherit current wind/capability/release ancestry without rebase, retarget, force-push, parent rewrite, or remote publication, and keep protected exact-head CI, review, approval, release, and normal downstream #4288 propagation open. |
-| 2026-08-11 | 1.14.22 | merge(rate-of-closure, #4282 #4283 #4285): normally merge exact remote ground-contract child `7a38b9838d743e051a6900620c4d6e754582aa89` first with exact current workflow/base head `b2baf6ef615b0d756a86ac4ca7eef2fc583210ee` second, preserving PR #4285's base and strict flight-to-ground contract while inheriting the complete current wind/capability, Ground/Tee parity, split-kinetics, and pinned-format ancestry; resolve the obsolete formatting-only kinetics hunk wholly to the validated parent facade, retain both append-only histories, and keep protected CI, approval, publication, release, qualified ground physics, and downstream #4288 propagation open. |
-| 2026-08-10 | 1.14.19 | fix(rate-of-closure, #4282 #4285): normally propagate exact quality-green responsive wind-workflow head `1e82f15026786ea0b08f78f4c001590ddce9ff39` into strict flight-to-ground contract child `e5bcbd1096d3be1f621a805c9d9f3fd321e375a5` without production/test conflict, topology change, or history rewrite; retain the fail-closed contract boundary and require fresh CI/review before normal #4288 propagation. |
-| 2026-08-10 | 1.14.18 | feat/fix(ground, #4199 #4268 #4282 #4285): normally propagate exact repaired responsive wind-workflow carrier `686016196a2f895058b8a566dff103a0fd32cd10` into original strict ground-contract child `788aa547651a3685a363ea401824a5d81477bafb` while preserving base `feat/4197-capability-observer`; retain the fail-closed ground schemas, fixture, migrations, legacy adapter, explicit dependency, import-alias repair, and pinned typing repairs while inheriting the complete deterministic workspace, variation, scalar-wind, capability, campaign-release, and hosted-mypy manifest-repair ancestry; normalize contract-checked scalar plot extraction so the composed skipped-import mypy profile cannot leak `Any`, enforce disabled-command invariants under optimized Python, and document the optional alias finder's reviewed catch-and-continue security boundary; preserve the established no-bounce/no-skid/no-roll/no-terrain/no-total-distance/no-ground-UI/no-Rust-WASM delivery boundary and require #4288 to acquire this exact ancestry through its own normal descendant merge. |
-| 2026-08-10 | 1.14.22 | fix(flight-ground, #4285 #4288): normally propagate exact quality-green strict-ground-contract head `a93edb4bfd6a8dc9334122cd2ae660983d5bf424` into flight-transfer child `108a841b1378c992defd3c7b7ee263d41a6c8b24` without production/test conflict, topology change, or history rewrite; retain the qualified cross-runtime transfer and explicit non-delivery boundaries, with fresh CI/review required before release. |
-| 2026-08-10 | 1.14.21 | fix(flight-ground, #4199 #4269 #4282 #4285 #4288): normally propagate exact repaired strict-ground parent `e5bcbd1096d3be1f621a805c9d9f3fd321e375a5` into original flight-transfer child `247215422a6d4b677552955b4923bc609a553259` while preserving base `feat/4268-ground-contract`; retain the qualified signed terminal state, physical sphere/terrain contact brackets, strict provenance, and Python/TypeScript/Rust/PyO3/WASM transfer while inheriting the deterministic workspace, variation, scalar-wind, capability, campaign-release, hosted-mypy, import-identity, scalar-extraction, command-invariant, and security-boundary repairs; preserve explicit non-delivery boundaries for bounce, skid, roll, terrain response, total distance, and UI. |
 | 2026-08-10 | 1.14.20 | fix(flight-ground, #4269 #4288): normally propagate exact corrected ground-contract parent `788aa547651a3685a363ea401824a5d81477bafb` into the flight-transfer descendant without changing its base; preserve the qualified Python/TypeScript/Rust/PyO3/WASM transfer while inheriting deterministic workspace timestamp parsing and canonical `swing_sim` import identity; retain explicit non-delivery boundaries for bounce, skid, roll, terrain response, total distance, and UI. |
 | 2026-08-09 | 1.14.19 | chore(flight-ground, #4269 #4288): normally propagate exact carrier-reconciled ground parent `6a2bc9d06f6f9a28a0d615b19d2ed4fc13871059` into the flight-transfer descendant without changing its base; preserve the qualified cross-runtime transfer and semantic facade assertions while completing corrected wind-to-ground ancestry; bind terminal samples before exact `FlightStatePoint` narrowing in the pinned-mypy regression; and retain explicit non-delivery boundaries for bounce, skid, roll, terrain response, total distance, and UI. |
 | 2026-08-09 | 1.14.18 | feat(flight-ground, #4269): propagate full signed terminal angular state and qualified physical sphere/terrain contact brackets across Python, TypeScript, Rust, PyO3, and WASM; preserve exact launch-origin evidence, vertical tee and terrain geometry, strict v1 surface/calibration/provenance data, interpolated terminal contact, typed unavailable outcomes, and bounded synchronous web integration. Reconcile the transfer integrator with the current capability and strict-ground ancestry through normal stacked merges without retargeting or rewriting either PR, and remove the resulting ground/flight package-facade cycle through direct record/type dependencies. |
@@ -3109,55 +2963,12 @@ Active development with stable core, continuous tool expansion, and web API in p
 | 2026-08-10 | 1.14.13 | fix(rate-of-closure, #4282): normally propagate exact corrected scalar-ensemble parent `cf52529b1e68479321bb93b1be3d59c77f782008` into the responsive PyQt6/React wind-workflow carrier without changing its base; retain the consolidated capability workflow and release authority while inheriting the strict cross-version UTC parser, variation export, immutable scalar rows, stable composite identity, availability, and typed wind adapter; route the child-owned capability-observation `StrEnum` through the shared Python 3.10 runtime contract and extend the compatibility regression. |
 | 2026-08-10 | 1.14.12 | feat(wind variation, #4199 #4281): add matching UI-neutral Python/TypeScript `scalar-ensemble/v1` contracts with structured provenance, labeled stages/categories/cohorts, unit-bearing variables, stable RFC3986 composite row IDs, immutable nullable raw rows, and overall/per-cohort availability; adapt every actual and perfect-information wind-strategy outcome with explicit status, true/estimated wind, launch/aim, target, landing, miss, cost, and information-delta values while leaving impact variables explicitly unavailable because the analysis begins at prescribed launch. |
 | 2026-08-10 | 1.14.11 | feat(variation, #4144 #4280): complete PyQt6/React selected-scatter CSV export parity with stable trial identity, typed outcome and explicit unavailable values; expose bounded accessible PyQt raw-row tables through one shared population helper; and split scalar scatter and distribution-matrix rendering into focused modules while preserving linked trial selection and all-trial arc analysis. |
-| 2026-08-11 | 1.14.23 | merge(rate-of-closure, #4282 #4283 #4285): use the already-merged PR #4283 head branch as the preserved dependency carrier for open draft PR #4285; normally merge exact remote/local capability-observer carrier child `b2baf6ef615b0d756a86ac4ca7eef2fc583210ee` first with exact published wind-workflow parent `3faf9e9493c9ba37168d9f65ab10cafeedc2a72f` second without reopening or retargeting a PR or rewriting history; reproduce the exact #4282 implementation while retaining the complete capability-observer history; keep publication, protected exact-head CI, and later #4285 propagation outside the bounded local audit. |
-| 2026-08-11 | 1.14.22 | merge(rate-of-closure, #4281 #4282): normally merge exact published responsive wind-workflow child `b2baf6ef615b0d756a86ac4ca7eef2fc583210ee` first with exact newly published wind scalar-adapter parent `9321c1d2e091b8c7e5a4a83aa9ad726290e7fb5a` second while preserving PR #4282's `feat/4199-wind-scalar-adapter` base and both histories; retain responsive PyQt6/React wind execution, progress/cancellation, controls, scatter, persistence/export, consolidated capability workflow, and strict campaign release authority while inheriting the current scalar-adapter, variation, workspace/toolstrip, launch-monitor/D-plane, split-kinetics, and Qt repairs; keep review, ordinary publication, protected exact-head CI, unresolved-thread checks, dependency integration, and release open. |
-| 2026-08-06 | 1.13.16 | feat(swing_sim, rate_of_closure, #4196): map desired flight to frame-explicit centered driver/iron delivery solution families with strict cross-runtime schemas, observed intervals/correlations, local sensitivities, complete residuals, model manifests, and rejected no-impact/miss diagnostics. |
-| 2026-08-06 | 1.13.15 | feat(swing_sim, rate_of_closure, #4195): add strict desired-flight inverse-solver contracts, deterministic bounded multi-objective search, ranked residual-rich candidates, typed infeasible/no-impact/nonconverged outcomes, and Python/TypeScript parity fixtures. |
-| 2026-08-06 | 1.13.14 | feat(swing_sim, rate_of_closure, #4194): add the canonical source-backed flight-result metric catalog, analytic landing/trajectory derivation, typed unavailable and qualified-ground boundaries, complete run manifests, deterministic Python/TypeScript exports, and cross-client parity fixtures. |
-| 2026-08-11 | 1.14.21 | merge/fix(rate-of-closure, #4281 #4282): normally merge exact published scalar-adapter parent `a7dce5f89b483303938f518b74b4028a1c68ba81` into exact remote responsive-workflow child `29e15d6ff631f7f30afcc745be783f2e716d7dcf` in child-first order; preserve the complete PyQt6/React wind workflow and capability stack while inheriting the parent's split kinetics and scalar-adapter size/parameter repairs; resolve obsolete monolithic kinetics wholly to the parent facade, retain both append-only documentation histories, apply pinned Ruff formatting only to seven surviving child-owned paths, and replace one skipped-import cache-return ignore with an explicit typed cast without changing behavior, APIs, schemas, frames, or units. |
-| 2026-08-10 | 1.14.16 | fix(rate-of-closure, #4281 #4282): normally propagate repaired wind scalar-adapter parent `d6fb04e07c2a625412e9208b07103acdc42c621b` into the responsive wind workflow without production/test conflict or history rewrite; retain current-state handoff policy and verify with 25 focused tests plus governance, size, and whitespace checks. |
-| 2026-08-10 | 1.14.15 | fix(rate-of-closure, #4282): preserve the strict campaign-manifest return contract under the repository's exact pinned-mypy skipped-import profile by explicitly typing the Pydantic-validated local as `CampaignManifest`; this repairs hosted quality-gate run `31395741841` without changing JSON parsing, validation, failure behavior, PR base, or dependency topology, and requires the resulting exact head to propagate normally into #4285. |
-| 2026-08-10 | 1.14.14 | feat/fix(rate-of-closure, #4199 #4281 #4282): normally propagate exact scalar-adapter parent `4a793c4c3f19aad43a3c215800b266be487ace49` into the responsive PyQt6/React wind-workflow carrier without changing its base; preserve the consolidated capability workflow and release authority while inheriting the complete variation, workspace, Python 3.10 compatibility, immutable scalar-row, stable identity, availability, and deterministic wind-scenario repairs; register the child capability tab in the parent's canonical navigation-state tuple so legacy saved orders reveal it without duplicating state; retain runtime invariant enforcement under optimized Python; and satisfy pinned typing/security analysis without weakening the fixed-argument builder boundary. |
-| 2026-08-10 | 1.14.13 | feat/fix(wind variation, #4144 #4199 #4280 #4281): normally propagate exact variation parent `8bcc49fc4e16e5e43be0b7f0f03c3017d5b79d0c` into the wind scalar-adapter child while preserving matching UI-neutral Python/TypeScript `scalar-ensemble/v1` contracts with structured provenance, unit-bearing variables, stable RFC3986 row IDs, immutable nullable raw rows, and overall/per-cohort availability; adapt actual and perfect-information wind outcomes with explicit status, true/estimated wind, launch/aim, target, landing, miss, cost, and information-delta values; enforce the deterministic scenario boundary identically in both clients with exact per-trial provenance, zero shear/turbulence, seed zero, and empty gusts; and retain the parent's result-local linked selection and complete visualization/workspace repairs. |
-| 2026-08-10 | 1.14.12 | feat/fix(variation, #4144 #4218 #4279 #4280): normally propagate exact workspace parent `6717e9e09d507dbc24bedb36177f1cdf0b4fd90b` into the variation-export child while preserving selected-scatter CSV parity with stable trial identity, typed unavailable outcomes, bounded accessible PyQt raw tables, focused PyQt/React visualization modules, linked selection, and all-trial arc analysis alongside the complete workspace/toolstrip, playback, plot, navigation, compatibility, and module-budget repairs; make linked trial selection result-local in both clients, clear it atomically when a rerun replaces the result, and reject or neutralize indices outside the current trial count so smaller reruns cannot crash or dim every new point. |
-| 2026-08-10 | 1.14.11 | feat/fix(rate-of-closure, #4203 #4218 #4279): normally propagate exact launch-registry parent `31cbc007d4c85b5479b7cd0fb0969124eab2af67` into the workspace/toolstrip child while preserving granular playback, path trails, module visibility, and independent plot controls; reuse the parent's single persisted impact-layer mapping and canonical navigation constants without duplicating state; and retain its focused triple-pendulum, plotting-catalog, and primary-navigation repairs. |
 | 2026-08-10 | 1.14.10 | fix(compatibility, #4279): make workspace UTC timestamp parsing deterministic across Python 3.10-3.12 with one anchored canonical grammar, consistent zero- through six-digit fractional-second parsing, and rejection of greater-than-microsecond precision instead of interpreter-dependent rejection or truncation. |
 | 2026-08-09 | 1.14.9 | fix(compatibility, #4279): route the child command/view `StrEnum` runtime imports and workspace-validation `UTC` import through `shared.python.compatibility` while preserving native enum typing under `TYPE_CHECKING`, all wire values, schemas, UTC serialization, and UI behavior; merge the parent and child regression into one nine-enum/two-UTC runtime-import contract exercised with real CPython 3.10.20. |
 | 2026-08-09 | 1.14.8 | feat(rate-of-closure, #4201): harden cross-runtime capability parsing and signed input; expose complete diagnostics, result exports, and quantitative scatter annotations; stabilize supported PyQt6/static-Vite entrypoints; and add a strict campaign manifest with generated schema, four-surface status, explicit gaps/dependencies, and fail-closed feature-stack/protected-parent/`main` release distinctions. |
 | 2026-08-08 | 1.14.7 | docs(rate-of-closure, #4197): record the capability stack consolidation onto `feat/4199-wind-workflow`, superseding per-slice stacked-branch cautions after #4283, #4289, and #4294 merged, while retaining the incomplete protected-release boundary. |
-| 2026-08-08 | 1.14.7 | docs(rate-of-closure, #4197): record the capability stack consolidation onto `feat/4199-wind-workflow`, superseding the per-slice stacked-branch cautions now that #4283, #4289, and #4294 are merged, and note that the unfinished CI matrix reflects an offline runner fleet rather than failures. |
 | 2026-08-08 | 1.14.6 | feat(rate-of-closure, #4197): add matched PyQt6/React Shot Optimizer workspaces with strict cross-runtime workflow persistence, off-main-thread qualified flight execution, truthful progress/cancellation, complete scalar-ensemble cohort retention, ranked alternatives, stage-qualified selectable scatter axes, zoom/autofit, bounded raw-row paging, lossless CSV/stable JSON, substantive hover guidance, and saved-layout migration that reveals newly registered modules. |
 | 2026-08-08 | 1.14.5 | feat(ball-flight, #4197): add profile/request-bound Python and TypeScript Waterloo/Penner capability evaluators with exact established parameter IDs, physical-domain and unit validation, sourced per-club spin defaults or paired variable spin inputs, shared gyro-projected positive-fade/right spin-axis convention across result/impact/variation producers, canonical target-frame conversion, target-aware complete scalar metrics, typed horizon nonconvergence, fail-fast invariants, and a shared 16-scalar tolerance fixture. Corrects the TypeScript ground-crossing interpolation so a descending launch from z=0 cannot extrapolate to a negative sample time, which previously raised an untyped `RangeError` that was absorbed as an `evaluator_exception`; such launches now report `nonconverged`, matching the Python runtime. |
-| 2026-08-11 | 1.14.21 | merge(rate-of-closure, #4280 #4281): normally merge exact published wind scalar-adapter child `a7dce5f89b483303938f518b74b4028a1c68ba81` first with exact newly published variation-export parent `e6c7460a01082631565fb9ed48aa32538bd7772c` second while preserving PR #4281's `feat/4144-variation-export-continuation` base and both histories; retain matching Python/TypeScript `scalar-ensemble/v1` provenance, availability, deterministic-scenario, wind-strategy adapter, and strict module-budget behavior while inheriting current variation visualization/export, workspace/toolstrip, launch-monitor/D-plane ancestry, split kinetics, and behavior-preserving Qt boundaries; keep review, ordinary publication, protected exact-head CI, unresolved-thread checks, dependency integration, and release open. |
-| 2026-08-11 | 1.14.20 | refactor(rate-of-closure, #4281): replace the private wind-scenario validator's five loose parameters with a scenario and one frozen, slotted expectation object; preserve validation order, messages, behavior, public API, and wire output while satisfying the four-parameter and 50-line production-function contracts. |
-| 2026-08-11 | 1.14.19 | refactor(rate-of-closure, #4281): split request/analysis validation from the 403-line wind-strategy plot adapter into a private collaborator, leaving 252-line and 186-line production modules below the mandatory 400-line ceiling while preserving the public API, scalar-ensemble wire order, row identity and values, availability, attributes, validation messages, and no-flight-rerun boundary. |
-| 2026-08-11 | 1.14.18 | merge(rate-of-closure, #4280 #4281): normally merge exact reviewed variation-export parent `3337945699966b63cb5cd8e52d7c3b194315e911` into exact remote wind scalar-adapter child `247046d55afcad3e6cd4f8029f854856c427f59c` with child-first parent order and no history rewrite; preserve matching Python/TypeScript `scalar-ensemble/v1` provenance, availability, deterministic-scenario, and wind-strategy plot-adapter behavior while inheriting the parent variation/export UI, kinetics split, Ground/Tee parity contracts, workspace/toolstrip repairs, and protected Ruff normalization; resolve the obsolete monolithic kinetics overlap to the validated parent `pendulum.sample(...)` facade, retain both append-only documentation histories, and require independent review plus fresh exact-head protected CI before publication. |
-| 2026-08-11 | 1.14.19 | merge(rate-of-closure, #4279 #4280): normally merge exact published variation-export child `3337945699966b63cb5cd8e52d7c3b194315e911` first with exact newly published workspace/toolstrip parent `efbca84095b617b4018732f7802c2da3f0525387` second while preserving PR #4280's `feat/4218-toolstrip-workspace` base and both histories; retain selected-scatter CSV parity, typed unavailable outcomes, bounded accessible tables, linked selection, and all-trial arc analysis while inheriting current workspace, launch-monitor/D-plane ancestry, split kinetics, and behavior-preserving Qt primitive-return boundaries; keep review, ordinary publication, protected exact-head CI, unresolved-thread checks, dependency integration, and release open. |
-| 2026-08-11 | 1.14.17 | merge(rate-of-closure, #4279 #4280): normally merge exact reviewed workspace/toolstrip parent `ccd0e026c580c93038fdf5c59d5d452a85ba27a0` into exact remote variation-export child `668ba96746f79f7a12e8092161bd610054197f58` with child-first parent order and no history rewrite; preserve selected-scatter CSV parity, typed unavailable outcomes, bounded accessible tables, linked selection, and all-trial arc analysis while inheriting the parent kinetics split, Ground/Tee parity contracts, and complete workspace/toolstrip behavior; resolve the obsolete monolithic kinetics overlap to the validated parent façade and normalize seven duplicate automation edits to exact protected Ruff 0.14.10 parent blobs while keeping their commit reachable; require independent review and fresh exact-head protected CI before publication. |
-| 2026-08-11 | 1.14.18 | merge(rate-of-closure, #4203 #4279): normally merge exact published workspace/toolstrip child `ccd0e026c580c93038fdf5c59d5d452a85ba27a0` first with exact newly published launch-registry parent `7abce9ad767fe8311da66a1e5998b892ea3ca9de` second while preserving PR #4279's `feat/4181-launch-monitor-registry` base and both histories; retain workspace, toolstrip, visibility, navigation, playback, and independent-plot behavior while inheriting current launch-monitor/D-plane ancestry, split kinetics, and behavior-preserving Qt primitive-return boundaries; keep review, ordinary publication, protected exact-head CI, unresolved-thread checks, dependency integration, and release open. |
-| 2026-08-11 | 1.14.16 | merge(rate-of-closure, #4279): ordinarily reconcile exact local workspace/toolstrip head `0b22c401a26c31441a599d8d9b39de123706e7ea` with divergent remote automation head `61fe2d556a5413e525d958612ccfd57e65b8d5a2`, preserving every commit and the existing stack topology; recognize 15 already-identical parent paths, normalize seven incompatible formatting edits back to protected Ruff 0.14.10 output, and resolve the sole obsolete pre-split kinetics conflict in favor of the current `pendulum.sample(...)` façade implementation; preserve workspace, toolstrip, visibility, navigation, playback, independent plots, physics, frames, units, schemas, and public contracts; require independent review and fresh exact-head protected CI before publication. |
-| 2026-08-11 | 1.14.15 | merge(rate-of-closure, #4203 #4279): normally propagate exact published launch-registry parent `3796b49e40b677fbac4e05739f8be49f905df2cb` into exact workspace/toolstrip child `7806a16f58e1c6999d32f0127a187fbb21f839a1` without rewriting stack topology; inherit only four static NumPy-array casts while preserving runtime arrays, physics, frames, units, public contracts, workspace behavior, and UI behavior; require fresh current-head protected CI and review. |
-| 2026-08-11 | 1.14.14 | merge(rate-of-closure, #4203 #4279): normally propagate exact launch-registry parent `0216a547aa79727091a2939b96e779e8ddbd7304` into the workspace/toolstrip child without rewriting stack topology; preserve the child's workspace, toolstrip, module-visibility, navigation, playback, and independent-plot behavior while inheriting the parent's identity-preserving kinetics split and pinned formatting repair; require fresh current-head protected CI and review. |
-| 2026-08-11 | 1.14.17 | docs(rate-of-closure, #4202 #4203): restore four exact append-only D-plane parent history rows omitted during the local current-parent reconciliation; preserve the candidate implementation, tests, topology, base, and quality evidence while keeping independent re-review, normal publication, protected exact-head CI, downstream propagation, and release open. |
-| 2026-08-11 | 1.14.16 | fix(rate-of-closure, #4203): close four exact-delta MyPy 1.13 Qt-stub boundaries by narrowing responsive-event handling, legend visibility, ball-setup event filtering, and visible status text to their already declared primitive return contracts; preserve values and UI behavior while retaining normal parent reconciliation and protected release gates. |
-| 2026-08-11 | 1.14.15 | merge(rate-of-closure, #4202 #4203): normally compose exact published launch-monitor-registry child `217e36dc93d30f79826847f958fbcd10805e58ed` with exact current D-plane parent `f3363aa88868f6a5c7e9ccfc682a9eca014e86c1`; retain the split typed kinetics facade at the sole formatting conflict, preserve parent behavior and history, and keep independent review, protected exact-head CI, downstream propagation, and release open. |
-| 2026-08-11 | 1.14.14 | fix(rate-of-closure, #4203): close four hosted MyPy `no-any-return` findings from the kinetics size split with explicit NumPy-array return narrowing at force-norm, RK4 concatenation, and app-frame projection boundaries; preserve exact runtime arrays, physics, frames, units, public contracts, and stack order. |
-| 2026-08-11 | 1.14.13 | refactor(rate-of-closure, #4203): split the 646-LOC swing-kinetics monolith into an identity-preserving 222-LOC public façade, 205-LOC pure-dynamics module, and 131-LOC immutable-series contract; preserve physics, frames, fixtures, UI behavior, and established imports while satisfying the changed-file 500-LOC gate. |
-| 2026-08-11 | 1.14.12 | style(rate-of-closure, #4203): apply repository-pinned Ruff 0.14.10 formatting to the eight files reported by current-head CI without changing physics, behavior, public contracts, schemas, UI layout, or stack order. |
-| 2026-08-10 | 1.14.14 | feat/fix(wind variation, #4144 #4199 #4280 #4281): normally propagate repaired variation parent `b90e5021a59e2081415b51ef29fbed06377bc201` into the wind scalar-adapter child without production/test conflict or history rewrite; preserve matching Python/TypeScript `scalar-ensemble/v1` provenance, availability, deterministic-scenario, and wind-strategy plot-adapter contracts; verify the reconciliation with 25 focused D-plane/impact tests and governance, size, and whitespace gates. |
-| 2026-08-10 | 1.14.13 | feat/fix(variation, #4144 #4218 #4279 #4280): normally propagate repaired workspace parent `61b7f48b5aeb7d57246b4963da3df086e79cbe15` into the variation-export child without feature-code conflict or history rewrite; preserve selected-scatter CSV parity, typed unavailable outcomes, bounded accessible tables, linked selection, all-trial arc analysis, and the complete workspace/toolstrip behavior; verify the reconciliation with 25 focused D-plane/impact tests and governance, size, and whitespace gates. |
-| 2026-08-10 | 1.14.12 | fix(rate-of-closure, d-plane, #4202 #4203 #4279): normally propagate exact repaired launch-registry parent `12dd76a8dbcc106c4683f2f2e53076f8dc6f1b76` into the workspace/toolstrip child without rewriting the stack; inherit explicit ndarray result boundaries while preserving numerical semantics, frames, schemas, and UI behavior; verify the reconciled tree with 25 focused D-plane/impact tests and governance, size, and whitespace gates. |
-| 2026-08-10 | 1.14.11 | feat/fix(rate-of-closure, #4203 #4218 #4279): normally propagate exact launch-registry parent `31cbc007d4c85b5479b7cd0fb0969124eab2af67` into the workspace/toolstrip child while preserving granular playback, path trails, module visibility, and independent plot controls; reuse the parent's single persisted impact-layer mapping and canonical navigation constants without duplicating state; and retain its focused triple-pendulum, plotting-catalog, and primary-navigation repairs. |
-| 2026-08-10 | 1.14.10 | fix(compatibility, #4279): make workspace UTC timestamp parsing deterministic across Python 3.10-3.12 with one anchored canonical grammar, consistent zero- through six-digit fractional-second parsing, and rejection of greater-than-microsecond precision instead of interpreter-dependent rejection or truncation. |
-| 2026-08-09 | 1.14.9 | fix(compatibility, #4279): route the child command/view `StrEnum` runtime imports and workspace-validation `UTC` import through `shared.python.compatibility` while preserving native enum typing under `TYPE_CHECKING`, all wire values, schemas, UTC serialization, and UI behavior; merge the parent and child regression into one nine-enum/two-UTC runtime-import contract exercised with real CPython 3.10.20. |
-| 2026-08-09 | 1.14.8 | feat(rate_of_closure, #4218 #4279): add a UI-neutral File/View/Tools command registry, strict atomic workspace documents, matched PyQt6/React top toolstrips, persistent module visibility and order, direct Impact/Swing/Flight navigation, deterministic replay/loop/speed controls, and independent per-plot canvases with zoom, Auto Fit, and movable or hidden legends. Propagate exact launch-registry parent `08a2fdd8ce6bbc8fbb8f121927a677d4addb6b11` normally while retaining its Linux-safe facade and Python 3.10 compatibility contracts, and type the Qt legend-visibility boundary explicitly for the pinned changed-file mypy gate. |
-| 2026-08-10 | 1.14.11 | test/fix(rate-of-closure, #4143 #4202 #4203 #4325): normally propagate repaired launch-registry parent `12dd76a8dbcc106c4683f2f2e53076f8dc6f1b76` into the shared Ground/Tee parity and rendered-evidence child without production/test-code conflict or history rewrite. |
-| 2026-08-10 | 1.14.10 | test(rate-of-closure, #4143): record deterministic Ground/Tee visual evidence through semantic Playwright checks and a hidden-window PyQt capture regression, retaining screenshots as external digested artifacts instead of brittle pixel baselines or repository binaries. |
-| 2026-08-10 | 1.14.9 | test(rate-of-closure, #4143): add one strict versioned SI golden fixture consumed by Python and React to pin Ground/Tee defaults, explicit overrides, physical height and center geometry, serialization, invalid finite-domain handling, and backward-compatible legacy migration without changing production behavior. |
-| 2026-08-10 | 1.14.8 | feat/fix/refactor(rate-of-closure, #4202 #4203): propagate the exact D-plane parent into the launch-monitor registry without rewriting the stack; preserve the responsive `SimulationViewControlsMixin` architecture while making `ImpactLayerControls` the single owner of persisted D-plane checkbox state; retain the existing automation compatibility seam as an identity alias; and repair the original child's three ungrandfathered size blockers through identity-preserving extractions for triple-pendulum dynamics, immutable plotting metadata, and versioned primary-navigation state. |
-| 2026-08-09 | 1.14.7 | fix(compatibility, #4203): route the PyQt torque-profile controller's UTC constant through the shared Python 3.10 compatibility module, preserving UTC timestamp serialization and workspace behavior while removing the remaining parent-owned `datetime.UTC` collection boundary. |
-| 2026-08-09 | 1.14.6 | fix(compatibility, #4203): route seven Rate/shared swing string-enum runtime imports through the existing Python 3.10 compatibility contract while retaining native enum typing under `TYPE_CHECKING`; preserve all wire values, schemas, physics, and UI behavior, and add a source-level regression exercised with real CPython 3.10.20. |
-| 2026-08-09 | 1.14.5 | fix(ci, #4203): keep the in-package swing flight and solver facade-contract tests in pytest's active package namespace by using relative imports, preventing editable Linux collection from crossing between `src.shared...` and `shared...` before assertions while leaving production APIs and physics unchanged. |
 | 2026-08-07 | 1.14.4 | feat(rate_of_closure, #4206): add validated manual reference AoA/path, targetward forward shaft lean, and tracked-reference versus registered generated-hosel shaft datums in PyQt6 and React; rotate pose, angular-rate components, and delivered face normals consistently; persist schema-v5 `manual_delivery`; export the reference-contact/reference-impact model boundary; and pin the representative Pitching Wedge decomposition in both runtimes. |
 | 2026-08-07 | 1.14.3 | fix(variation, #4144): preserve trial identity when filtering finite landing coordinates so carry/lateral values from different incomplete trials cannot form fictitious points; use one paired-row contract for Python/TypeScript ellipse analysis, PyQt/React rendering, and exact plotted-count status. Clarify that the wedge kernel's 20 mm example is synthetic and separately pin the generated Pitching Wedge face-center/hosel cross-check and current UI-state limitations. |
 | 2026-08-06 | 1.14.2 | feat(rate_of_closure, #4192 #4234): complete the shared spatial-target workflows in PyQt6 and React with canonical cross-tab state, versioned JSON/CSV/manifest persistence, no-run 2D/3D rendering, continuous aerial passage, surface-projected landing assessment, field-linked validation, stale-solver protection, high-DPI canvases, responsive wrapped PyQt forms, collapsible engineering detail/layer controls, and movable or hideable legends; keep aerial requests fail-closed where solver/variation objectives remain ground-only. |
@@ -3170,16 +2981,8 @@ Active development with stable core, continuous tool expansion, and web API in p
 | 2026-08-06 | 1.13.13 | fix(ball-flight, #4199): migrate strategy output to v2; separate policy-fixed true-wind counterfactuals from preset-oracle regret and add failure-inclusive target-hold, miss-distance CVaR, and directional risk metrics with Python/TypeScript parity. |
 | 2026-08-06 | 1.13.12 | feat(rate_of_closure, #4200): add deterministic timestamp interpolation and accessible play/pause/scrub/speed/restart/Launch/Apex/Landing controls to PyQt6 and React; preserve Matplotlib camera state with mutable markers; add a dependency-free rotatable/zoomable orthographic web canvas with a locked physical metre scale and one cancellable animation loop. |
 | 2026-08-06 | 1.13.9 | feat(rate_of_closure, swing_sim, #4182 #4183 #4189): add typed reference-frame-explicit 3D D-plane geometry, face-center/contact/reference analyses, exact-versus-planar spin-loft residuals, persistent PyQt6/React engineering layers, and shaded sector exports. |
-| 2026-08-06 | 1.13.11 | feat(rate_of_closure, #4200): add deterministic timestamp interpolation and accessible play/pause/scrub/speed/restart/Launch/Apex/Landing controls to PyQt6 and React; preserve Matplotlib camera state with mutable markers; add a dependency-free rotatable/zoomable orthographic web canvas with a locked physical metre scale and one cancellable animation loop. |
-| 2026-08-10 | 1.13.11 | fix(swing-sim, d-plane, #4202): add explicit ndarray result boundaries to the private vector conversion and horizontal-projection helpers, closing the exact changed-file MyPy `no-any-return` failures without changing numerical semantics, DbC validation, frames, schemas, or UI behavior. |
-| 2026-08-10 | 1.13.10 | feat/fix(rate-of-closure, swing-sim, #4179 #4182 #4183 #4189 #4202): retain typed reference-frame-explicit 3D D-plane geometry, face-center/contact/reference analyses, exact-versus-planar spin-loft residuals, persistent PyQt6/React engineering layers, and shaded sector exports while propagating the Python 3.10 UTC repair and source-wide AST guard through the exact impact-visualization parent; extract persisted D-plane layer controls to restore the protected simulation-view module budget. |
-| 2026-08-06 | 1.13.9 | feat(rate_of_closure, swing_sim, #4182 #4183 #4189): add typed reference-frame-explicit 3D D-plane geometry, face-center/contact/reference analyses, exact-versus-planar spin-loft residuals, persistent PyQt6/React engineering layers, and shaded sector exports. |
-| 2026-08-10 | 1.13.9 | feat/fix(rate-of-closure, #4162 #4167 #4173 #4174 #4178 #4179): retain exact-event pose/twist/wrist interpolation, the versioned impact-scene contract, locked-scale PyQt6 and React views, named cameras, accessible metrics, and PNG/SVG/JSON exports while propagating the Python 3.10 UTC repair and source-wide AST guard through the exact turf-physics parent. |
-| 2026-08-10 | 1.13.8 | feat/fix(golf-club, rate-of-closure, #4166 #4167 #4173 #4174 #4178): retain the passive provenance-gated turf proxy, nine-point wedge contact wrench, strict profile persistence, convergence diagnostics, and explicit force-coupling boundary while propagating the Python 3.10 UTC repair and source-wide AST guard through the exact stacked parent. |
 | 2026-08-05 | 1.13.7 | feat(rate_of_closure, #4162): add exact-event pose/twist/wrist interpolation; a versioned impact-scene contract; locked-scale orbitable wedge, shaft, ball, contact, orientation, screw-axis, and velocity-decomposition views in PyQt6 and React; named cameras, accessible metric definitions, and PNG/SVG/JSON exports. |
-| 2026-08-10 | 1.13.7 | fix(rate-of-closure, #4167 #4173 #4174): propagate the Python 3.10 UTC compatibility repair and source-wide AST guard through the impact-inspector parent into swept wedge ground clearance without rewriting the stacked child or changing its ground-contact contracts. |
 | 2026-08-06 | 1.13.6 | feat(golf-club, rate_of_closure, #4166): add a passive, provenance-gated compliant turf proxy; nine-point wedge contact wrench; strict profile persistence; cancellation and refinement diagnostics; and a retained-Rate adapter with explicit force-coupling limitations. |
-| 2026-08-10 | 1.13.6 | fix(rate-of-closure, #4167 #4173): propagate the Python 3.10 UTC compatibility repair into the impact-inspector child without rewriting its history; torque-profile persistence now uses the shared compatibility export and a source-wide AST guard prevents direct, aliased, or module-attribute `datetime.UTC` regressions. |
 | 2026-08-06 | 1.13.5 | refactor(gui, ci): deduplicate Rotation Converter plot helpers and extract Movement Optimizer motion helpers, restoring the protected module-size budget inherited by the stacked Rate PRs. |
 | 2026-08-05 | 1.13.4 | feat(rate_of_closure, golf-club, #4158 #4160 #4163): integrate frame-explicit wedge contact/shaft kinematics into retained Rate runs; add honest impact-or-closest-approach jump controls and engineering readouts to PyQt6 and React; restore manual web angular velocity; and select the documented 30 ms square pose for flat automatic speed plateaus. |
 | 2026-08-05 | 1.5.5 | fix(ci, #4155): make the Python tool-cache guard inspect `/opt/hostedtoolcache` and optionally require the interpreter's declared link library; run that stronger semantic preflight immediately before the Rust/PyO3 job provisions Python, with Linux fixture and workflow-order contracts. |
diff --git a/docs/development/RATE_OF_CLOSURE_CAMPAIGN_HANDOFF.md b/docs/development/RATE_OF_CLOSURE_CAMPAIGN_HANDOFF.md
index 9a6195ba7..e3ae4f788 100644
--- a/docs/development/RATE_OF_CLOSURE_CAMPAIGN_HANDOFF.md
+++ b/docs/development/RATE_OF_CLOSURE_CAMPAIGN_HANDOFF.md
@@ -1,5 +1,38 @@
 # Rate of Closure Ball-Flight Campaign Handoff
 
+## 2026-08-11 PR #4342 current-parent propagation candidate
+
+- PR `#4342` retains branch `feat/4274-regional-plan-io` and base
+  `feat/4274-regional-surface-ui`. Exact current child
+  `c1f47f2ef68b3db102da5416aaac17a40f675207` is first and exact reviewed
+  local #4339 candidate `db335937afc4b587d235eb705e315f577519c5e6` is
+  second in a normal no-ff merge.
+- Child-owned canonical request import/export, bounded UTF-8, native atomic
+  save, browser-qualified download, tests, and limitations remain intact while
+  inheriting current editor, wire, regional-physics, and complete ground
+  ancestry.
+- The default shared canonical encoder still rejects floats and integers beyond
+  JavaScript's safe range. The capability-observation facade alone selects a
+  separately named extended finite-float policy that reuses the shared
+  recursion, keeps integers bounded, emits exact exponent-free `1e20` and
+  `1e21` tokens matching TypeScript, and rejects non-finite values.
+- Local qualification passes all `909` Rate-of-Closure Python tests, all `110`
+  React files / `686` tests, the complete Cargo workspace, `47` focused Python
+  compatibility/regional-I/O tests, and `12` focused React capability tests.
+  Pinned Ruff 0.14.10 check/format passes `17` changed Python files; pinned
+  MyPy 1.13 and Bandit pass `12` changed production files. TypeScript,
+  zero-warning ESLint, the 199-module Vite build, Rust format and
+  warning-denied clippy, both changed-file LOC gates, manifest/docs/blocking-
+  gate/assertion/minimum-test governance, marker scans, and diff checks pass.
+  One untouched manual-delivery UI test timed out in the first concurrent full
+  run, then passed alone and in the single complete rerun.
+- The candidate has not been rebased, retargeted, rewritten, force-pushed, or
+  published and remains `not_released`.
+- Execution/playback, result interchange, measured calibration, model-input
+  persistence, changing geometry or velocity, TypeScript/compiled regional
+  physics, downstream parity, protected exact-head evidence, review, approval,
+  dependency integration, and release remain open.
+
 ## 2026-08-11 PR #4339 current-parent propagation candidate
 
 - PR `#4339` retains branch `feat/4274-regional-surface-ui` and base
@@ -29,41 +62,217 @@
   physics, downstream parity, protected exact-head evidence, review, approval,
   dependency integration, and release remain open.
 
+
+## 2026-08-11 PR #4342 append-only preservation repair
+
+Independent audit of local merge `e7fedfb18de1550eed3484ed2fc99d0baaecdca1`
+found that three older parent ancestry sections had been summarized instead of
+retained verbatim. This docs-only follow-up restores the full parent text below
+without changing production code, tests, schemas, manifest state, PR base, or
+release claims. The exact parent sections and SPEC rows are now governed by
+byte-for-byte comparison against
+`d21741e312b849a63f73cabf351a15d9de80fb94`.
+
+## 2026-08-11 PR #4335 current-parent ancestry candidate
+
+The clean dedicated `feat/4271-regional-wire-contract` worktree starts from
+exact live PR #4335 child `74a053d2d544da9f44a88007660ad28c0127f285`
+and normally merges exact newly published PR #4332 parent
+`04ccf08dd990de1cd056a3420e67772773a4be2e` as its second parent. PR #4335
+keeps base `feat/4271-regional-surface-transitions`; neither branch is rebased,
+retargeted, rewritten, force-pushed, or published by this reconciliation.
+Production physics, wire contracts, golden fixtures, numerical ordering, and
+public APIs merge byte-exactly; only SPEC, manifest, and handoff records require
+truthful reconciliation.
+
+The child retains the strict cross-runtime regional-plan request/result wire
+contract, canonical JSON/SHA-256 evidence, fail-closed parsing, and Python
+resolver adapter. The parent retains bounded coplanar regional transitions and
+the complete reconciled impact/bounce/skid/roll ancestry. Changing normals,
+height or surface-velocity discontinuities, terrain deformation, torsional-spin
+damping, roll-to-skid transitions, internal transition-ledger export, regional
+UI, compiled or TypeScript regional physics, downstream parity, protected CI,
+review, normal stack integration, and main release remain open. This local
+merge is not release evidence and requires independent review before an
+ordinary fast-forward publication.
+
+Merged-tree qualification is `132` focused Python ground tests and `9`
+focused React regional/ground-contract tests passing. Pinned Ruff 0.14.10 check
+and format pass all `45` ground Python files. Pinned MyPy 1.13 passes the exact
+two-file isolated CI boundary and all `31` production modules under the coherent
+whole-package profile that disables only redundant-cast warnings; those casts
+remain required by the isolated protected profile. Bandit reports no
+medium/high finding. Documentation governance, manifest validation/layout and
+all `8` manifest contracts, changed-production and minimum-test policies, the
+official 500-LOC PR-delta gate (`6` files, zero violations), diff checks, and
+mandatory maxima of `392` module lines, `46` function lines, and `4` parameters
+all pass.
+
+## 2026-08-11 PR #4332 current-parent ancestry candidate
+
+The clean dedicated `feat/4271-regional-surface-transitions` worktree starts
+from exact live PR #4332 child `1a48d749af508843fac2a5102f4dd56294429bda`
+and normally merges exact newly published `feat/4271-ground-skid-roll` parent
+`0ea6740965068542e9d8c7449e06ec07d88969e0` as its second parent. PR #4332
+keeps base `feat/4271-ground-skid-roll`; neither branch is rebased, retargeted,
+rewritten, force-pushed, or published by this reconciliation. The only textual
+merge conflict is the independent SPEC-version collision. SPEC 1.14.27 retains
+the regional child record after the parent's 1.14.26 entry; production physics,
+contracts, schemas, numerical ordering, and public APIs merge without conflict.
+
+The child retains bounded coplanar material overlays, explicit precedence,
+exact quadratic boundary splitting, base-edge precedence, state and energy
+continuity, strict `surface_transition` evidence, request-bound transition
+limits, and randomized piecewise-analytic coverage. The current parent retains
+its reviewed impact, bounce, skid, roll, resistance, qualified-rest, edge,
+composition, and passive-ledger behavior. Changing normals, height or
+surface-velocity discontinuities, terrain deformation, torsional-spin damping,
+roll-to-skid transitions, regional UI, compiled regional physics, downstream
+parity, protected CI, review, normal stack integration, and main release remain
+open. This local merge is not release evidence and requires independent review
+before an ordinary fast-forward publication.
+
+Merged-tree qualification is `121` focused Python ground/regional tests and
+`5` focused React ground-contract tests passing. Pinned Ruff 0.14.10 check and
+format pass all `41` ground Python files; pinned MyPy 1.13 passes all `28`
+ground production modules; and Bandit reports no medium/high finding. The
+campaign manifest validator and all `8` manifest contracts, documentation
+governance, changed-production policy, minimum test contract, both parent and
+child diff checks, and the official 500-LOC changed-file gate (`14` files,
+zero violations) pass. Production maxima are `392` lines per module, `46`
+lines per function, and `4` parameters excluding `self`/`cls`. The manifest
+now records open PR #4332 at its still-published child head and open PR #4304
+at its exact newly published parent head; neither record claims protected or
+main-release evidence.
+
+## 2026-08-11 PR #4304 current-parent ancestry candidate
+
+The clean dedicated `feat/4271-ground-skid-roll` worktree first fast-forwarded
+to exact live PR #4304 child `52d9a6a978d8e6b8b19ef92f02f265c9058b00ad`,
+then normally merged exact live `feat/4270-ground-impact-bounce` parent
+`cf6e72bad98e5f36f782254942e6895b8b71e670`. The PR base remains unchanged;
+neither branch was rebased, retargeted, rewritten, force-pushed, or published.
+The automatic merge was conflict-free. Its only child-tree change before this
+documentation reconciliation is deterministic formatting in the existing
+skid/roll regression test; production physics, contracts, schemas, numerical
+ordering, and public APIs remain unchanged.
+
+Local qualification on the merge candidate is `115` focused ground
+tests passing on CPython 3.13, pinned Ruff 0.14.10 clean across the ground
+package and tests, and pinned MyPy 1.13 clean across all `25` ground production
+modules. The manifest validator and all `8` manifest tests, documentation
+governance, exact-parent and exact-child diff checks, changed-production policy,
+the official changed-file 500-LOC budget, and mandatory production limits of
+400 lines per module, 50 lines per function, and four parameters all pass. The
+local merge is a PUBLISH candidate for independent review only; protected
+exact-head CI, approval, normal child propagation, and publication remain open.
+
+
+## 2026-08-11 PR #4342 current-parent reconciliation candidate
+
+- The clean `feat/4274-regional-plan-io` worktree normally merges exact
+  published PR #4342 child `8e1c7ccd99a7c4886c5fb9ccc7e4d94a6d7e3833`
+  with exact newly published PR #4339 parent
+  `d21741e312b849a63f73cabf351a15d9de80fb94` as its second parent.
+- PR #4342 retains base `feat/4274-regional-surface-ui`; no rebase, retarget,
+  rewrite, force-push, publication, protected-evidence, or release claim is
+  made by this reconciliation.
+- Strict regional request import/export is preserved. The child widget module
+  delegates canonical-precision controls to the parent's frozen
+  `NumberInputSpec` and three-parameter helper while retaining inclusive safe
+  number bounds and eleven-decimal presentation.
+- Merged-tree qualification is 53 focused Python/PyQt/shared-ground tests and
+  14 focused React tests passing. Ruff, MyPy, Bandit, TypeScript, zero-warning
+  ESLint, the 199-module production build, campaign manifest and eight manifest
+  tests, documentation/tool-manifest governance, changed-Python, minimum-test,
+  exact-diff assertion, 500-LOC, parent/child diff, and whitespace gates pass.
+  Structural maxima are 396 module lines, 40 function lines, and three
+  parameters.
+- Independent review, ordinary fast-forward publication, fresh protected
+  CI/approval, dependency integration, and release remain required.
+
 ## 2026-08-11 PR #4339 structural helper repair candidate
 
-This local follow-up descends exact independently reviewed reconciliation
-candidate `eedad6d23b517eb4c99d4ba9000ff6555101099f`. Independent review found
-one publication blocker: the regional PyQt number-input helper accepted six
-parameters, above the campaign's mandatory four-parameter limit. RED-first
-tests now bind the helper's signature and all observable widget settings. A
-frozen `NumberInputSpec` validates finite positive step and finite ordered
-bounds before widget construction; the three-parameter helper preserves field
-names, values, units, ranges, increments, tooltips, and presentation order.
-
-Eight focused Python/PyQt tests and 25 focused React tests pass; the complete
-React suite passes all 108 files / 672 tests. Pinned Ruff check/format, MyPy on
-all six affected production modules, Bandit medium/high screening, TypeScript,
-zero-warning ESLint, documentation/manifest governance, changed-production,
-test-assertion, minimum-test, 500-LOC, and diff gates pass. Mandatory maxima
-are 400 module lines, 42 function lines, and three parameters. A broader
-Rate-of-Closure Python run reached the ten-minute workstation lifecycle bound
-without yielding an assertion failure; it is not counted as passing evidence.
-The monorepo-wide lane separately exposed unrelated baseline failures and
-xdist worker termination. No schema, digest, regional physics, UI behavior,
-PR base, protected evidence, or release boundary changes.
+- Parent head `d21741e312b849a63f73cabf351a15d9de80fb94` replaces the regional
+  PyQt six-parameter input helper with a frozen validated `NumberInputSpec`
+  and three-parameter helper without changing field behavior or ordering.
+- Eight focused Python/PyQt tests, 25 focused React tests, and the complete 672
+  React tests pass with Ruff, MyPy, Bandit, TypeScript, ESLint, build,
+  documentation, manifest, size, structural, and diff gates clean.
+- No schema, digest, regional physics, PR base, protected evidence, or release
+  boundary changes.
 
 ## 2026-08-11 PR #4339 current-parent ancestry candidate
 
-The clean `feat/4274-regional-surface-ui` worktree starts from exact published
-PR #4339 child `cbb9c0a6bdc6a50f59f7a661139b9d53e1892980` and normally merges
-exact newly published PR #4335 parent
-`9e01ccc3e891cc45907293751a192624195a77a5` as its second parent. The PR
-continues to target `feat/4271-regional-wire-contract`; neither branch is
-rebased, retargeted, rewritten, or force-pushed. Feature payloads merge without
-conflict; only SPEC, manifest, and handoff records require truthful
-reconciliation. Independent review, ordinary fast-forward publication, fresh
-protected exact-head CI, approval, dependency integration, and release remain
-open.
+- The parent worktree normally merged exact editor child
+  `cbb9c0a6bdc6a50f59f7a661139b9d53e1892980` with exact published #4335
+  parent `9e01ccc3e891cc45907293751a192624195a77a5` while retaining the
+  `feat/4271-regional-wire-contract` base.
+- UI, wire, resolver, regional physics, and ground ancestry were preserved;
+  protected exact-head review, integration, and release remained open.
+
+## 2026-08-11 PR #4342 delta-MyPy follow-up
+
+- Protected CI on exact head
+  `cffe349ac0a8054f1d168cb36684fd00bc5f8a49` rejected one redundant Boolean
+  cast in the regional atomic-write adapter.
+- The cast is removed and the typed helper's direct return is unchanged.
+  Focused persistence tests and the CI-equivalent pinned MyPy command pass.
+- No wire, validation, file, UI, or physics behavior changes. Fresh protected
+  CI/review and dependency ordering remain open.
+
+## 2026-08-11 regional request I/O protected publication
+
+- Branch `feat/4274-regional-plan-io` is published normally as draft PR
+  [#4342](https://github.com/D-sorganization/Tools/pull/4342), targeting exact
+  PR #4339 branch `feat/4274-regional-surface-ui` at parent head
+  `cbb9c0a6bdc6a50f59f7a661139b9d53e1892980`.
+- Reviewed implementation head
+  `d748e7a5ef3da5e6ce7737ff6829e0f14665fe97` includes canonical PyQt6/React
+  request import/export, safe-number parity, and bounded strict UTF-8 native
+  reads. Publication documentation commits change no runtime code.
+- Protected CI, review, #4274, dependency ordering, integration, and release
+  remain open.
+
+## 2026-08-11 bounded regional request read follow-up
+
+Independent review of local safe-number commit
+`10c394f6b1fd2927e7f3b1f96cc097cae6bfd380` found that native request import
+checked file size with `stat()` and then performed a separate unbounded text
+read. A concurrently grown or replaced file could therefore allocate beyond
+the 1 MiB wire cap before the strict parser rejected it.
+
+Native import now opens one binary handle, reads at most the wire cap plus one
+sentinel byte, rejects overflow, strictly decodes UTF-8, and only then delegates
+to the unchanged canonical v1 parser. Tests simulate content growth after the
+metadata check and reject invalid UTF-8 explicitly. The complete regional and
+atomic persistence set is 28 tests passing. This changes no schema, digest,
+physics, browser behavior, or write semantics. Static and governance evidence
+is recorded in this commit: Ruff, Ruff format, Black, focused MyPy, manifest,
+module-size, documentation-governance, and diff checks pass. The child remains
+local and unprotected.
+
+## 2026-08-11 regional request I/O safe-number follow-up
+
+Independent review of local commit `e39edf4b50b1fb9811b0032bec4758c7a08c9b74`
+found two cross-runtime representation gaps. The PyQt6 precedence spin box
+silently narrowed otherwise valid v1 integers above 1,000,000, and Python
+accepted integer-valued wire numbers above JavaScript's exact safe range.
+
+The native editor now uses an exact decimal integer entry for precedence and
+preserves every v1 value from zero through 9,007,199,254,740,991 without a
+floating-point conversion. Shared Python canonical-number and ground-record
+validation now reject magnitudes outside that same range, matching React while
+retaining all finite fractional values inside it. PyQt6 floating-point editors
+publish the same bounds. Regression coverage pins exact maximum-precedence
+open/save round-trip and matched native/browser refusal of unsafe material
+numbers. This changes no schema version, provenance algorithm, physics, or
+browser filesystem limitation. Final local evidence is 132 shared ground tests,
+21 regional editor/I/O tests, five atomic-workspace tests, 23 focused React
+tests, the whole-shell tooltip sweep, and eight manifest tests passing. Ruff,
+Ruff format, Black, focused MyPy, TypeScript, zero-warning ESLint, the 199-module
+production build, module and changed-file size budgets, documentation
+governance, and diff checks pass. The child remains local and unprotected.
 
 ## 2026-08-10 PR #4339 stale-validation invalidation follow-up
 
@@ -81,6 +290,43 @@ and diff checks are clean. The known polynomial-generator empty-legend warning
 is unrelated. This changes no wire schema, provenance digest, physics, or
 persistence boundary; fresh protected CI is still required before merge.
 
+## 2026-08-10 issue #4274 canonical regional request I/O local child
+
+Local branch `feat/4274-regional-plan-io` started directly from published PR
+#4339 head `a9ace5052bcd54b78f79b62d5d9ac26debedb4b1`, then fast-forwarded
+normally to corrected exact parent `cbb9c0a6bdc6a50f59f7a661139b9d53e1892980`.
+The parent's stale-validation invalidation and tooltip follow-up are preserved.
+This child has not been pushed, opened as a PR, reviewed, merged, or released.
+
+The matched Ground Surfaces editors now import and export the canonical
+`ground-regional-material-plan-request/v1` document. PyQt6 uses native Open
+and Save As dialogs plus a shared flush/fsync/atomic-replace UTF-8 writer;
+cancel is a no-op, failed reads and writes preserve the last-known-good editor
+or file, and the recent path advances only after success. React validates a
+bounded browser File before committing state and downloads exact canonical
+bytes while revoking its object URL. Browser filesystem limitations are
+explicit, and workspace persistence remains separate.
+
+Import accepts only editor producer/provider v1 evidence, the fixed target
+frame origin/downrange axis, and one to eight rows. It never relabels external
+evidence. An unchanged import round-trips the exact request and provenance;
+after any edit, validation binds a fresh draft digest. No measured calibration
+is synthesized and neither path executes physics.
+
+RED-first coverage now includes exact-byte round-trip, native cancel and
+atomic rollback, corruption, duplicate keys, oversize input, unsupported
+producer/axis qualification, stale-provenance rebinding, transactional PyQt6
+population, browser rollback, URL cleanup, and pre-allocation size rejection.
+Focused evidence is 25 Python/PyQt tests and 12 React tests passing. With
+adjacent wire and manifest coverage, 44 Python and 16 React tests pass.
+Focused MyPy, Ruff, TypeScript, zero-warning ESLint, the 199-module production
+build, release-manifest validation, and module budgets are clean.
+
+Issue #4274 remains open. Measured calibration, workspace model-input
+persistence, execution/playback, result interchange, visualization, changing
+geometry, compiled regional physics, downstream parity, protected CI/review,
+integration of this child into the protected stack, and release remain explicit
+gaps.
 ## 2026-08-10 issue #4274 matched regional surface editor local child
 
 Local branch `feat/4274-regional-surface-ui` started from published PR #4335
@@ -125,63 +371,7 @@ terrain/interval visualization, TypeScript or compiled regional physics,
 UpstreamDrift parity, protected CI/review, parent integration, and main release
 remain open.
 
-## 2026-08-11 Current regional-surface parent propagated into regional wire
-
-- Open draft PR `#4335` retains branch `feat/4271-regional-wire-contract` and
-  base `feat/4271-regional-surface-transitions`. Exact remote regional-wire
-  child `9e01ccc3e891cc45907293751a192624195a77a5` is merged first and exact current
-  regional-surface parent `5ac3c1f233de30dd42eeac57073cbdb149bde652`
-  second by an ordinary no-ff merge.
-- Strict Python/TypeScript regional-plan schemas, canonical JSON/SHA-256
-  fixture evidence, fail-closed parsing, Python resolver binding, tests,
-  contracts, limitations, and both append-only histories remain intact.
-- Focused local qualification passes all `132` Python ground/regional-wire
-  tests, `9` focused React regional/ground-contract tests, and all `137`
-  `tools-core` Rust tests. TypeScript, zero-warning ESLint, the 190-module Vite
-  build, Rust format and warning-denied clippy, pinned Ruff 0.14.10 across six
-  changed Python files, pinned MyPy 1.13 and Bandit medium/high across four
-  production files, the 500-LOC changed-file gate, manifest parsing plus eight
-  contracts, documentation governance, exact wire-tree preservation,
-  history/marker, and diff checks pass.
-- The candidate is unpublished and `not_released`; changing geometry or
-  velocity, internal transition-ledger export, regional UI, compiled or
-  TypeScript regional physics, UpstreamDrift parity, protected exact-head CI,
-  review, approval, dependency integration, and release remain open.
-
-## 2026-08-11 PR #4335 current-parent ancestry candidate
 
-The clean dedicated `feat/4271-regional-wire-contract` worktree starts from
-exact live PR #4335 child `74a053d2d544da9f44a88007660ad28c0127f285`
-and normally merges exact newly published PR #4332 parent
-`04ccf08dd990de1cd056a3420e67772773a4be2e` as its second parent. PR #4335
-keeps base `feat/4271-regional-surface-transitions`; neither branch is rebased,
-retargeted, rewritten, force-pushed, or published by this reconciliation.
-Production physics, wire contracts, golden fixtures, numerical ordering, and
-public APIs merge byte-exactly; only SPEC, manifest, and handoff records require
-truthful reconciliation.
-
-The child retains the strict cross-runtime regional-plan request/result wire
-contract, canonical JSON/SHA-256 evidence, fail-closed parsing, and Python
-resolver adapter. The parent retains bounded coplanar regional transitions and
-the complete reconciled impact/bounce/skid/roll ancestry. Changing normals,
-height or surface-velocity discontinuities, terrain deformation, torsional-spin
-damping, roll-to-skid transitions, internal transition-ledger export, regional
-UI, compiled or TypeScript regional physics, downstream parity, protected CI,
-review, normal stack integration, and main release remain open. This local
-merge is not release evidence and requires independent review before an
-ordinary fast-forward publication.
-
-Merged-tree qualification is `132` focused Python ground tests and `9`
-focused React regional/ground-contract tests passing. Pinned Ruff 0.14.10 check
-and format pass all `45` ground Python files. Pinned MyPy 1.13 passes the exact
-two-file isolated CI boundary and all `31` production modules under the coherent
-whole-package profile that disables only redundant-cast warnings; those casts
-remain required by the isolated protected profile. Bandit reports no
-medium/high finding. Documentation governance, manifest validation/layout and
-all `8` manifest contracts, changed-production and minimum-test policies, the
-official 500-LOC PR-delta gate (`6` files, zero violations), diff checks, and
-mandatory maxima of `392` module lines, `46` function lines, and `4` parameters
-all pass.
 
 ## 2026-08-10 PR #4335 isolated-MyPy return typing
 
@@ -231,66 +421,8 @@ This child remains `not_released`. Protected CI, independent review, normal
 stack integration, UI, TypeScript/Rust/PyO3/WASM regional physics, changing
 normals/heights/velocities, internal transition-ledger wire export,
 UpstreamDrift parity, and main release remain open.
-## 2026-08-11 Current skid/roll parent propagated into regional surfaces
-
-- Open draft PR `#4332` retains branch
-  `feat/4271-regional-surface-transitions` and base
-  `feat/4271-ground-skid-roll`. Exact remote regional-surface child
-  `04ccf08dd990de1cd056a3420e67772773a4be2e` is merged first and exact current
-  skid/roll parent `3f861c1faaa7455faee92eaa7f813174667208e0`
-  second by an ordinary no-ff merge.
-- Bounded coplanar material overlays, precedence, exact boundary splitting,
-  transition evidence, continuity, regional production, tests, fixtures,
-  contracts, limitations, and both append-only histories remain intact.
-- Focused local qualification passes all `121` Python ground/regional tests,
-  `20` focused React ground-contract/transfer tests, and all `137` `tools-core`
-  Rust tests. TypeScript, zero-warning ESLint, the 189-module Vite build, Rust
-  format and warning-denied clippy, pinned Ruff 0.14.10 across 14 changed
-  Python files, pinned MyPy 1.13 and Bandit medium/high across 12 production
-  files, the 500-LOC changed-file gate, manifest parsing plus eight contracts,
-  documentation governance, exact child-tree preservation, history/marker,
-  and diff checks pass.
-- The candidate is unpublished and `not_released`; changing normals,
-  discontinuous geometry or surface velocity, regional UI, compiled regional
-  physics, UpstreamDrift parity, protected exact-head CI, review, approval,
-  dependency integration, and release remain open.
-
-## 2026-08-11 PR #4332 current-parent ancestry candidate
-
-The clean dedicated `feat/4271-regional-surface-transitions` worktree starts
-from exact live PR #4332 child `1a48d749af508843fac2a5102f4dd56294429bda`
-and normally merges exact newly published `feat/4271-ground-skid-roll` parent
-`0ea6740965068542e9d8c7449e06ec07d88969e0` as its second parent. PR #4332
-keeps base `feat/4271-ground-skid-roll`; neither branch is rebased, retargeted,
-rewritten, force-pushed, or published by this reconciliation. The only textual
-merge conflict is the independent SPEC-version collision. SPEC 1.14.27 retains
-the regional child record after the parent's 1.14.26 entry; production physics,
-contracts, schemas, numerical ordering, and public APIs merge without conflict.
 
-The child retains bounded coplanar material overlays, explicit precedence,
-exact quadratic boundary splitting, base-edge precedence, state and energy
-continuity, strict `surface_transition` evidence, request-bound transition
-limits, and randomized piecewise-analytic coverage. The current parent retains
-its reviewed impact, bounce, skid, roll, resistance, qualified-rest, edge,
-composition, and passive-ledger behavior. Changing normals, height or
-surface-velocity discontinuities, terrain deformation, torsional-spin damping,
-roll-to-skid transitions, regional UI, compiled regional physics, downstream
-parity, protected CI, review, normal stack integration, and main release remain
-open. This local merge is not release evidence and requires independent review
-before an ordinary fast-forward publication.
 
-Merged-tree qualification is `121` focused Python ground/regional tests and
-`5` focused React ground-contract tests passing. Pinned Ruff 0.14.10 check and
-format pass all `41` ground Python files; pinned MyPy 1.13 passes all `28`
-ground production modules; and Bandit reports no medium/high finding. The
-campaign manifest validator and all `8` manifest contracts, documentation
-governance, changed-production policy, minimum test contract, both parent and
-child diff checks, and the official 500-LOC changed-file gate (`14` files,
-zero violations) pass. Production maxima are `392` lines per module, `46`
-lines per function, and `4` parameters excluding `self`/`cls`. The manifest
-now records open PR #4332 at its still-published child head and open PR #4304
-at its exact newly published parent head; neither record claims protected or
-main-release evidence.
 ## 2026-08-10 issue #4271 coplanar regional-material local child
 
 Local branch `feat/4271-regional-surface-transitions` starts from exact current
@@ -327,555 +459,10 @@ regional physics, UpstreamDrift parity, protected CI, review, normal stack
 integration, and main release remain open. Region plans and from/to identity
 records are execution-scoped non-wire data in this child.
 
-## 2026-08-11 Current impact/bounce parent propagated into skid/roll
-
-- Open draft PR `#4304` retains branch `feat/4271-ground-skid-roll` and base
-  `feat/4270-ground-impact-bounce`. Exact remote skid/roll child
-  `0ea6740965068542e9d8c7449e06ec07d88969e0` is merged first and exact current
-  impact/bounce parent `62beb3e1aa951645fd556a53f9cbae4bb46c47e5`
-  second by an ordinary no-ff merge.
-- The child-owned arbitrary-plane skid, static-feasible pure roll, rolling
-  resistance, qualified rest, edge localization, strict result composition,
-  passive ledgers, tests, fixture, and limitations remain intact with both
-  append-only handoff/SPEC histories.
-- Local verification passes `154` focused ground/flight tests, all eight
-  campaign-manifest tests, and all `137` `tools-core` Rust tests. Pinned Ruff
-  0.14.10 covers 18 PR-delta Python files; pinned MyPy 1.13 and Bandit
-  medium/high cover 12 production files. Rust format/clippy, manifest, docs,
-  exact child-tree, history/marker, and diff gates pass.
-- This remains a local `not_released` candidate. Regional/changing-normal
-  surfaces, deformation, torsional damping, roll-to-skid transitions, UI,
-  compiled physics, publication, protected exact-head CI, review, approval,
-  downstream integration, and release remain open.
-
-## 2026-08-10 PR #4302 corrected-flight-transfer propagation
-
-Draft PR #4302 remains on `feat/4270-ground-impact-bounce` with unchanged base
-`feat/4269-flight-ground-transfer`. Exact corrected #4288 parent
-`247215422a6d4b677552955b4923bc609a553259` is incorporated by the normal merge
-containing this handoff. The child preserves its passive restitution/Coulomb
-sphere-plane impulse, deterministic repeated ballistic hops, exact physical
-contact, capture-to-skid handoff, cancellation/failure semantics, and airborne
-segment evidence while inheriting the corrected transfer ancestry, deterministic
-workspace timestamps, and canonical `swing_sim` import identity. No branch was
-rebased, retargeted, rewritten, or force-pushed.
-
-The child implementation is the scoped Python reference prefix from
-`docs/specs/GROUND_IMPACT_BOUNCE.md`: it includes full spin, moving-boundary
-accounting, exact first contact, bounded repeat-hop execution, typed airborne
-segments, and a shared golden fixture. It remains partial and `not_released`.
-Issue #4271 retains skid, roll, rest, total distance, and final
-`GroundSimulationResult`; terrain deformation, UI execution, TypeScript bounce
-physics, compiled Rust/PyO3/WASM bounce parity, and UpstreamDrift adapters remain
-excluded. Protected CI, independent review, and dependency-ordered collapse
-remain separate release gates.
-
-Merged-tree validation is `987` Python tests, `106` React files / `661` tests,
-and the complete `tools-core` Rust suite at `137` tests (`111` unit, `20`
-transfer, `6` wire). The affected cross-version suite is also `146` tests on
-real CPython 3.10.20. The 189-module Vite production build, TypeScript,
-zero-warning ESLint, Ruff check/format across 90 files, pinned mypy 1.13 across
-all 17 ground and nine transfer production modules, Rust workspace format plus
-warning-denied `tools-core` clippy, campaign-manifest validator plus eight
-contracts, documentation governance, 11-file 500-LOC budget, marker scan, and
-diff checks are clean. Hosted checks and review apply to the new exact merge
-head only.
-
-
-## 2026-08-10 PR #4288 corrected-ground propagation
-
-Draft PR #4288 remains on `feat/4269-flight-ground-transfer` with unchanged
-base `feat/4268-ground-contract`. Exact corrected #4285 parent
-`788aa547651a3685a363ea401824a5d81477bafb` is incorporated by the normal merge
-containing this handoff. The child retains its qualified signed terminal state,
-physical sphere/terrain contact brackets, strict provenance, and cross-runtime
-Python/TypeScript/Rust/PyO3/WASM contracts while inheriting deterministic
-workspace timestamps and canonical `swing_sim` import identity. No branch was
-rebased, retargeted, rewritten, or force-pushed.
-
-This propagation does not add bounce, skid, roll, terrain response, total
-distance, or UI execution. Protected CI, independent review, and normal
-dependency-ordered collapse remain separate release gates.
-
-Merged-tree validation is `951` Python tests, `106` React files / `661` tests,
-and `26` focused Rust transfer/wire tests. The affected cross-version suite is
-also `110` tests on real CPython 3.10.20. The 189-module Vite production build,
-TypeScript, zero-warning ESLint, Ruff check/format across 82 files, pinned mypy
-1.13 across the six-file transfer and 13-file parent production namespaces,
-Rust workspace format plus warning-denied `tools-core` clippy, campaign-manifest
-validator plus eight contracts, documentation governance, 13-file 500-LOC
-budget, marker scan, and diff checks are clean. Hosted checks and review apply
-to the new exact merge head only.
-
-
-## 2026-08-10 PR #4282 workspace timestamp propagation
-
-Exact corrected #4281 parent
-`cf52529b1e68479321bb93b1be3d59c77f782008` is incorporated into
-`feat/4199-wind-workflow` by the normal merge containing this handoff. PR #4282
-keeps base `feat/4199-wind-scalar-adapter`; neither branch was rebased,
-retargeted, rewritten, or force-pushed. The consolidated wind/capability
-workflow inherits the deterministic Python 3.10-3.12 UTC parser, variation
-export, and scalar-ensemble history without changing its own behavior.
-
-The monotonic SPEC assigns 1.14.10 to compatibility, 1.14.11 to variation,
-1.14.12 to the scalar adapter, and 1.14.13 to this workflow carrier. The
-reconciled tree passes `924` combined Python tests, `28` direct compatibility
-tests on real CPython 3.10.20, and `104` React files / `642` tests. The Vite
-production build, TypeScript, zero-warning ESLint, Ruff, format, pinned mypy
-1.13, campaign-manifest validator and its eight contracts, documentation
-governance, source-size, conflict-marker, and diff gates are clean. Protected
-CI, review, and downstream propagation remain open.
-
-
-## 2026-08-10 PR #4280 workspace timestamp propagation
-
-Exact parent `05383d333b6fd87eaf5e37305476f50b505c2c2e` is incorporated into
-`feat/4144-variation-export-continuation` through the normal merge containing
-this handoff. PR #4280 keeps base `feat/4218-toolstrip-workspace`; neither
-branch was rebased, retargeted, force-pushed, or rewritten. The reconciled tree
-retains all variation export/accessibility behavior and adds the strict
-cross-version workspace timestamp parser.
-
-SPEC 1.14.10 remains the parent compatibility entry and SPEC 1.14.11 becomes
-the child variation entry. The reconciled tree passes `778` Rate tests, `27`
-real-Python-3.10.20 compatibility tests, `1 file / 8` focused React tests,
-TypeScript, focused zero-warning ESLint, Ruff, format, and pinned mypy 1.13.
-Documentation, size, and diff gates must remain clean in the merge commit.
-Protected CI, review, and later propagation remain open.
-
-
-## 2026-08-09 PR #4280 variation-export propagation
-
-Draft #4280 remains based on `feat/4218-toolstrip-workspace` and includes exact
-corrected parent `3f67ed466fefc8991db9c4409f921f25e1c37142` through a normal
-merge. The child retains complete selected-scatter-axis CSV export parity,
-typed unavailable rows, PyQt accessible raw tables, and the focused
-table/scatter/matrix split.
-
-
-## 2026-08-10 PR #4279 fractional timestamp compatibility
-
-Descendant PR #4281's exact-head Python 3.10 lane failed the workspace instant
-ordering regression because that interpreter rejects one-, two-, four-, and
-five-digit fractional seconds before comparison. The correction belongs on
-the earliest carrier, PR #4279: enforce one anchored UTC grammar, parse zero
-through six fractional digits consistently, and reject precision beyond
-microseconds instead of allowing newer interpreters to truncate it.
-Persisted JSON remains unchanged and UTC-only validation stays authoritative.
-
-Local evidence is `778 passed` for the full Rate suite, `45 passed` for the
-workspace/compatibility suites, and `27 passed` for the direct compatibility
-suite on CPython 3.10.20. Ruff, format, pinned mypy 1.13, docs governance, and
-the 400-line budget pass. Publication is not completion: verify the exact
-remote head/base, push normally, then propagate in stack order through #4280,
-#4281, #4282, and every later descendant that carries the workspace.
-
-
-## 2026-08-09 Capability results stabilization
-
-The isolated `feat/4201-capability-results-diagnostics` continuation is based
-on exact wind-workflow carrier head `18fe89201`. It closes two audit findings
-without changing the optimizer or physics: both clients expose every ranked
-diagnostic and parameter unit, both provide result CSV plus versioned result
-JSON, and React scalar scatters now carry numeric scales and cohort legends.
-Raw observation exports remain lossless and distinct. This is feature-stack
-implementation evidence only; #4197/#4201 remain open through integration,
-hosted CI, review, downstream parity, and release to `main`.
-
-Local evidence: 813 Rate Python/PyQt tests and 104 React files / 628 tests,
-Ruff check/format, targeted mypy for the new export and tab modules, TypeScript,
-zero-warning ESLint, and the 188-module Vite production build passed. The
-pre-existing mypy 1.19.1/Python 3.13 internal serialization assertion on
-`capability_results.py` reproduces unchanged at carrier `18fe89201`.
-
-
-## 2026-08-09 Web-release stability continuation
-
-Local branch `feat/4201-web-release-stability` is based exactly on campaign
-carrier `18fe89201`. It fixes direct-path execution of
-`src/rate_of_closure/launch_web.py`, adds a subprocess delegation smoke
-contract, and reconciles the supported release boundary to the files that
-exist: static Vite web output plus PyQt6/PyInstaller desktop output. Rate has
-no `src-tauri` project, so its stale Tauri scripts, CLI dependency, lockfile
-entries, and current documentation claims were removed instead of inventing an
-unqualified wrapper.
-
-The load-sensitive `FlightExplorerPanel` assertion now explicitly settles the
-real lazy Wind Strategy import inside React `act`; it does not extend a timeout
-or replace the child component with a mock. This is test synchronization only,
-not a UI or physics change. The branch must remain local until reviewed and
-folded into the established campaign carrier; no PR, push, or merge belongs to
-this handoff slice.
-
-Exact local evidence: 813 Rate Python/PyQt tests passed with 15 existing
-warnings; 9 launcher/registration tests passed; 102 React files / 624 tests
-passed; and five sequential focused Flight Explorer runs passed 25/25 tests.
-TypeScript type-check, zero-warning ESLint, the 187-module static build,
-focused Ruff check/format, targeted mypy, package installation/audit, and diff
-checks pass. `tests/test_dry_compliance.py` retains two unrelated baseline
-failures for the Movement Optimizer and Optimizer GUI PyQt launchers.
-
-
-## 2026-08-09 Machine-readable campaign authority
-
-`docs/release/rate_of_closure_campaign.v1.json` is now the canonical current
-state for the primary Rate, impact, flight, variation, club-builder, wedge,
-toolstrip, design-quality, parity, and ground programs. Its normalized carrier
-and evidence tables replace status inference from the chronological entries
-below. `scripts/rate_campaign_manifest.py` supplies strict Pydantic validation
-and a generated JSON Schema; it rejects missing programs, unresolved evidence,
-placeholders, malformed SHAs, and contradictory release claims.
-
-The authority records the current campaign as **not released**. Capability PRs
-#4294, #4289, and #4283 are merged into feature parent #4282, but those were not
-protected `main` merges. Current carrier head `18fe89201` contains the hermetic
-`swing_core` parity-lane correction and its focused workflow regression; a
-fresh hosted parity run, top-down stack propagation, #4133 reconciliation,
-#4119 current-main conflict resolution, protected checks, installed-package
-evidence, and UpstreamDrift parity are still required.
-
-Maintain the JSON authority and the current-state handoffs in the same commit
-whenever a carrier SHA, test result, limitation, supported surface, or release
-stage changes. Historical detail below remains useful provenance but must not
-override a contradictory validated manifest record.
-
-The four reviewed stabilization slices are composed on implementation head
-`2c1a77baa`: strict workflow parsing/signed editing, complete capability
-diagnostics and result exports, static-web entrypoint/package truth, and this
-manifest authority. Combined local evidence is 828 Rate Python/PyQt tests and
-104 React files / 642 tests, TypeScript, zero-warning ESLint, the 188-module
-Vite build, Ruff, targeted mypy, deterministic manifest/schema validation, and
-nine manifest/parity contracts. This remains local evidence; #4282 still needs
-a normal push, exact-head hosted CI, review, and dependency-ordered release.
-
-The first exact-head hosted quality gate (`31340032608`) passed checkout,
-dependency installation, Ruff, and formatting, then failed mypy 1.13 because
-`--follow-imports=skip` exposed Pydantic and Qt boundary returns as `Any`. The
-loader now casts the validated model and the playback adapter converts elapsed
-milliseconds to a concrete `float`; these are static typing fixes, not schema
-or runtime behavior changes. The exact Python 3.12/mypy 1.13 command passes all
-54 delta files; Ruff, 62 focused regression tests, and eight campaign-manifest
-tests pass locally.
-
-
-## 2026-08-11 Current flight-transfer parent propagated into ground impact
-
-- Open draft PR `#4302` retains branch `feat/4270-ground-impact-bounce` and base
-  `feat/4269-flight-ground-transfer`. Exact remote bounce child
-  `cf6e72bad98e5f36f782254942e6895b8b71e670` is merged first and exact current
-  transfer parent `c39297d37a17d5d8d3520ed62c5563cdcc609cab` second by an
-  ordinary no-ff merge.
-- Passive sphere-plane impact, restitution/friction, deterministic repeated
-  bounce, exact contact, capture-to-skid handoff, cancellation/failure evidence,
-  and the full current transfer ancestry are retained with both append-only
-  handoff/SPEC histories.
-- Three inherited PR-delta Python files receive AST-equivalent pinned Ruff
-  0.14.10 normalization; analysis, persistence, and contract assertion behavior
-  are unchanged.
-- Local verification passes `1129` Rate/ground/flight Python tests, `25`
-  focused React transfer tests, and all `137` `tools-core` Rust tests, plus
-  TypeScript, zero-warning ESLint, the 189-module build, Rust format/clippy,
-  pinned Ruff 0.14.10, pinned MyPy 1.13, Bandit medium/high, manifest parsing,
-  history preservation, and diff checks.
-- This remains a local `not_released` candidate. Skid, roll, rest, total
-  distance, final ground results, UI, TypeScript bounce physics, compiled bounce
-  runtimes, publication, protected exact-head CI, review, approval, dependency
-  integration, and release remain open.
-
-## 2026-08-10 PR #4285 workspace timestamp propagation
-
-Draft PR #4285 remains on `feat/4268-ground-contract` with unchanged base
-`feat/4197-capability-observer`. Exact corrected #4282 carrier
-`5f77af4add23547a21cc3fabce98ae9ad4260427` is incorporated through the normal
-merge containing this handoff. The strict ground-contract child retains its
-schemas, canonical fixture, migrations, legacy-result adapter, declared
-dependency, and pinned typing repairs while inheriting the deterministic Python
-3.10-3.12 UTC parser and the corrected variation, scalar, wind, capability, and
-campaign-release ancestry. No branch was rebased, retargeted, rewritten, or
-force-pushed.
-
-The preceding exact-head Python 3.12 job then exposed an import-order defect,
-not a physics failure: embedded `src.shared.python.swing_sim` test packages and
-canonical `shared.python.swing_sim` imports could become distinct module trees,
-so ground and impact contract collection failed. The shared alias registry now
-coalesces `swing_sim`; a subprocess identity regression was RED before the fix
-and passes with both affected public-API suites afterward. The contemporaneous
-file-size run was cancelled during checkout before its budget step, so no size
-violation was reported.
-
-Merged-tree evidence is `915` Rate/ground/impact Python tests, the full
-`12`-test shared-alias architecture suite, `28` focused tests on CPython 3.12,
-`28` direct compatibility tests on real CPython 3.10.20, and `104` React files /
-`642` tests. The 188-module Vite build, TypeScript, zero-warning ESLint, Ruff,
-format, pinned mypy 1.13, campaign-manifest validator plus eight contracts,
-documentation governance, the 44-file 500-LOC budget, marker scan, and diff
-checks are clean. Hosted CI and review still apply only to the new exact head.
-
-This is ancestry reconciliation, not new landing physics. Qualified bounce,
-skid, roll, terrain profiles, total distance, UI execution, Rust/WASM parity,
-protected CI, review, and release remain open. PR #4288 must normally merge the
-resulting exact #4285 head before its transfer behavior is evaluated as a
-descendant of the corrected carrier.
-
-## 2026-08-11 Current strict-ground parent propagated into flight transfer
-
-- Open draft PR `#4288` retains branch `feat/4269-flight-ground-transfer` and
-  base `feat/4268-ground-contract`. Exact remote child
-  `df716ac740beb9d47dced66f867dc3dde1fa5367` is merged first and exact current
-  parent `584e4e53a7990053a346938c4e45538786887ec7` second by an ordinary
-  no-ff merge.
-- The child-owned Python/TypeScript/Rust/PyO3/WASM transfer, physical contact
-  evidence, production, tests, fixtures, and limitations remain intact. Only
-  append-only handoff/SPEC conflicts require aggregation, preserving every dated
-  record from both parents.
-- Focused verification passes 85 Python contract/transfer tests, 25 React tests,
-  and 137 Rust library and transfer/wire tests. Pinned Ruff 0.14.10 covers 13
-  PR-delta Python files, pinned MyPy 1.13 covers nine production files, and
-  Bandit reports no medium/high findings. TypeScript, exact-delta ESLint, the
-  189-module build, Rust format, docs, strict manifest, diff, exact transfer-tree,
-  and both-parent preservation checks pass.
-- This bounded local audit does not authorize publication or release. Protected
-  exact-head CI, review, approval, and release remain open.
-
-## 2026-08-11 PR #4288 append-only history preservation repair
-
-This documentation-only follow-up to exact local candidate
-`1eb4188ff1a87a3d8e9a55aeca68194ca04e0bb4` restores every omitted #4285
-parent handoff section and specification history row verbatim. It repairs one
-interleaved rate-of-closure paragraph without changing implementation, tests,
-schemas, manifest data, merge topology, or #4288's
-`feat/4268-ground-contract` base. The repaired candidate remains local pending
-independent review, ordinary publication, fresh protected CI, approval,
-dependency integration, and release.
-
-## 2026-08-11 PR #4288 current-ground-parent ancestry candidate
-
-Draft PR #4288 remains on `feat/4269-flight-ground-transfer` with unchanged
-base `feat/4268-ground-contract`. Exact published child
-`afb1bdcfc6701caaf1f7bc3497a6a37dd9698c14` normally merges exact newly
-published #4285 parent `e09ab96280cab363b36bd0e7db4a3cd064dc2527`
-second. The child keeps its bounded TypeScript integrator and qualified
-cross-runtime terminal-state/contact transfer; the parent contributes the
-current strict-ground, wind/capability, split-kinetics, plotting-catalog, and
-format ancestry. Seven child-owned Python files receive pinned Ruff 0.14.10
-normalization with no AST change, and the child TypeScript facade is retained
-with standard line endings. No branch is rebased, retargeted, rewritten, force-pushed, or
-released. Independent review, ordinary publication, fresh protected exact-head
-CI, approval, dependency integration, and release remain required.
-
-## 2026-08-11 Current capability carrier propagated into ground contract
-
-- Open draft PR `#4285` retains `feat/4268-ground-contract` over
-  `feat/4197-capability-observer`. Exact remote child
-  `e09ab96280cab363b36bd0e7db4a3cd064dc2527` is merged first and exact current
-  capability-observer carrier `d0b827667e61f0583e1dad0b3cbbca6819624d3c`
-  second by an ordinary no-ff merge.
-- The merge preserves the complete strict flight-to-ground contract tree and
-  all current wind/capability ancestry. Only append-only handoff/SPEC documents
-  conflict; every dated record from both histories is retained.
-- Focused validation passes 214 Python/PyQt tests. Pinned Ruff 0.14.10 covers
-  all 27 PR-delta Python files, pinned MyPy 1.13 covers all 20 production files,
-  and Bandit reports no medium/high findings. Docs, strict manifest, diff, exact
-  ground-tree, and append-only preservation checks pass.
-- This local candidate is not published or released. Protected exact-head CI,
-  review, approval, and normal descendant propagation into #4288 remain open.
-
-## 2026-08-11 Current workflow/base ancestry reconciled into ground contract
-
-- Draft PR `#4285` remains on `feat/4268-ground-contract`, based on
-  `feat/4197-capability-observer`. Exact remote child
-  `7a38b9838d743e051a6900620c4d6e754582aa89` is the first parent and exact
-  published base/workflow head
-  `b2baf6ef615b0d756a86ac4ca7eef2fc583210ee` is merged second.
-- GitHub ancestry established that the published base contains exact merged
-  #4283 head `9bbb98e16e435a0d4c74153b909f2ebfefbbce7a` before this reconciliation.
-  No branch is rebased, retargeted, force-pushed, or rewritten.
-- Ground-contract behavior is preserved. The only source conflict is an
-  obsolete formatting-only edit to monolithic kinetics and resolves to the
-  parent's validated split facade; all append-only documentation histories
-  remain present.
-- The retained child `float` narrowing selects the plotting catalog as changed.
-  Its literal registry is split by scalar and series responsibility behind the
-  unchanged public catalog facade, restoring the strict 400/50/4 production
-  budgets while preserving registry order and behavior.
-- Complete local evidence is 1,539 passed with one optional Rust-wheel skip for
-  Rate/shared-swing Python, 106 React files / 648 tests plus type-check, lint,
-  and the 188-module build, and 12 Rust tests. The focused contract/parity/
-  kinetics/wind/manifest suite passes 146 tests and the catalog/plot regression
-  suite passes 57 tests.
-- This local candidate is not published or released. Independent exact-head
-  review, ordinary publication, fresh protected CI, required approval, and a
-  later normal descendant merge into #4288 remain release gates.
-
-## 2026-08-11 merged capability-observer carrier receives current #4282
-
-- Merged PR `#4283` used `feat/4197-capability-observer`, which remains the
-  current base of open draft PR `#4285`. Exact remote/local carrier child
-  `b2baf6ef615b0d756a86ac4ca7eef2fc583210ee` is the first parent and exact
-  published #4282 head `3faf9e9493c9ba37168d9f65ab10cafeedc2a72f`
-  is the second parent of this normal local no-ff merge.
-- The carrier is already an ancestor of #4282, so all implementation merges
-  without conflict. The exact #4282 responsive wind/capability/manifest tree and
-  the complete merged capability-observer history are preserved; only this
-  current carrier handoff and SPEC release row are intentionally added.
-- Focused validation is green: 86 Python/PyQt capability, wind, and campaign-
-  manifest tests and all 86 tests in 12 focused React capability/wind suites
-  pass. React type-check, zero-warning lint, and the production build pass. Docs
-  governance, strict campaign-manifest validation, and exact-parent diff hygiene
-  pass. The exact #4282 delta is documentation-only, so changed-Python
-  Ruff/MyPy/Bandit gates are not applicable.
-- This bounded audit does not authorize publication. Independent review,
-  ordinary branch publication, protected exact-head CI,
-  and later #4285 propagation remain open.
-
-## 2026-08-11 current scalar-adapter propagation into wind workflow PR #4282
-
-- PR `#4282` remains on `feat/4199-wind-workflow`, based on
-  `feat/4199-wind-scalar-adapter`. Exact published child
-  `b2baf6ef615b0d756a86ac4ca7eef2fc583210ee` is the first parent and exact
-  newly published scalar-adapter head
-  `9321c1d2e091b8c7e5a4a83aa9ad726290e7fb5a` is the second parent of this
-  normal local merge.
-- Production and test paths merge automatically. Responsive PyQt6/React wind
-  execution, progress/cancellation, controls, scatter, persistence/export,
-  consolidated capability workflow, and strict campaign release authority
-  remain authoritative while inheriting the current scalar-adapter, variation,
-  workspace/toolstrip, launch-monitor/D-plane, split-kinetics, and Qt repairs.
-  Only append-only handoffs and SPEC conflict textually; both histories remain.
-- Focused pre-commit regression is green: 94 Python/PyQt wind, responsive-
-  layout, navigation, playback, and manifest tests pass, as do all 44 tests in
-  eight focused React wind/responsive suites. Docs governance and strict
-  campaign-manifest generation/validation also pass.
-- Exact post-commit PR-delta gates are also green: pinned Ruff 0.14.10 checks
-  and formats 57 changed Python files, pinned MyPy 1.13.0 accepts 38 production
-  files, and pinned Bandit 1.7.7 finds no medium/high issues in 37 scanned source
-  files. The 500-LOC budget, changed-Python policy, minimum-test contract,
-  changed-test assertions, diff hygiene, React type-check/lint, and production
-  web build all pass.
-- No issue or epic is completed by this propagation. Independent review,
-  normal publication,
-  protected exact-head CI, unresolved-thread checks, dependency integration,
-  and release remain open.
-
-## 2026-08-11 Exact #4281 parent propagation into PR #4282
-
-- Draft PR `#4282` retains branch `feat/4199-wind-workflow` and base
-  `feat/4199-wind-scalar-adapter`. Exact remote child
-  `29e15d6ff631f7f30afcc745be783f2e716d7dcf` normally merges exact published
-  parent `a7dce5f89b483303938f518b74b4028a1c68ba81` second without rewriting either
-  history or the dependency stack.
-- Responsive PyQt6/React wind execution, progress/cancellation, controls,
-  scatter, persistence/export, and the consolidated capability stack remain
-  authoritative. The parent contributes its split kinetics facade, Ground/Tee
-  parity, and strict 252-line public/213-line private scalar-adapter design.
-- The obsolete formatting-only child edit to monolithic kinetics resolves to
-  the parent facade. All documentation history remains present. Seven
-  child-owned Ruff-format failures remain the only known current-head code
-  formatting correction. The skipped-import MyPy profile additionally replaces
-  one now-unused cache-return ignore with an explicit typed cast that preserves
-  the cached object and runtime behavior. Independent review and protected CI
-  remain required.
-- Local evidence is 74 focused Python/PyQt wind, kinetics, and Ground/Tee
-  tests; 29 focused React tests; and 8 campaign-manifest tests. TypeScript,
-  focused zero-warning ESLint, Ruff check/format, pinned MyPy 1.13, Bandit's
-  medium/high threshold, strict wind module/signature/function budgets,
-  manifest, documentation, policy, SPEC, and diff gates pass.
-
-## 2026-08-11 current variation propagation into wind scalar PR #4281
-
-- PR `#4281` remains on `feat/4199-wind-scalar-adapter`, based on
-  `feat/4144-variation-export-continuation`. Exact published child
-  `a7dce5f89b483303938f518b74b4028a1c68ba81` is the first parent and exact
-  newly published variation head `e6c7460a01082631565fb9ed48aa32538bd7772c`
-  is the second parent of this normal local merge.
-- Production and test paths merge automatically. Matching Python/TypeScript
-  `scalar-ensemble/v1` provenance, availability, deterministic-scenario,
-  wind-strategy adapter, and strict module-budget behavior remain authoritative
-  while inheriting current variation visualization/export, workspace/toolstrip,
-  launch-monitor/D-plane ancestry, split kinetics, and Qt typing repairs. Only
-  the three canonical handoffs and SPEC conflict textually; both append-only
-  histories are retained.
-- Focused pre-commit regression is green: 77 Python wind/variation tests and all
-  20 tests in the five focused React wind/variation suites pass. Exact
-  post-commit PR-delta gates are also green: pinned Ruff 0.14.10 checks/formats
-  five changed Python files, pinned MyPy 1.13.0 accepts three production files,
-  and pinned Bandit 1.7.7 finds no issues in those three files. Both 500- and
-  strict 400-LOC budgets, docs governance, changed-Python policy, minimum-test
-  contract, changed-test assertions, diff hygiene, React type-check/lint, and
-  the production web build pass. The manifest artifact/checker/test exist on
-  neither side of this pre-manifest stack and are not applicable.
-- No issue or epic is completed by this propagation. Independent review,
-  normal publication, protected exact-head CI, unresolved-thread checks,
-  dependency integration, and release remain open.
-
-## 2026-08-11 PR #4281 wind adapter module-budget repair
-
-- The public `wind_strategy_plot_adapter.py` is reduced from 403 to 252 lines
-  by extracting its request/analysis validation into the 213-line private
-  `_wind_strategy_plot_validation.py` collaborator. Both now satisfy the
-  mandatory below-400 production-module ceiling.
-- A frozen, slotted scenario-expectation object reduces the private scenario
-  validator from five loose parameters to two cohesive parameters. Every
-  production function is at or below four parameters and 50 lines.
-- Public imports, the `scalar-ensemble/v1` dataset, variable and cohort order,
-  composite row identity, values, availability, attributes, validation
-  messages, and the rule that adaptation never invokes flight physics are
-  unchanged.
-- Local focused evidence is 14 Python adapter tests and 16 React wind tests,
-  plus Ruff/format, pinned MyPy 1.13, Bandit, and the explicit line budget.
-  Independent review, exact-head protected CI, ordinary publication, and
-  dependency order remain required.
-
-## 2026-08-11 reviewed variation parent propagation into wind scalar PR #4281
-
-- PR `#4281` remains on `feat/4199-wind-scalar-adapter`, based on
-  `feat/4144-variation-export-continuation`. Exact remote child
-  `247046d55afcad3e6cd4f8029f854856c427f59c` normally merges exact reviewed
-  parent `3337945699966b63cb5cd8e52d7c3b194315e911` in child-first order,
-  preserving both histories without rebase, reset, retarget, force-push, or
-  parent rewrite.
-- Matching Python/TypeScript `scalar-ensemble/v1` provenance, availability,
-  deterministic-scenario, and wind-strategy plot-adapter behavior remain
-  authoritative. The parent contributes selected-scatter export and linked
-  visualization behavior, the kinetics split, Ground/Tee parity contracts,
-  workspace/toolstrip repairs, and protected Ruff normalization.
-- The sole feature-code conflict referenced obsolete monolithic kinetics and
-  is resolved to the validated parent `pendulum.sample(...)` facade. Both
-  append-only handoff/SPEC histories are retained. This local reconciliation
-  still requires independent review, fresh exact-head protected CI,
-  unresolved-thread checks, dependency order, and ordinary publication before
-  any merge or release claim.
-
-## 2026-08-10 Second #4280 propagation into wind scalar PR #4281
-## 2026-08-10 Repaired wind-scalar parent propagation into PR #4282
-
 Status verified 2026-08-08. This isolated integration is published as draft
 [PR #4217](https://github.com/D-sorganization/Tools/pull/4217). No source PR
 branch was rewritten.
 
-## 2026-08-11 PR #4304 current-parent ancestry candidate
-
-The clean dedicated `feat/4271-ground-skid-roll` worktree first fast-forwarded
-to exact live PR #4304 child `52d9a6a978d8e6b8b19ef92f02f265c9058b00ad`,
-then normally merged exact live `feat/4270-ground-impact-bounce` parent
-`cf6e72bad98e5f36f782254942e6895b8b71e670`. The PR base remains unchanged;
-neither branch was rebased, retargeted, rewritten, force-pushed, or published.
-The automatic merge was conflict-free. Its only child-tree change before this
-documentation reconciliation is deterministic formatting in the existing
-skid/roll regression test; production physics, contracts, schemas, numerical
-ordering, and public APIs remain unchanged.
-
-Local qualification on the merge candidate is `115` focused ground
-tests passing on CPython 3.13, pinned Ruff 0.14.10 clean across the ground
-package and tests, and pinned MyPy 1.13 clean across all `25` ground production
-modules. The manifest validator and all `8` manifest tests, documentation
-governance, exact-parent and exact-child diff checks, changed-production policy,
-the official changed-file 500-LOC budget, and mandatory production limits of
-400 lines per module, 50 lines per function, and four parameters all pass. The
-local merge is a PUBLISH candidate for independent review only; protected
-exact-head CI, approval, normal child propagation, and publication remain open.
-
 ## 2026-08-10 PR #4304 corrected-impact propagation
 
 Draft PR #4304 remains on `feat/4271-ground-skid-roll` with unchanged base
@@ -918,140 +505,6 @@ repair. All `115` ground tests, Ruff, formatting, a finding-free local scan of
 the affected file, documentation governance, the `370`-line source-size
 check, and diff gates pass before an ordinary guarded fast-forward
 publication; protected CI and review remain open.
-## 2026-08-10 Final repaired #4285 propagation into PR #4288
-
-## Inherited current-parent campaign history
-
-## 2026-08-10 PR #4282 exact-head mypy repair
-
-- Hosted run `31395741841`, quality job `93478091420`, checked merge ref
-  `ae75c4b006072b20a713e45b177defcc16a380fe` for exact #4282 head
-  `aa6eeffb0395f7ed7954f2315b1c625cada552d8`.
-- Checkout, installation, Ruff, and format passed. Pinned mypy 1.13 failed only
-  `scripts/rate_campaign_manifest.py:335:5 [no-any-return]` because the exact
-  skipped-import profile treats Pydantic's validator result as `Any`.
-- The validated local now has an explicit `CampaignManifest` annotation. This
-  preserves the strict JSON contract and avoids both a suppression and the
-  redundant cast reported by the non-skipped profile.
-- The failed run belongs to the obsolete head. Publish normally, preserve the
-  PR base/draft state, and propagate the resulting exact head into #4285.
-
-## 2026-08-10 Exact #4281 propagation into wind workflow PR #4282
-
-- PR `#4282` remains on `feat/4199-wind-workflow`, based on
-  `feat/4199-wind-scalar-adapter`.
-- Original child `5f77af4add23547a21cc3fabce98ae9ad4260427` normally merges exact
-  parent `4a793c4c3f19aad43a3c215800b266be487ace49`; preserve that parent order
-  and do not rebase, retarget, force-push, rewrite a parent, change draft state,
-  or merge the PR.
-- The responsive PyQt6/React wind workflow, consolidated capability stack, and
-  campaign release authority remain additive with the parent's strict
-  scalar-ensemble wind adapter, variation visualization, result-local
-  selection, workspace/toolstrip, and Python 3.10 compatibility repairs.
-- The canonical navigation-state tuple now includes the child-owned capability
-  tab while retaining the parent's legacy-order visibility migration; no
-  duplicate navigation constant remains.
-- Static review replaces optimized-away UI/capability assertions with explicit
-  runtime invariant failures and records the fixed-argument executable builder
-  subprocess boundary using Bandit's native suppression syntax. The separate
-  exact-CI Pydantic return repair is recorded above.
-- SPEC 1.14.14 is the unique combined child entry. Verification passes 73
-  focused Python tests, the complete 1,657-test Rate/shared-swing/golf-club
-  matrix with two explicit optional build123d skips, 643 React tests across 105
-  files plus type-check/lint/build, 12 Rust tests, 31 real-CPython-3.10
-  compatibility tests, and 43 post-review invariant/GUI/manifest tests.
-  Ruff/format, pinned MyPy 1.13 across 37 changed production files, Bandit,
-  exact-parent size, manifest, docs, minimum-test, assertions, CI-policy
-  secrets, compilation, and diff gates pass. Independent re-review found no
-  actionable findings after 94 Python workflow/navigation tests, 20
-  manifest/GUI tests, 61 React workflow tests, and independent static checks.
-  Protected current-head CI, unresolved threads, and repository review remain
-  release gates after publication.
-## 2026-08-09 PR #4282 corrected-parent propagation
-
-Draft #4282 remains on `feat/4199-wind-workflow` with unchanged base
-`feat/4199-wind-scalar-adapter`. Exact corrected #4281 parent
-`958770049f0124dac0426a6dd62fd4edbf437e7a` is incorporated by the normal
-local merge containing this handoff; no branch was rebased, retargeted,
-force-pushed, or published. The result preserves the responsive PyQt6/React
-wind-strategy workflow, the consolidated capability stack, release-manifest
-authority, and launcher/type-gate repairs while inheriting the complete
-Python 3.10 compatibility, variation-export, and scalar-ensemble corrections.
-
-## 2026-08-09 Capability workflow hardening continuation
-
-Local branch `feat/4201-capability-contract-hardening` starts at exact carrier
-`18fe89201d657116bbca99922297c14968356c44`. It makes the Python and TypeScript
-`capability-optimization-workflow/v1` readers fail closed on JSON primitive
-type mismatches and pins the identical accept/reject matrix in a shared golden
-fixture. The Python reader validates the complete nested wire document before
-calling model constructors, so numeric strings, booleans-as-numbers, numeric
-text identifiers/provenance, and fractional integer fields cannot be silently
-normalized. Integral JSON numbers remain portable between runtimes.
-
-The React Shot Optimizer also replaces browser-managed number inputs with
-decimal text drafts. This makes leading-minus and decimal entry reliable while
-preserving numeric commit and existing workflow validation. RED evidence was
-Python `8 failed, 16 passed` and React `2 failed, 26 passed`; final focused
-evidence is Python `43 passed`, React `9 files / 69 tests passed`, with Ruff,
-format, mypy, TypeScript, ESLint, structural, and diff checks clean. This
-continuation also builds 187 Vite modules without a size warning; it is local
-only and has not changed GitHub state.
-
-## 2026-08-10 Final repaired #4282 propagation into PR #4285
-
-- Draft PR `#4288` keeps branch `feat/4269-flight-ground-transfer`, base
-  `feat/4268-ground-contract`, and original child
-  `108a841b1378c992defd3c7b7ee263d41a6c8b24` first.
-- Exact quality-green strict-contract head
-  `a93edb4bfd6a8dc9334122cd2ae660983d5bf424` is the second parent.
-- Only the four canonical handoff/spec files conflicted; transfer production
-  and tests did not. No history or topology rewrite was used.
-- Fresh exact-head protected CI and review remain mandatory; queued checks are
-  not release evidence.
-
-## 2026-08-10 PR #4288 exact repaired-ground propagation
-
-Draft PR #4288 remains on `feat/4269-flight-ground-transfer` with unchanged
-base `feat/4268-ground-contract`. Original child
-`247215422a6d4b677552955b4923bc609a553259` normally incorporates exact repaired
-#4285 parent `e5bcbd1096d3be1f621a805c9d9f3fd321e375a5` second in the
-merge containing this handoff. The child retains its qualified signed terminal state,
-physical sphere/terrain contact brackets, strict provenance, and cross-runtime
-Python/TypeScript/Rust/PyO3/WASM contracts while inheriting deterministic
-workspace timestamps, canonical `swing_sim` import identity, the hosted-mypy
-manifest repair, and the complete variation, wind, capability, and campaign
-ancestry. No branch was rebased, retargeted, rewritten, or force-pushed.
-
-This propagation does not add bounce, skid, roll, terrain response, total
-distance, or UI execution. Protected CI, independent review, and normal
-dependency-ordered collapse remain separate release gates.
-
-Exact composed-tree verification is 1,080 Python tests passed with six explicit
-optional installed-`tools_core` wheel skips, 107 React files / 662 tests,
-26 direct Rust transfer/wire tests, TypeScript, zero-warning ESLint, and the
-189-module production build. Ruff check/format passes all 28 changed Python
-files; pinned mypy 1.13 passes all 21 changed production files. Campaign
-manifest validation, documentation governance, and diff checks are clean. The
-missing local wheel is not accelerated installed-package evidence and remains
-an explicit release boundary; the direct Rust suites executed rather than
-being relabeled as wheel parity.
-
-## 2026-08-10 Exact repaired #4282 propagation into PR #4285
-
-Draft PR #4285 remains on `feat/4268-ground-contract` with unchanged base
-`feat/4197-capability-observer`. Original child
-`788aa547651a3685a363ea401824a5d81477bafb` normally incorporates exact
-repaired #4282 carrier `686016196a2f895058b8a566dff103a0fd32cd10`
-second. That parent contains merged capability PR #4283 commit
-`c1827bbdc50a6e11cc475db2636b4e47a4c15416`, exact observer head
-`9bbb98e16e435a0d4c74153b909f2ebfefbbce7a`, and the hosted pinned-mypy
-manifest repair. The strict ground-contract child retains its
-schemas, canonical fixture, migrations, legacy-result adapter, declared
-dependency, and pinned typing repairs while inheriting the deterministic Python
-3.10-3.12 UTC parser and the corrected variation, scalar, wind, capability, and
-campaign-release ancestry. No branch was rebased, retargeted, rewritten, or
-force-pushed.
 
 ## 2026-08-09 PR #4304 corrected implementation evidence
 
@@ -1064,29 +517,6 @@ all pass locally. This registry/handoff commit is documentation-only and makes
 no material runtime, physics, numerical, schema, or API change. Fresh
 protected CI and review remain required after an ordinary guarded push.
 
-## 2026-08-10 Exact repaired #4282 propagation into PR #4285 (continued)
-
-The inherited ground descendant passes 1,703 Python tests with two optional `build123d`
-skips, 643 React tests across 105 files plus type-check/lint/build, 12 Rust tests,
-and 77 real-CPython-3.10 ground and compatibility tests. Ruff/format pass 78
-changed Python files; pinned mypy and Bandit pass 52 changed production files.
-Manifest, docs, minimum-test, assertions, 500-LOC, changed-file secrets, Python
-3.10 compilation, and diff checks are clean. Hosted CI and review still apply
-only to the new exact head. That repaired ground head is now incorporated into
-#4288; normal propagation of the reviewed #4288 head into #4298 is the next
-ancestry gate.
-
-The repaired parent owns the explicit Pydantic manifest annotation. The
-composed tree additionally normalizes contract-checked scalar plot values,
-enforces disabled-command invariants under optimized Python, and documents the
-optional alias finder's reviewed catch-and-continue security boundary.
-
-This is ancestry reconciliation, not new landing physics. Qualified bounce,
-skid, roll, terrain profiles, total distance, UI execution, Rust/WASM parity,
-protected CI, review, and release remain open. The exact repaired #4285 head is
-now incorporated into #4288; the transfer behavior is evaluated on that
-corrected descendant before normal propagation into #4298.
-
 ## 2026-08-09 PR #4304 isolated-MyPy correction
 
 Hosted quality-gate run `31354071845` (job `93350276996`) failed exact
@@ -1236,32 +666,6 @@ terrain deformation, UI, TypeScript physics, Rust/PyO3/WASM, and UpstreamDrift
 adapters remain excluded. Protected CI, required review, ordinary parent
 integration, and downstream delivery are still required.
 
-## 2026-08-09 PR #4285 corrected-carrier reconciliation
-
-Draft #4285 remains on `feat/4268-ground-contract` with unchanged base
-`feat/4197-capability-observer`. Exact corrected wind carrier
-`bb101cedd555d07d493aae998b46050c68660cdd` is incorporated through the
-normal local merge containing this handoff; no branch was rebased, retargeted,
-force-pushed, or published. The ground parent retains its strict schemas,
-canonical fixture, migrations, legacy-result adapter, explicit dependency, and
-pinned-mypy repairs while acquiring the corrected wind/scalar/variation,
-Python 3.10 compatibility, capability, and campaign-release ancestry.
-
-This is an ancestry reconciliation, not new landing physics. Qualified bounce,
-skid, roll, terrain profiles, total distance, UI, Rust/WASM parity, protected
-CI, review, and release remain open. PR #4288 must normally merge the resulting
-exact #4285 head and rerun its transfer and combined compatibility gates. That
-next propagation is where the descendant acquires the observer-to-wind
-reconciliation; neither PR base should change.
-Draft #4282 remains on `feat/4199-wind-workflow` with unchanged base
-`feat/4199-wind-scalar-adapter`. Exact corrected #4281 parent
-`958770049f0124dac0426a6dd62fd4edbf437e7a` is incorporated by the normal
-local merge containing this handoff; no branch was rebased, retargeted,
-force-pushed, or published. The result preserves the responsive PyQt6/React
-wind-strategy workflow, the consolidated capability stack, release-manifest
-authority, and launcher/type-gate repairs while inheriting the complete
-Python 3.10 compatibility, variation-export, and scalar-ensemble corrections.
-
 ## 2026-08-09 PR #4288 corrected-ground-parent propagation
 
 Draft #4288 remains on `feat/4269-flight-ground-transfer` with unchanged base
@@ -1329,443 +733,6 @@ against a `1e-12` absolute tolerance. This branch does not change the wind
 workflow. The hosted Rust `-lpython3.11` linker failure remains runner/toolchain
 infrastructure.
 
-## Additional inherited parent campaign history
-
-This is an ancestry reconciliation, not new landing physics. Qualified bounce,
-skid, roll, terrain profiles, total distance, UI, Rust/WASM parity, protected
-CI, review, and release remain open. PR #4288 must normally merge the resulting
-exact #4285 head and rerun its transfer and combined compatibility gates. That
-next propagation is where the descendant acquires the observer-to-wind
-reconciliation; neither PR base should change.
-Draft #4282 remains on `feat/4199-wind-workflow` with unchanged base
-`feat/4199-wind-scalar-adapter`. Exact corrected #4281 parent
-`958770049f0124dac0426a6dd62fd4edbf437e7a` is incorporated by the normal
-local merge containing this handoff; no branch was rebased, retargeted,
-force-pushed, or published. The result preserves the responsive PyQt6/React
-wind-strategy workflow, the consolidated capability stack, release-manifest
-authority, and launcher/type-gate repairs while inheriting the complete
-Python 3.10 compatibility, variation-export, and scalar-ensemble corrections.
-## 2026-08-11 current workspace propagation into variation PR #4280
-
-- PR `#4280` remains on `feat/4144-variation-export-continuation`, based on
-  `feat/4218-toolstrip-workspace`. Exact published child
-  `3337945699966b63cb5cd8e52d7c3b194315e911` is the first parent and exact
-  newly published workspace head `efbca84095b617b4018732f7802c2da3f0525387`
-  is the second parent of this normal local merge.
-- Production and test paths merge automatically. Variation selected-scatter
-  CSV parity, typed unavailable outcomes, bounded accessible tables, linked
-  selection, and all-trial arc analysis remain authoritative while inheriting
-  the parent's current workspace/toolstrip, launch-monitor/D-plane ancestry,
-  split kinetics, and Qt typing repairs. Only the three canonical handoffs and
-  SPEC conflict textually; both append-only histories are retained.
-- Focused pre-commit regression is green: 56 Python variation tests and all 10
-  tests in the three focused React variation suites pass. Exact post-commit
-  PR-delta gates are also green: pinned Ruff 0.14.10 checks/formats five changed
-  Python files, pinned MyPy 1.13.0 accepts four production files, and pinned
-  Bandit 1.7.7 finds no issues in those four files. Docs governance, 500-LOC
-  budget, changed-Python policy, minimum-test contract, changed-test assertions,
-  diff hygiene, React type-check/lint, and the production web build pass. The
-  manifest artifact/checker/test exist on neither side of this pre-manifest
-  stack, so that later release gate is structurally not applicable.
-- No issue or epic is completed by this propagation. Independent review,
-  normal publication, protected exact-head CI, unresolved-thread checks,
-  dependency integration, and release remain open.
-
-## 2026-08-11 reviewed workspace parent propagation into variation PR #4280
-
-- PR `#4280` remains on `feat/4144-variation-export-continuation`, based on
-  `feat/4218-toolstrip-workspace`. Exact remote child
-  `668ba96746f79f7a12e8092161bd610054197f58` normally merges exact reviewed
-  parent `ccd0e026c580c93038fdf5c59d5d452a85ba27a0` in child-first order,
-  preserving every commit and both histories without rebase, reset, retarget,
-  force-push, or parent rewrite.
-- Variation selected-scatter CSV parity, typed unavailable outcomes, bounded
-  accessible tables, linked selection, and all-trial arc analysis remain
-  authoritative. The parent contributes its validated kinetics split,
-  Ground/Tee parity contracts, protected Ruff normalization, complete
-  workspace/toolstrip behavior, and remote-history reconciliation.
-- The sole feature-code conflict was obsolete monolithic kinetics source and
-  is resolved to the current `pendulum.sample(...)` façade. Both append-only
-  handoff histories are retained. Seven duplicate child automation edits fail
-  protected Ruff 0.14.10 formatting and are normalized to exact reviewed
-  parent blobs without removing their commit from history. This local merge
-  still requires independent review, fresh exact-head protected CI,
-  unresolved-thread checks, dependency order, and ordinary publication before
-  any release or merge claim.
-
-## 2026-08-10 Second #4279 propagation into variation PR #4280
-
-- PR `#4280` stays on `feat/4144-variation-export-continuation`, based on
-  `feat/4218-toolstrip-workspace`.
-- Exact repaired parent `61b7f48b5aeb7d57246b4963da3df086e79cbe15`
-  is incorporated through a normal merge commit. No feature code conflicts and
-  no branch/PR topology or parent history is rewritten.
-- Preserve the variation/export implementation and both append-only handoff
-  histories. Fresh protected CI, review, and dependency order remain open.
-- Post-reconciliation evidence is 25 focused D-plane/impact tests plus docs
-  governance, changed-file-size, and whitespace gates.
-
-## 2026-08-10 Exact #4279 propagation into variation PR #4280
-
-- PR `#4280` remains the immediate child on
-  `feat/4144-variation-export-continuation`, based on
-  `feat/4218-toolstrip-workspace`.
-- Original child `f90836e342efc8be624739802375af2876d11e5f` normally merges
-  exact parent `6717e9e09d507dbc24bedb36177f1cdf0b4fd90b`; preserve that
-  parent order and do not rebase, retarget, force-push, rewrite a parent, or
-  change draft state.
-- The variation selected-scatter export, typed unavailable rows, accessible
-  raw tables, linked trial selection, and focused PyQt/React visualization
-  modules merge cleanly with the parent's full workspace/toolstrip, playback,
-  plots, navigation, compatibility, and module-budget repairs.
-- Staged review found a release-blocking stale linked selection after rerunning
-  a smaller study. PyQt now clears selection before each result replacement
-  and validates public selection indices against the current trial count;
-  React clears on result identity change and bounds the value shared with all
-  linked views. Larger-to-smaller rerun regressions cover both clients.
-- SPEC 1.14.12 is the unique combined child release entry. Verification is
-  green across 37 focused variation tests, 1,528 complete
-  Rate/shared-swing/golf-club tests with two explicit optional build123d skips,
-  546 React tests across 90 files plus type-check/lint/build, exact-parent
-  Ruff/Ruff-format/Black/pinned MyPy/Bandit, file-size, docs, minimum-test,
-  assertion, detect-secrets, diff, real CPython 3.10.20 compilation, and 30
-  compatibility regressions. There is no Rust delta from the exact parent.
-  Independent staged re-review found no actionable findings after checking
-  both the first replacement-render React boundary and validated PyQt values.
-  Current-head protected CI, review threads, and repository review remain
-  release gates after publication.
-## 2026-08-11 current #4203 propagation into workspace PR #4279
-
-- PR `#4279` remains on `feat/4218-toolstrip-workspace`, based on
-  `feat/4181-launch-monitor-registry`. Exact published child
-  `ccd0e026c580c93038fdf5c59d5d452a85ba27a0` is the first parent and exact
-  newly published #4203 head `7abce9ad767fe8311da66a1e5998b892ea3ca9de`
-  is the second parent of this normal local merge.
-- Production and test paths merge automatically. Workspace/toolstrip,
-  module-visibility, navigation, granular playback, and independent plots are
-  preserved while the descendant inherits the parent's current launch-monitor,
-  D-plane, split-kinetics, and Qt typing repairs. Only the three canonical
-  handoffs and SPEC conflict textually; both append-only histories are retained.
-- Focused pre-commit regression is green: 142 exact PR-delta Python tests pass,
-  and all 32 tests in the eight changed React suites pass. Exact post-commit
-  PR-delta gates are also green: pinned Ruff 0.14.10 checks/formats 27 changed
-  Python files, pinned MyPy 1.13.0 accepts 18 production files, and pinned
-  Bandit 1.7.7 finds no medium/high issues in those 18 files. Docs governance,
-  500-LOC budget, changed-Python policy, minimum-test contract, changed-test
-  assertions, diff hygiene, React type-check/lint, and the production web build
-  pass. The manifest artifact/checker/test exist on neither side of this
-  pre-manifest stack, so that later release gate is structurally not applicable.
-- No issue or epic is completed by this propagation. Independent review,
-  normal publication, protected exact-head CI, unresolved-thread checks,
-  dependency integration, and release remain open.
-
-## 2026-08-11 remote automation reconciliation for workspace PR #4279
-
-- PR `#4279` remains on `feat/4218-toolstrip-workspace`, based on
-  `feat/4181-launch-monitor-registry`. Exact local head
-  `0b22c401a26c31441a599d8d9b39de123706e7ea` ordinarily merges exact remote
-  head `61fe2d556a5413e525d958612ccfd57e65b8d5a2`; both histories and every
-  commit remain reachable without reset, rebase, force-push, retarget, or
-  parent rewrite.
-- The remote automation commit spans 23 paths but is formatting-only. Fifteen
-  paths were already byte-identical through the current parent, including its
-  six existing `.codex-worktrees` gitlinks. Its seven unique formatting edits
-  do not match protected Ruff 0.14.10 output and are normalized back to the
-  pinned form. The sole content conflict was between obsolete pre-split
-  kinetics source and the current façade; resolution retains the current
-  `pendulum.sample(...)` implementation and parent module split.
-- No workspace/toolstrip feature code is lost. Module visibility, navigation,
-  granular playback, independent plots, physics, frames, units, schemas, and
-  public contracts remain unchanged. Fresh exact-head protected CI,
-  independent review, unresolved-thread checks, dependency order, and normal
-  publication are still required; this local merge is not a release claim.
-
-## 2026-08-11 hosted MyPy repair propagation into workspace PR #4279
-
-- PR `#4279` remains on `feat/4218-toolstrip-workspace`, based on
-  `feat/4181-launch-monitor-registry`. Exact child
-  `7806a16f58e1c6999d32f0127a187fbb21f839a1` normally merges exact published
-  parent `3796b49e40b677fbac4e05739f8be49f905df2cb` without rebase, retarget,
-  force-push, or history rewrite.
-- There are no feature-code conflicts. The only inherited production changes
-  are four static NumPy-array casts that close hosted MyPy return boundaries;
-  workspace/toolstrip behavior and numerical, frame, unit, schema, rendering,
-  and public-API semantics remain unchanged.
-- Preserve both append-only histories below. This local merge requires fresh
-  merged-tree protected CI, required review, unresolved-thread checks, and
-  dependency order before publication or merge.
-
-## 2026-08-11 latest #4203 propagation into workspace PR #4279
-
-- PR `#4279` remains on `feat/4218-toolstrip-workspace`, based on
-  `feat/4181-launch-monitor-registry`; exact parent
-  `0216a547aa79727091a2939b96e779e8ddbd7304` is incorporated by a normal
-  two-parent merge without rebase, retarget, force-push, or history rewrite.
-- There are no feature-code conflicts. The child retains its workspace,
-  toolstrip, module-visibility, navigation, granular playback, and independent
-  plot behavior while inheriting the parent's pinned formatting and
-  identity-preserving 222/205/131-line kinetics split.
-- Preserve both append-only histories below. Fresh merged-tree protected CI,
-  required review, unresolved-thread checks, and dependency order remain
-  release gates; this local merge is not a completion or publication claim.
-
-## 2026-08-10 Second #4203 propagation into workspace PR #4279
-
-- PR `#4279` remains the immediate child on
-  `feat/4218-toolstrip-workspace`, based on
-  `feat/4181-launch-monitor-registry`.
-- Exact repaired parent head
-  `12dd76a8dbcc106c4683f2f2e53076f8dc6f1b76` is incorporated through a
-  normal merge commit; no feature code conflicts and no branch or PR topology
-  is rewritten.
-- Preserve both append-only handoff histories. Fresh child protected CI,
-  review, and earlier dependency gates remain required.
-- Post-reconciliation evidence is 25 focused D-plane/impact tests plus docs
-  governance, changed-file-size, and whitespace gates.
-
-## 2026-08-10 Exact #4203 propagation into workspace PR #4279
-
-- PR `#4279` remains the immediate child on branch
-  `feat/4218-toolstrip-workspace`, based on
-  `feat/4181-launch-monitor-registry`.
-- Original child `05383d333b6fd87eaf5e37305476f50b505c2c2e` normally merges
-  exact parent `31cbc007d4c85b5479b7cd0fb0969124eab2af67`; preserve that
-  two-parent order and do not rebase, retarget, force-push, or change draft
-  state.
-- The child workspace/toolstrip, module-visibility, granular playback, path
-  trail, and independent plot behavior remain authoritative. The parent
-  `ImpactLayerControls`, navigation-state constants, plotting catalog, and
-  triple-pendulum extractions are integrated without duplicate state.
-- Verification is green across 1,339 Python tests with six explicit skips, 545
-  React tests plus type-check/lint/build, the 40-test post-format PyQt rerun,
-  exact-parent Ruff/Ruff-format/Black/pinned MyPy/Bandit, the 500-line budget,
-  docs/minimum-test/assertion/detect-secrets/diff gates, real CPython 3.10.20
-  compile/import checks, and 30 compatibility regressions. There is no Rust
-  delta from the exact parent. Independent staged review found no actionable
-  findings after 76 additional focused PyQt/navigation/workspace tests.
-  Current-head protected CI, unresolved-thread checks, and required repository
-  review remain release gates after guarded publication.
-
-## 2026-08-11 #4203 append-only SPEC preservation repair
-
-- Independent review of local candidate `e20b4f630...` found four exact
-  current-parent history rows missing from SPEC: 2026-08-10 versions 1.13.11,
-  1.13.9, 1.13.7, and 1.13.6. The current documentation-only follow-up restores
-  those rows verbatim and advances the declared/changelog version together.
-- Runtime files, tests, ordered merge parents, PR base, and validation evidence
-  are byte-unchanged. Re-review, normal publication, exact-head protected CI,
-  and downstream propagation remain required.
-
-## 2026-08-11 #4203 current-parent reconciliation
-
-- The current D-plane base is exact head
-  `f3363aa88868f6a5c7e9ccfc682a9eca014e86c1`; #4203's exact published child
-  is `217e36dc93d30f79826847f958fbcd10805e58ed`. They are being composed in
-  that order by a normal merge while retaining base `feat/4189-dplane`.
-- The parent's nine changed files merge without semantic conflict except for
-  formatting around the inherited kinetics origin offset. The split child
-  facade remains authoritative because the obsolete parent expression reaches
-  through `source.inner`; the equivalent explanatory comment is preserved.
-- The exact CI-pinned MyPy 1.13 delta found four additional Qt-stub `Any`
-  boundaries after the merge. Explicit `bool`/`str` narrowing now closes the
-  responsive-event, legend, ball-setup event-filter, and status-text contracts
-  without changing their runtime values or interaction behavior.
-- No issue or epic is complete from this reconciliation. Independent review,
-  ordinary publication, protected CI, downstream propagation, and release
-  gates remain open.
-
-## 2026-08-11 #4203 exact-head format completion
-
-- Exact published head `7d69a545ae555679f0318940e67c1786626d6794`
-  failed only the Ruff format step. The exact pinned 0.14.10 reproduction found
-  eleven noncompliant changed files: four inherited files plus seven altered by
-  the automated pre-commit repair.
-- The pending correction is mechanical Ruff formatting of those eleven files.
-  It changes no runtime behavior or public contract and still requires AST
-  equivalence, focused tests, independent review, ordinary fast-forward
-  publication, and fresh protected exact-head CI.
-
-## 2026-08-11 #4203 hosted MyPy NumPy boundary repair
-
-- Exact PR head `0216a547aa79727091a2939b96e779e8ddbd7304` reached runtime merge
-  ref `aede309` but failed CI Standard run `31477542889`, job `93734652129`,
-  with four `no-any-return` findings introduced when kinetics was split for the
-  size budget.
-- Explicit `numpy.ndarray` casts now close only the static return boundaries
-  around force norms, the RK4 derivative concatenation, and the app-frame
-  projection. There is no numerical operation, copy, schema, frame, unit,
-  rendering, or public API change.
-- The acceptance profile is the entire 102-file PR-base changed-source MyPy
-  set, followed by focused kinetics/presentation/PyQt tests, the full Rate
-  suite, and the repository's Ruff/format, size, Bandit, documentation, and
-  diff gates. A three-file MyPy pass alone is insufficient release evidence.
-- That local profile is now green: MyPy 102 files; Ruff/format 141 files;
-  Bandit 101 production files with no medium/high finding; 28 focused and 701
-  full Rate tests; plus module/file size, documentation, minimum-test, and diff
-  gates.
-- The fix remains local pending independent review and ordinary protected CI.
-  Do not resume #4279 propagation from an unreviewed or unpublished commit.
-
-## 2026-08-11 #4203 kinetics size-budget repair
-
-- The exact `572bf525d` PR tree passes its base-relative candidate scan, but
-  `python scripts/check_file_size_budget.py --max-loc 500 --changed-only
-  --base-ref HEAD~1 --baseline-file scripts/monolith_baseline.txt` reproduces
-  the proactive release failure: Ruff-formatted `simulation/kinetics.py` is
-  selected at 646 LOC and is not grandfathered.
-- A behavior-preserving extraction leaves a 222-LOC stable public façade and
-  delegates immutable result validation (131 LOC) and pure pendulum dynamics
-  (205 LOC) to focused private modules. Public export identity and the existing
-  `_reaction_forces` compatibility seam are test-pinned.
-- Existing numerical parity, inverse/forward dynamics, energy, reaction-force,
-  presentation, and PyQt behavior remain the acceptance basis. This local-only
-  implementation does not change PR base, stack order, protected review/CI
-  requirements, or any downstream issue completion state.
-- Local regression evidence is 28 focused kinetics/presentation/PyQt tests
-  and all 701 Rate-of-Closure Python tests passing.
-
-## 2026-08-11 #4203 pinned-Ruff formatting repair
-
-- No material handoff change: the eight Python files reported by CI Standard
-  run `31468208320`, job `93705508050`, receive only the repository-pinned
-  Ruff 0.14.10 formatting output.
-- No physics, behavior, frame, schema, UI, release-stack, or dependency
-  contract changes. Current-head protected CI and review remain release
-  requirements; queued checks remain unresolved rather than passing.
-
-## 2026-08-10 #4143 child receives repaired launch-registry parent
-
-- Ready PR `#4325` stays on `feat/4143-tee-parity-fixture`, based on
-  `feat/4181-launch-monitor-registry`.
-- Exact parent `12dd76a8dbcc106c4683f2f2e53076f8dc6f1b76` is incorporated by a
-  normal merge commit. There is no production/test-code conflict and no
-  rebase, retarget, force-push, or parent rewrite.
-- Preserve the shared parity fixture and deterministic web/PyQt evidence.
-  Fresh exact-head CI, review, dependency order, and release to `main` remain
-  required before #4143 can close.
-
-## 2026-08-10 #4143 Python/React golden ball-setup parity
-
-- The bounded `feat/4143-tee-parity-fixture` branch starts at exact PR #4203
-  head `31cbc007d4c85b5479b7cd0fb0969124eab2af67`, preserving its draft state,
-  base, and stack order.
-- A single `ball_setup_golden_v1.json` fixture declares schema/version, metre
-  units, the ground-plane-to-ball-bottom reference, ball radius, Driver/Tee and
-  iron/Ground defaults, explicit club-default overrides, Ground zero effective
-  height, center/serialization geometry, invalid finite-domain cases, and a
-  legacy simulation-run migration.
-- Python and React independently consume every case through their public
-  configuration/persistence boundaries. Verification is 18 passing Python
-  tests, 24 passing React tests, and green TypeScript, ESLint, Vite production
-  build, Ruff check, and Ruff format.
-- Recorded visual evidence is stored under
-  `C:\Users\diete\AppData\Local\Temp\rate-4143-visual-evidence-8050eeba`.
-  Playwright captured the 1600 x 1200 default Driver/Tee and rerun
-  explicit-Ground React states after semantic control/diagram and zero-error
-  checks. A hidden 1400 x 900 PyQt harness captured the same states after
-  canonical center, editor, and tee-artist assertions. The browser manifest
-  SHA-256 is `43df78e04b47e1b3209ff7a574718f90847ccda6dde5afd863d43191a950ccf7`;
-  the PyQt manifest SHA-256 is
-  `07822495dbcfa7568615ccb2728481210c28963614434c80f6997210c325a6f9`.
-  PNGs remain external evidence rather than oversized repository binaries.
-- #4143 remains open for protected CI/review and release to `main`. The strict
-  campaign release manifest does not exist in this exact #4203 history; it was
-  added later on a divergent branch and is not backported by this bounded
-  slice.
-
-## 2026-08-10 Second propagation into launch-monitor registry
-
-- Draft child PR `#4203` retains base `feat/4189-dplane` and receives exact
-  repaired parent head `7d8d2f06dc797021d01939691e58f8425b652b33`
-  through a normal merge commit. No branch history, PR base, or draft state is
-  rewritten.
-- The inherited repair closes the parent head's two pinned MyPy
-  `no-any-return` findings with explicit ndarray boundaries and makes no
-  numerical, frame, schema, or UI change.
-- Parent quality-gate success is not child release evidence. Current-head child
-  CI, review, all earlier ancestors, and #4189 acceptance remain required.
-- Reconciled child-tree evidence is 25 focused D-plane/impact tests, docs
-  governance, changed-file size, and whitespace checks. The local Windows
-  MyPy 1.15/installed-NumPy stub combination is incompatible with the branch's
-  Python 3.11 target, while WSL currently fails to start with `E_FAIL`; the
-  successful pinned typing evidence remains the exact parent hosted gate.
-
-## 2026-08-10 Propagation into launch-monitor registry
-
-- Immediate child PR `#4203` keeps base `feat/4189-dplane`.
-- Its original head `08a2fdd8ce6bbc8fbb8f121927a677d4addb6b11`
-  normally merges exact parent `#4202` head
-  `b443fdbed7064c5db0320106013c8413e3e24356`; no branch rewrite, retarget,
-  or force push is permitted.
-- The semantic reconciliation keeps #4203's responsive
-  `SimulationViewControlsMixin`, while the parent's `ImpactLayerControls`
-  helper becomes the single owner of persisted D-plane checkbox state. The
-  automation compatibility mapping aliases that helper state exactly.
-- The launch-monitor registry, Python 3.10 compatibility layer, frame-explicit
-  D-plane contracts, responsive layout, and exports remain additive. Both
-  affected PyQt modules satisfy the protected 500-line limit.
-- The untouched original child had three additional ungrandfathered size
-  blockers: swing sources at 540 LOC, the plotting catalog at 533 LOC, and the
-  main window at 528 LOC. Narrow extractions move triple-pendulum dynamics,
-  plotting metadata, and versioned primary-tab state into focused modules,
-  while identity-pinned re-exports preserve every established import seam. The
-  resulting module pairs are 282/282, 459/98, and 494/85 lines.
-- Focused evidence is 36 passing PyQt simulation/layout tests, 38
-  plotting/navigation tests, and 21 simulation source/export tests. Combined
-  evidence is 1,249 passing Python tests with six explicit optional skips, 521
-  React tests and all web gates, 12 `swing-core` tests, real CPython 3.10
-  checks, scoped static analysis, docs/minimum-test/assertion governance,
-  changed-file size, detect-secrets, and diff checks. The protected size gate
-  passes all 107 changed candidates; a separate full-tree audit retains two
-  untouched non-candidate monoliths (`kinetics.py` and
-  `torque_profile_panel.py`). Independent staged review found no actionable
-  findings after 95 additional focused tests. Current-head protected CI and
-  required repository review remain pending release evidence.
-
-Status verified 2026-08-07. This isolated integration is published as draft
-[PR #4217](https://github.com/D-sorganization/Tools/pull/4217). No source PR
-branch was rewritten.
-## 2026-08-10 PR #4281 workspace timestamp propagation
-
-Exact parent `f90836e342efc8be624739802375af2876d11e5f` is incorporated into
-`feat/4199-wind-scalar-adapter` through the normal merge containing this
-handoff. PR #4281 keeps base `feat/4144-variation-export-continuation`; no
-branch was rebased, retargeted, rewritten, or force-pushed. The child retains
-its scalar ensemble and wind adapter while inheriting the deterministic UTC
-parser and complete variation parent.
-
-The monotonic SPEC now assigns 1.14.10 to compatibility, 1.14.11 to variation,
-and 1.14.12 to the wind child. The reconciled tree passes `794` Rate tests,
-`27` direct compatibility tests on real CPython 3.10.20, and `2 files / 10`
-focused React tests, plus TypeScript, focused zero-warning ESLint, Ruff,
-format, pinned mypy 1.13, documentation, size, and diff gates. Protected CI,
-review, and later propagation remain open.
-
-## 2026-08-09 PR #4281 scalar-ensemble parent propagation
-
-Draft #4281 remains based on `feat/4144-variation-export-continuation` and now
-includes exact corrected #4280 parent
-`38ed58cab96842a3007e76a855db83ee2452b8fd` through the normal merge containing
-this handoff. No branch was rebased, retargeted, force-pushed, or published.
-The child retains its matching Python/TypeScript `scalar-ensemble/v1` contract
-and wind-strategy adapter while inheriting the full variation-export and
-Python 3.10 compatibility history. SPEC 1.14.12 records this child source delta
-independently and monotonically above parent 1.14.11.
-
-SPEC 1.14.13 records the current #4282 propagation monotonically above the
-corrected parent's 1.14.12, 1.14.11, and 1.14.10 entries. Protected CI, review,
-publication, installed-package evidence, downstream UpstreamDrift parity, and
-the remaining scientific/accessibility/performance release gates remain
-separate. This local merge is not a protected release.
-
-Focused evidence is 89 ground-contract, compatibility, scalar-adapter, and
-responsive-wind tests on Python 3.11 plus the same 89 on real CPython 3.10.20.
-Ruff check/format passes 34 focused Python files, pinned mypy 1.13 passes 23
-production modules, and ground modules/functions remain within 400/50-line
-budgets without placeholders. The inherited campaign manifest validates and
-all nine manifest/parity contracts pass. Documentation governance, ancestry,
-SPEC-order, and final diff assertions are required in the same local merge.
-
 ## 2026-08-09 Ground-contract stack recovery
 
 Protected quality-gate run `31341468033` on PR #4285 exact head
@@ -2366,10 +1333,8 @@ an explicit mocked-integrator regression test for that boundary.
 Current exact local evidence is 906 Python/PyQt/shared-swing tests passed with
 one expected optional-Rust skip and 15 existing warnings, plus 91 React test
 files / 555 tests passed. Ruff, formatting, Black, focused mypy, TypeScript,
-zero-warning ESLint, the 166-module production build, and `git diff --check`
-passed. The function budget passed, but the 403-line wind adapter did not meet
-the mandatory below-400 module ceiling at that historical head; the
-2026-08-11 repair above corrects that blocker. The adapter is plot-ready model
+zero-warning ESLint, the 166-module production build, `git diff --check`, and
+the production module/function budgets pass. The adapter is plot-ready model
 infrastructure, not an end-user workflow; #4199 remains open for worker,
 progress/cancellation, client controls, strategy/scatter displays,
 persistence, and exports.
@@ -2764,6 +1729,7 @@ existing local SciPy/NumPy compatibility warning also remains environmental.
 All new source files are below 400 lines. The oversized append-only SPEC and
 handoff registries plus the preserved Waterloo and `from_imperial` public
 signatures predate #4269 and are explicitly retained for compatibility.
+
 ## 2026-08-09 PR #4302 deterministic-digest scanner repair
 
 The protected run at exact PR #4302 head
@@ -2776,477 +1742,3 @@ The digest, fixture, physics, scanner scope, and baseline remain unchanged.
 Commit and push this bounded repair normally before propagating it into #4304,
 #4305, or later ground children. The contemporaneous file-size cancellation
 occurred in checkout and remains infrastructure evidence, not a code failure.
-## 2026-08-11 parent sections preserved during #4282 reconciliation
-
-## 2026-08-10 Exact #4280 propagation into wind scalar PR #4281
-
-- PR `#4281` remains on `feat/4199-wind-scalar-adapter`, based on
-  `feat/4144-variation-export-continuation`.
-- Original child `cf52529b1e68479321bb93b1be3d59c77f782008` normally merges
-  exact parent `8bcc49fc4e16e5e43be0b7f0f03c3017d5b79d0c`; preserve that
-  parent order and do not rebase, retarget, force-push, rewrite a parent, or
-  change draft state.
-- The Python/TypeScript `scalar-ensemble/v1` contract and wind-strategy plot
-  adapter retain stable row identity, provenance, typed unavailable values,
-  and per-cohort availability while inheriting the complete result-local
-  selection, visualization, workspace, compatibility, and module-budget tree.
-  Both implementations enforce the same deterministic scenario boundary:
-  exact per-trial provenance, zero shear/turbulence, seed zero, and no gusts.
-- SPEC 1.14.13 is the unique combined child entry. Verification passes 27
-  focused Python contract/adapter/selection tests, 1,549 complete
-  Rate/shared-swing/golf-club tests with two explicit optional build123d skips,
-  and all 556 React tests across 92 files plus type-check/lint/build. Exact-
-  parent Ruff/format/pinned MyPy/Bandit, the two-file 500-line budget, docs,
-  minimum-test, assertions, detect-secrets, diff, and CPython 3.10 compilation
-  pass. Independent re-review found no actionable findings after checking the
-  exact provenance derivation and all five deterministic-scenario regressions.
-  Protected current-head CI, unresolved threads, and repository review remain
-  release gates after publication.
-
-## 2026-08-11 latest #4203 propagation into workspace PR #4279
-
-- PR `#4279` remains on `feat/4218-toolstrip-workspace`, based on
-  `feat/4181-launch-monitor-registry`; exact parent
-  `0216a547aa79727091a2939b96e779e8ddbd7304` is incorporated by a normal
-  two-parent merge without rebase, retarget, force-push, or history rewrite.
-- There are no feature-code conflicts. The child retains its workspace,
-  toolstrip, module-visibility, navigation, granular playback, and independent
-  plot behavior while inheriting the parent's pinned formatting and
-  identity-preserving 222/205/131-line kinetics split.
-- Preserve both append-only histories below. Fresh merged-tree protected CI,
-  required review, unresolved-thread checks, and dependency order remain
-  release gates; this local merge is not a completion or publication claim.
-
-## 2026-08-10 Second #4203 propagation into workspace PR #4279
-
-- PR `#4279` remains the immediate child on
-  `feat/4218-toolstrip-workspace`, based on
-  `feat/4181-launch-monitor-registry`.
-- Exact repaired parent head
-  `12dd76a8dbcc106c4683f2f2e53076f8dc6f1b76` is incorporated through a
-  normal merge commit; no feature code conflicts and no branch or PR topology
-  is rewritten.
-- Preserve both append-only handoff histories. Fresh child protected CI,
-  review, and earlier dependency gates remain required.
-- Post-reconciliation evidence is 25 focused D-plane/impact tests plus docs
-  governance, changed-file-size, and whitespace gates.
-
-## 2026-08-10 Exact #4203 propagation into workspace PR #4279
-
-- PR `#4279` remains the immediate child on branch
-  `feat/4218-toolstrip-workspace`, based on
-  `feat/4181-launch-monitor-registry`.
-- Original child `05383d333b6fd87eaf5e37305476f50b505c2c2e` normally merges
-  exact parent `31cbc007d4c85b5479b7cd0fb0969124eab2af67`; preserve that
-  two-parent order and do not rebase, retarget, force-push, or change draft
-  state.
-- The child workspace/toolstrip, module-visibility, granular playback, path
-  trail, and independent plot behavior remain authoritative. The parent
-  `ImpactLayerControls`, navigation-state constants, plotting catalog, and
-  triple-pendulum extractions are integrated without duplicate state.
-- Verification is green across 1,339 Python tests with six explicit skips, 545
-  React tests plus type-check/lint/build, the 40-test post-format PyQt rerun,
-  exact-parent Ruff/Ruff-format/Black/pinned MyPy/Bandit, the 500-line budget,
-  docs/minimum-test/assertion/detect-secrets/diff gates, real CPython 3.10.20
-  compile/import checks, and 30 compatibility regressions. There is no Rust
-  delta from the exact parent. Independent staged review found no actionable
-  findings after 76 additional focused PyQt/navigation/workspace tests.
-  Current-head protected CI, unresolved-thread checks, and required repository
-  review remain release gates after guarded publication.
-
-## 2026-08-11 #4203 append-only SPEC preservation repair
-
-- Independent review of local candidate `e20b4f630...` found four exact
-  current-parent history rows missing from SPEC: 2026-08-10 versions 1.13.11,
-  1.13.9, 1.13.7, and 1.13.6. The current documentation-only follow-up restores
-  those rows verbatim and advances the declared/changelog version together.
-- Runtime files, tests, ordered merge parents, PR base, and validation evidence
-  are byte-unchanged. Re-review, normal publication, exact-head protected CI,
-  and downstream propagation remain required.
-
-## 2026-08-11 #4203 current-parent reconciliation
-
-- The current D-plane base is exact head
-  `f3363aa88868f6a5c7e9ccfc682a9eca014e86c1`; #4203's exact published child
-  is `217e36dc93d30f79826847f958fbcd10805e58ed`. They are being composed in
-  that order by a normal merge while retaining base `feat/4189-dplane`.
-- The parent's nine changed files merge without semantic conflict except for
-  formatting around the inherited kinetics origin offset. The split child
-  facade remains authoritative because the obsolete parent expression reaches
-  through `source.inner`; the equivalent explanatory comment is preserved.
-- The exact CI-pinned MyPy 1.13 delta found four additional Qt-stub `Any`
-  boundaries after the merge. …11026 tokens truncated… PR #4217
-was squash-merged there. It adds one
-UI-neutral registry for 17 File/View/Tools commands, a strict versioned
-workspace document with atomic file persistence, matched PyQt/React top
-toolstrips, persistent module visibility/order, theme and shortcut surfaces,
-and direct Impact/Swing/Flight navigation. File actions that do not yet have a
-complete client adapter remain visibly disabled with a reason rather than
-pretending to save incomplete state.
-
-The same continuation corrects the interaction defects reported against the
-live Swing and Plots views. Playback now has deterministic replay-from-end,
-Restart, granular 0.05x through 4.00x speed, pause, and loop behavior. The
-full swing path is opt-in so a persistent trail does not obscure the current
-frame. Each managed plot now owns a distinct figure/canvas, zoom state,
-Auto Fit action, wheel zoom, and independently movable or hideable legend;
-the plot workspace presents all managed plots instead of reusing one selected
-canvas. PyQt small-window testing caught the new playback editor compressing
-below the 64 px readability floor; the explicit editor minimum fixes that case
-and the three-case layout suite passes.
-
-Two read-only cross-repository audits are now tracked as separate programs:
-
-- [#4260](https://github.com/D-sorganization/Tools/issues/4260), with
-  #4261-#4266, establishes one impact/flight authority and a machine-readable
-  parity contract across Tools PyQt, Tools React, UpstreamDrift PyQt, and
-  UpstreamDrift React.
-- [#4267](https://github.com/D-sorganization/Tools/issues/4267), with
-  #4268-#4276, defines qualified landing, bounce, skid, roll, and total-distance
-  modeling with editable ground profiles and exact UpstreamDrift adapters.
-
-The parity audit found that UpstreamDrift PyQt reuses Tools, while the
-UpstreamDrift React launcher has no native Rate React route. UpstreamDrift's
-Tools gitlink `ff4240217005e1415ca409fd124e50b64ee642d2` also predates the
-current integration head by 184 commits, and its sibling/vendor resolution is
-ambiguous. The ground audit found a useful existing fail-closed
-`GroundModelResult` boundary plus reusable putting/terrain primitives, but no
-qualified end-to-end ground solver. Before bounce can be correct, airborne
-flight must terminate against physical terrain plus ball radius and preserve
-the full terminal angular-velocity vector; the current relative launch-plane
-event and spin-free trajectory state do neither. Those prerequisites are
-explicit in #4269 and must not be hidden by UI-derived estimates.
-
-The final local verification pass is green. The complete Rate-of-Closure and
-shared swing-model run passed 890 tests with one expected skip because the
-optional `swing_core` Rust wheel is not installed; the remaining 15 warnings
-are the existing Hypothesis collection warning. React passed 89 files / 545
-tests, zero-warning ESLint, TypeScript checking, and the production Vite build.
-Ruff, Black, targeted mypy, `git diff --check`, and the repository structural
-limits also pass: every changed production Python file is at most 400 lines and
-every changed production Python function is at most 50 lines. Rendered PyQt
-inspection confirmed independent plot canvases, responsive single-column
-reflow at the tested desktop width, independent 125%/100% zoom state, working
-Auto Fit, and the opt-in trail/playback controls. These are local validation
-results only; they do not establish protected CI, review, merge, or release
-status.
-
-### 2026-08-07 variation export and completion audit continuation
-
-The post-toolstrip branch `feat/4144-variation-export-continuation` is published
-as [draft PR #4280](https://github.com/D-sorganization/Tools/pull/4280), based
-on exact parent head `c36ca36e91f34fa849d2508708bf9dd6c0cdc392`. It keeps #4279 unchanged
-while closing one remaining #4144 parity gap: selected scalar scatter data can
-now be exported as CSV from both clients, retaining every raw trial, typed
-outcome, and unavailable cell rather than only the finite points drawn on the
-canvas. PyQt also has a bounded read-only raw-trial table matching the web
-workflow, and the table population is shared with the matrix view.
-
-The complete post-change local gates passed:
-
-- Python/PyQt/shared swing suite: `890 passed, 1 skipped, 15 warnings`; the
-  skip is the optional `swing_core` wheel and the warnings are the existing
-  Hypothesis collection and empty polynomial-preview legend warnings.
-- React: `89` files / `545` tests passed.
-- Ruff check/format, Black, targeted mypy, TypeScript, zero-warning ESLint,
-  the `166`-module Vite production build, and `git diff --check` passed.
-- Every changed production file is below 400 lines and every changed
-  production function is at most 50 lines.
-
-A live GitHub/source reconciliation covered every requested epic in this
-campaign. No epic yet satisfies its own definition of done: most implementation
-is still on feature branches, #4119 is the only Rate platform PR targeting
-`main` and is currently dirty, #4203 and #4279 remain draft/unstable, and only
-formal club-builder child #4147 is closed. The variation request is
-substantively implemented, but #4142/#4144 remain open because bounded
-large-ensemble execution, nonlinear global sensitivity, localized execution,
-the immutable UpstreamDrift consumer pin, protected CI, and default-branch
-release are incomplete.
-
-The literal universal-runner audit also found two uncovered many-evaluation
-paths. Wind strategy analysis retains all paired outcomes but has no user
-workflow or universal plot adapter; capability optimization retains aggregates
-but not individual sample rows. The next safe model slice is a UI-neutral
-scalar-ensemble contract with unique composite row IDs, unit-bearing variable
-metadata, caller-defined cohorts, paired-finite scatter extraction, and exact
-availability accounting. Wind integration must accept both its immutable
-request and analysis so launch definitions and provenance are not inferred.
-Issue #4199 already owns the required controls, scatter, strategy table,
-progress/cancellation, and export workflow.
-
-The first narrow #4199 implementation slice is published as
-[draft PR #4281](https://github.com/D-sorganization/Tools/pull/4281) from branch
-`feat/4199-wind-scalar-adapter`, stacked on exact PR #4280 head
-`d71b0ea01b5659d3049ff05627c41f06481207e4`. Implementation commit
-`4a28114aa` introduces an exact
-cross-runtime `scalar-ensemble/v1` wire contract and pure wind-strategy
-adapters. The contract preserves structured provenance, unit-bearing variable
-definitions, caller-defined cohorts, RFC3986 composite identities, nullable
-raw rows, and exact scatter availability. The adapters validate the immutable
-request against the stored paired analysis, preserve completed,
-nonconverged, and invalid outcomes, and never invoke a flight model. React has
-an explicit mocked-integrator regression test for that boundary.
-
-Current exact local evidence is 906 Python/PyQt/shared-swing tests passed with
-one expected optional-Rust skip and 15 existing warnings, plus 91 React test
-files / 555 tests passed. Ruff, formatting, Black, focused mypy, TypeScript,
-zero-warning ESLint, the 166-module production build, and `git diff --check`
-passed. The function budget passed, but the 403-line wind adapter did not meet
-the mandatory below-400 module ceiling at that historical head; the
-2026-08-11 repair above corrects that blocker. The adapter is plot-ready model
-infrastructure, not an end-user workflow; #4199 remains open for worker,
-progress/cancellation, client controls, strategy/scatter displays,
-persistence, and exports.
-
-### 2026-08-07 ground and four-surface audit refinement
-
-The rolling-ground and cross-application parity requests remain tracked by the
-existing [ground epic #4267](https://github.com/D-sorganization/Tools/issues/4267)
-and [parity epic #4260](https://github.com/D-sorganization/Tools/issues/4260);
-no duplicate epic or child issue is required. The latest exact-path audit and
-acceptance refinements are attached to
-[the ground epic](https://github.com/D-sorganization/Tools/issues/4267#issuecomment-5222725556)
-and [the parity epic](https://github.com/D-sorganization/Tools/issues/4260#issuecomment-5222726010).
-
-The scientific implementation order is contractual: #4268 defines the
-surface/contact/trajectory/result transfer state, then #4269 corrects physical
-terrain contact and preserves terminal full angular velocity. Only then may
-#4270/#4271 qualify the 3D impulse, repeated bounce, skid, and pure-roll
-phases. Carry remains first physical contact. Final downrange, final lateral,
-horizontal displacement, surface path length, and launch-monitor-style total
-distance are distinct quantities; no implementation may silently assume
-`total distance = carry + roll distance`.
-
-Reusable UpstreamDrift scope is deliberately narrow: its split terrain
-material/elevation/normal/region package can feed a one-way versioned DTO
-adapter. Current scalar landing, heuristic putting-roll, duplicate legacy
-terrain, and Rust tangential-loss implementations are reference material, not
-the qualified physics authority. Upstream surface defaults remain illustrative
-until citations, calibration, uncertainty, and applicability are recorded.
-
-The parity matrix must distinguish seven product identities: standalone Rate
-PyQt6 and React, the Upstream Rate PyQt provider and React route, Upstream Shot
-Tracer PyQt6 and React, and the legacy Upstream ball-flight GUI. Current
-Upstream `main` (`0782853295e005af68818617e4725eb980890f43`) pins Tools at
-`ff4240217005e1415ca409fd124e50b64ee642d2`, exposes no native Rate React route,
-and contains contradictory vendor-first and sibling-first Tools resolvers.
-These facts are current audit evidence, not completion; #4260, #4267, and all
-children remain open.
-
-## 2026-08-10 PR #4202 D-plane ndarray typing repair
-
-- The repair is based on exact published PR `#4202` head
-  `b443fdbed7064c5db0320106013c8413e3e24356` and retains base
-  `feat/4162-wedge-impact-visualization`.
-- CI Standard run `31384810375`, job `93442745760`, reported two pinned MyPy
-  1.13 `no-any-return` errors in private NumPy conversion/projection helpers.
-  Explicit ndarray local boundaries resolve those errors without changing
-  geometry, validation, frames, serialized contracts, or UI behavior.
-- RED/GREEN evidence is exact: two errors reproduced before the edit and zero
-  afterward. Twenty-four focused D-plane/impact tests, seven metadata/pre-push
-  contract tests, scoped Ruff/Ruff-format/Black, docs governance, minimum-test,
-  module-size, changed-file-size, and diff checks pass. Three exploratory
-  CI-workflow contract tests retain unrelated older-branch toolcache/env drift;
-  no workflow file is changed here.
-- This repair does not alter the stacked base, publish the branch, change its
-  draft state, or authorize a merge. Parent-first ordering, protected CI, and
-  review remain required.
-
-## 2026-08-10 Propagation into 3D D-plane geometry
-
-- Immediate child PR `#4202` keeps base
-  `feat/4162-wedge-impact-visualization`.
-- Its original head `b4abec03bccfbdd87ddf91427159c5c2332c21dd`
-  normally merges exact parent `#4179` head
-  `6704a3e541a3e74c28b4a284530d1a21269dd340`; no branch rewrite, retarget, or
-  force push is permitted.
-- The Python 3.10 UTC repair and AST guard are inherited alongside the typed,
-  frame-explicit D-plane, spin-loft, visualization, and export contracts.
-- Persisted D-plane layer controls are extracted into a focused helper so the
-  simulation view satisfies the protected 500-line module budget without a
-  behavior or compatibility-seam change.
-- Combined-stack verification is green: 93 focused and 825 scoped Python tests
-  (two optional `build123d` skips), 360 React tests and all web gates, real
-  CPython 3.10.20 compilation/UTC, scoped Ruff/Black/MyPy, and repository
-  governance gates. The exact parent's 12 unchanged `swing-core` tests remain
-  applicable because this child has no Rust delta. The inherited 17-error broad
-  MyPy Qt/NumPy baseline in 11 untouched files remains outside scope.
-  Protected CI and required review remain release gates.
-
-## 2026-08-10 Propagation into wedge impact visualization
-
-- Immediate child PR `#4179` keeps base `feat/4166-wedge-turf-physics`.
-- Its original head `0eb804e70887c788421332369e42792411aff55a`
-  normally merges exact parent `#4178` head
-  `bfa83aedc88ead380babc73a699377d98b971006`; no branch rewrite, retarget, or
-  force push is permitted.
-- The Python 3.10 UTC repair and AST guard are inherited alongside the
-  exact-event, locked-scale, exportable impact-scene contract.
-- Combined-stack verification is green: 58 focused and 739 scoped Python tests
-  (two optional `build123d` skips), 347 React tests and all web gates, real
-  CPython 3.10.20 compilation/UTC, scoped Ruff/Black/MyPy, and repository
-  governance gates. The exact parent's 12 unchanged `swing-core` tests remain
-  applicable because this child has no Rust delta. The inherited 17-error broad
-  MyPy Qt/NumPy baseline in 11 untouched files remains outside scope.
-  Protected CI and required review remain release gates.
-
-## 2026-08-10 Propagation into wedge turf physics
-
-- Immediate child PR `#4178` keeps base `feat/4161-wedge-ground-clearance`.
-- Its original head `aaae3f73e17dbfaad5cca1dc6f49559b3aebe9d5`
-  normally merges exact parent `#4174` head
-  `9ea93e92563280ec34bca682ad44d7409edd7a02`; no branch rewrite, retarget, or
-  force push is permitted.
-- The Python 3.10 UTC repair and AST guard are inherited alongside the passive,
-  provenance-gated turf-contact model and explicit force-coupling boundary.
-- Combined-stack verification is green: 56 focused and 732 scoped Python tests
-  (two optional CAD-dependency skips), real CPython 3.10.20 checks, scoped
-  static analysis, and repository governance gates. With no web or Rust delta,
-  the exact parent's green 345 React and 12 Rust tests remain applicable. The
-  inherited 17-error broad MyPy Qt/NumPy baseline in 11 untouched files remains
-  outside scope.
-
-## 2026-08-10 Propagation into wedge ground clearance
-
-- Immediate child PR `#4174` keeps base `feat/4163-impact-inspector`.
-- Its original head `880a6465fc872cf3d6650283db154ddc41793a31`
-  normally merges exact parent `#4173` head
-  `9ddaff3b6bca542fd7a2befc7d7b0ae53910a60a`; no branch rewrite, retarget, or
-  force push is permitted.
-- The Python 3.10 UTC repair and AST guard are inherited alongside the swept
-  wedge ground-clearance model, persistence, PyQt, and React surfaces.
-- Combined-stack verification is green: 56 focused and 703 scoped Python tests
-  (two optional CAD-dependency skips), 345 React tests and all web gates, 12
-  Rust tests, real CPython 3.10.20 checks, scoped static analysis, and
-  repository governance gates. The inherited 17-error broad MyPy Qt/NumPy
-  baseline in 11 untouched files remains outside scope.
-
-## 2026-08-10 Propagation into impact inspector
-
-- Immediate child PR `#4173` keeps base `feat/4144-variation-visualizations`.
-- Its original head `3c43955aaeb3964ff8c3ef2748d626baae518b76`
-  normally merges exact parent `#4167` head
-  `22b66b560652b78de84141344c4ddd9a92a83b26`; no branch rewrite, retarget, or
-  force push is permitted.
-- The Python 3.10 UTC compatibility repair and its AST guard are inherited
-  additively alongside the existing shared wedge impact-inspector contract.
-- Combined-stack verification is green: 63 focused and 562 total Rate Python
-  tests; 334 React tests plus type-check, lint, and production build; 12 Rust
-  tests; real CPython 3.10.20 compile/UTC checks; scoped static analysis and
-  repository governance gates. The broad MyPy sweep retains 17 pre-existing
-  Qt/NumPy typing findings in 11 untouched files. Current-head protected CI and
-  required review remain pending release evidence.
-
-## Dependency position
-
-PR `#4167` (`feat/4144-variation-visualizations`) is the base-most open Rate
-feature above the already merged `feat/investigation-suite` carrier. Later
-wedge, D-plane, launch-monitor, workspace, wind, capability, and ground work
-depends on this line and must receive any repair through ordinary parent
-propagation; child branches must not be rewritten.
-
-## 2026-08-10 Python 3.10 repair
-
-- Protected CI at exact pre-repair head
-  `edaa56358a9ccf47809533fcab28e6415b336771` collected 13 Rate test modules
-  unsuccessfully because `datetime.UTC` does not exist on Python 3.10.
-- The torque-profile controller now consumes the repository's existing
-  `shared.python.compatibility.UTC` export.
-- A source-tree AST guard rejects future direct imports and unaliased or
-  aliased `datetime.UTC` module-attribute access anywhere under
-  `src/rate_of_closure`.
-- Local evidence is green: 27 focused controller/history/AST tests and the
-  complete 554-test Rate suite on Python 3.13; real CPython 3.10.20 compatibility
-  import; Ruff check/format; focused pinned MyPy 1.13; detect-secrets;
-  touched-file size and diff checks.
-
-## Truthful release state
-
-This is an actionable compatibility repair, not completion of issue `#4144` or
-the variation epic. Current-head protected CI, required review, dependency
-propagation, and ordinary merges remain required. Runner download timeouts,
-missing toolcache/link libraries, cancelled jobs, and queued jobs are tracked
-as infrastructure and are never counted as green evidence.
-
-Every implementation commit must update this file, both other canonical
-handoffs, and `SPEC.md`, or explicitly record no material handoff change and
-the reason.
-
-## 2026-08-11 additional parent sections preserved during #4282 reconciliation
-
-## 2026-08-09 PR #4279 Python 3.10 compatibility completion
-
-Exact corrected #4203 parent
-`08a2fdd8ce6bbc8fbb8f121927a677d4addb6b11` is incorporated into #4279 by
-normal local merge `a340fabefa443d47325c5538f342683b38c01ade`; the stack base is
-unchanged and nothing was rebased, retargeted, force-pushed, or published.
-The two child-owned string-enum modules and child workspace UTC validator now
-use the existing shared Python 3.10 compatibility contract at runtime. Native
-enum typing remains explicit under `TYPE_CHECKING`, and wire IDs, schemas,
-timestamp serialization, physics, and UI behavior are unchanged.
-
-The reconciled test covers nine `StrEnum` runtime imports, both `UTC` runtime
-imports, and direct execution of all three child workspace modules. Python
-3.11 passes `126` focused tests. Real CPython 3.10.20 passes all `14`
-compatibility cases and imports 10 dotted Rate/shared targets. Ruff
-check/format and pinned mypy 1.13 pass all affected files; documentation
-governance and diff checks are required in the same commit. Protected CI,
-review, publication, and normal propagation to #4280 and later descendants
-remain outstanding.
-
-## 2026-08-09 PR #4279 parent propagation
-
-The #4279 continuation remains based on `feat/4181-launch-monitor-registry` and
-now contains exact parent head
-`08a2fdd8ce6bbc8fbb8f121927a677d4addb6b11` through a normal merge. Source
-changes applied cleanly; the overlapping handoff/specification histories were
-reconciled monotonically. The result retains the parent's Linux-safe facade
-and Python 3.10 compatibility contracts plus the child's workspace,
-top-toolstrip, module-navigation, playback, and independent plot controls. No
-branch was rebased, retargeted, force-pushed, or published by this continuation.
-
-Current focused evidence is `126 passed` for both facade modules and all
-Python/PyQt tests changed by #4279, plus `8 React files / 32 tests passed` for
-the matching workspace surfaces. Ruff check/format is clean across all 28
-relevant Python files. CI-pinned mypy 1.13 is clean across the 18 changed
-production files and both facade tests after making the child's
-`legend_visible()` Qt boundary return an explicit `bool`. Documentation
-governance, staged/unstaged diff checks, and the affected `29`-test simulation
-GUI rerun pass. Protected CI, review, and publication remain separate release
-requirements.
-
-## 2026-08-09 Python 3.10 compatibility follow-up
-
-The descendant Python 3.10 lane revealed seven unconditional
-`enum.StrEnum` imports already present at PR #4203. The earliest owning parent
-now routes runtime imports through `shared.python.compatibility.StrEnum` while
-using the stdlib type only under `TYPE_CHECKING`. The change preserves every
-enum value and schema while removing the Python 3.11-only import boundary.
-Local evidence is 64 focused contract/physics/compatibility tests, Ruff and
-format, pinned mypy 1.13 across eight changed files, and a real CPython 3.10.20
-probe of the shared fallback and seven module import declarations. Parent is
-published at exact head `ab7de5a47977417e02926c3fbc7476002e82b690`;
-propagate it through #4279, #4280, #4281, and #4282 before retrying descendants.
-
-The same scan found one parent-owned `datetime.UTC` import in the PyQt torque
-profile controller. It now consumes the established compatibility constant,
-preserving UTC serialization while allowing Python 3.10 collection. This
-follow-up belongs on #4203 before the child stack is propagated.
-
-## 2026-08-09 Launch-registry parent CI repair
-
-The earliest campaign parent, draft PR #4203, remains based on
-`feat/4189-dplane` at exact published head
-`912ebc9d69b05763a76c2c8f198d943737e2d3fb`. CI run `31199764932` passed its
-quality gate, then all Python versions failed before assertions because pytest
-collected the flight and solver contract modules through `src.shared...` while
-their absolute dotted facade aliases requested the editable `shared...`
-namespace. The scoped repair converts only those two test imports to relative
-package imports. Production APIs and physics are unchanged. The same run's
-Rust failure is the known runner inability to link `-lpython3.11`. Validate
-the two contract modules under importlib collection, publish a new exact head,
-and propagate it through #4279, #4280, #4281, and #4282 in normal order.
-Both modules now pass all `12` tests on Windows and WSL Python 3.11 with
-importlib collection. Ruff check/format and pinned mypy 1.13 are clean. A
-test-only `Any` cast keeps frozen-dataclass metadata introspection explicit
-without weakening the runtime contract.
diff --git a/docs/release/rate_of_closure_campaign.v1.json b/docs/release/rate_of_closure_campaign.v1.json
index 1f60a6630..133582e1f 100644
--- a/docs/release/rate_of_closure_campaign.v1.json
+++ b/docs/release/rate_of_closure_campaign.v1.json
@@ -134,7 +134,7 @@
       "pr": 4282,
       "branch": "feat/4199-wind-workflow",
       "base_branch": "feat/4199-wind-scalar-adapter",
-      "head_sha": "b2baf6ef615b0d756a86ac4ca7eef2fc583210ee",
+      "head_sha": "18fe89201d657116bbca99922297c14968356c44",
       "state": "open",
       "merge_commit_sha": null,
       "protection_evidence": "not_recorded"
@@ -154,7 +154,7 @@
       "pr": 4285,
       "branch": "feat/4268-ground-contract",
       "base_branch": "feat/4197-capability-observer",
-      "head_sha": "584e4e53a7990053a346938c4e45538786887ec7",
+      "head_sha": "6a2bc9d06f6f9a28a0d615b19d2ed4fc13871059",
       "state": "open",
       "merge_commit_sha": null,
       "protection_evidence": "not_recorded"
@@ -164,7 +164,7 @@
       "pr": 4288,
       "branch": "feat/4269-flight-ground-transfer",
       "base_branch": "feat/4268-ground-contract",
-      "head_sha": "c39297d37a17d5d8d3520ed62c5563cdcc609cab",
+      "head_sha": "4972e55e0bb6e5b6bf7da0f899eed5d4f54e7d9d",
       "state": "open",
       "merge_commit_sha": null,
       "protection_evidence": "not_recorded"
@@ -194,7 +194,7 @@
       "pr": 4302,
       "branch": "feat/4270-ground-impact-bounce",
       "base_branch": "feat/4269-flight-ground-transfer",
-      "head_sha": "62beb3e1aa951645fd556a53f9cbae4bb46c47e5",
+      "head_sha": "cf6e72bad98e5f36f782254942e6895b8b71e670",
       "state": "open",
       "merge_commit_sha": null,
       "protection_evidence": "not_recorded"
@@ -204,7 +204,7 @@
       "pr": 4304,
       "branch": "feat/4271-ground-skid-roll",
       "base_branch": "feat/4270-ground-impact-bounce",
-      "head_sha": "3f861c1faaa7455faee92eaa7f813174667208e0",
+      "head_sha": "0ea6740965068542e9d8c7449e06ec07d88969e0",
       "state": "open",
       "merge_commit_sha": null,
       "protection_evidence": "not_recorded"
@@ -214,7 +214,7 @@
       "pr": 4332,
       "branch": "feat/4271-regional-surface-transitions",
       "base_branch": "feat/4271-ground-skid-roll",
-      "head_sha": "5ac3c1f233de30dd42eeac57073cbdb149bde652",
+      "head_sha": "04ccf08dd990de1cd056a3420e67772773a4be2e",
       "state": "open",
       "merge_commit_sha": null,
       "protection_evidence": "not_recorded"
@@ -224,7 +224,7 @@
       "pr": 4335,
       "branch": "feat/4271-regional-wire-contract",
       "base_branch": "feat/4271-regional-surface-transitions",
-      "head_sha": "8f933ed8dcb29e55ece4ec6bb1e60813f6794d57",
+      "head_sha": "9e01ccc3e891cc45907293751a192624195a77a5",
       "state": "open",
       "merge_commit_sha": null,
       "protection_evidence": "not_recorded"
@@ -234,7 +234,17 @@
       "pr": 4339,
       "branch": "feat/4274-regional-surface-ui",
       "base_branch": "feat/4271-regional-wire-contract",
-      "head_sha": "d21741e312b849a63f73cabf351a15d9de80fb94",
+      "head_sha": "db335937afc4b587d235eb705e315f577519c5e6",
+      "state": "open",
+      "merge_commit_sha": null,
+      "protection_evidence": "not_recorded"
+    },
+    {
+      "id": "pr-4342",
+      "pr": 4342,
+      "branch": "feat/4274-regional-plan-io",
+      "base_branch": "feat/4274-regional-surface-ui",
+      "head_sha": "c1f47f2ef68b3db102da5416aaac17a40f675207",
       "state": "open",
       "merge_commit_sha": null,
       "protection_evidence": "not_recorded"
@@ -423,6 +433,42 @@
         "src/rate_of_closure/AGENT_HANDOFF.md",
         "docs/development/RATE_OF_CLOSURE_CAMPAIGN_HANDOFF.md"
       ]
+    },
+    {
+      "id": "ground-regional-ui-helper-local-d217",
+      "kind": "local",
+      "commit_sha": "d21741e312b849a63f73cabf351a15d9de80fb94",
+      "outcome": "passed",
+      "commands": [
+        "python -m pytest tests/rate_of_closure/test_regional_surface_plan.py tests/rate_of_closure/test_gui.py tests/rate_of_closure/test_layout_minsize.py -q",
+        "npm test -- --run",
+        "npm run type-check && npm run lint && npm run build",
+        "uvx --from ruff==0.14.10 ruff check src/rate_of_closure tests/rate_of_closure && uvx --from ruff==0.14.10 ruff format --check src/rate_of_closure tests/rate_of_closure"
+      ],
+      "summary": "The current PR #4339 parent passed focused Python/PyQt and React qualification plus the complete 672-test React suite after replacing its six-parameter number-input helper with a frozen specification and three-parameter helper. This remains local evidence, not protected release evidence.",
+      "source_paths": [
+        "AGENT_HANDOFF.md",
+        "src/rate_of_closure/AGENT_HANDOFF.md",
+        "docs/development/RATE_OF_CLOSURE_CAMPAIGN_HANDOFF.md"
+      ]
+    },
+    {
+      "id": "ground-regional-io-local-8e1c",
+      "kind": "local",
+      "commit_sha": "8e1c7ccd99a7c4886c5fb9ccc7e4d94a6d7e3833",
+      "outcome": "passed",
+      "commands": [
+        "python -m pytest tests/rate_of_closure/test_regional_surface_plan.py tests/rate_of_closure/test_regional_surface_plan_io.py tests/rate_of_closure/test_workspace_files.py -q",
+        "npm test -- --run src/components/RegionalSurfacePlanPanel.test.tsx",
+        "npm run type-check && npm run lint && npm run build",
+        "uvx --from ruff==0.14.10 ruff check src/rate_of_closure tests/rate_of_closure && uvx --from ruff==0.14.10 ruff format --check src/rate_of_closure tests/rate_of_closure"
+      ],
+      "summary": "The published PR #4342 child passed strict native atomic and bounded browser regional request I/O qualification, safe-number parity, bounded UTF-8 import, static analysis, build, and governance gates before parent reconciliation. Fresh protected evidence remains open.",
+      "source_paths": [
+        "AGENT_HANDOFF.md",
+        "src/rate_of_closure/AGENT_HANDOFF.md",
+        "docs/development/RATE_OF_CLOSURE_CAMPAIGN_HANDOFF.md"
+      ]
     }
   ],
   "programs": [
@@ -1204,7 +1250,8 @@
         "pr-4304",
         "pr-4332",
         "pr-4335",
-        "pr-4339"
+        "pr-4339",
+        "pr-4342"
       ],
       "supported_surfaces": {
         "tools.pyqt6": "implemented_unverified",
@@ -1217,14 +1264,16 @@
         "ground-skid-roll-local-f475",
         "ground-regional-local-1a48",
         "ground-regional-wire-local-74a0",
-        "ground-regional-ui-local-cbb9"
+        "ground-regional-ui-local-cbb9",
+        "ground-regional-ui-helper-local-d217",
+        "ground-regional-io-local-8e1c"
       ],
-      "evidence_gap": "The contract, flight-transfer, impact/bounce, static-plane skid/roll, regional-surface, regional-plan wire, and regional-editor carriers remain open on draft PRs 4285, 4288, 4302, 4304, 4332, 4335, and 4339. Exact local evidence exists, but fresh protected evidence for the reconciled heads, execution/playback, result interchange, workspace model-input persistence, changing geometry, compiled-runtime physics, and downstream parity remain absent.",
+      "evidence_gap": "The contract, flight-transfer, impact/bounce, static-plane skid/roll, regional-surface, regional-plan wire, regional-editor, and regional request-I/O carriers remain open on draft PRs 4285, 4288, 4302, 4304, 4332, 4335, 4339, and 4342. Exact local evidence exists, but fresh protected evidence for the reconciled heads, execution/playback, result interchange, workspace model-input persistence, measured calibration, changing geometry, compiled-runtime physics, and downstream parity remain absent.",
       "limitations": [
         "Airborne flight must first terminate against physical terrain and preserve terminal angular velocity.",
         "Current UpstreamDrift terrain and putting implementations are reference inputs rather than the qualified authority.",
         "The #4270 rigid impact law does not consume firmness, grass, compressibility, moisture, or rolling-resistance fields.",
-        "Published open PR #4304 qualifies one immutable planar profile and optional finite tangent-axis edge; open PR #4332 adds bounded coplanar material overlays; open PR #4335 adds a strict Python/TypeScript plan wire boundary and Python resolver; and open PR #4339 adds matched session-only PyQt6/React request editors. Changing normals, height or velocity discontinuities, terrain deformation, torsional-spin damping, roll-to-skid transitions, internal transition-ledger export, execution/playback, result interchange, measured calibration, workspace model-input persistence, compiled or TypeScript regional physics, and downstream adapters remain unimplemented."
+        "Published open PR #4304 qualifies one immutable planar profile and optional finite tangent-axis edge; open PR #4332 adds bounded coplanar material overlays; open PR #4335 adds a strict Python/TypeScript plan wire boundary and Python resolver; open PR #4339 adds matched PyQt6/React request editors; and open PR #4342 adds editor-qualified canonical request import/export with native atomic semantics and explicit browser limitations. Changing normals, height or velocity discontinuities, terrain deformation, torsional-spin damping, roll-to-skid transitions, internal transition-ledger export, execution/playback, result interchange, measured calibration, workspace model-input persistence, compiled or TypeScript regional physics, and downstream adapters remain unimplemented."
       ],
       "depends_on_issues": [
         4191
diff --git a/docs/specs/GROUND_SKID_ROLL.md b/docs/specs/GROUND_SKID_ROLL.md
index c8b7e1a28..1e39b2750 100644
--- a/docs/specs/GROUND_SKID_ROLL.md
+++ b/docs/specs/GROUND_SKID_ROLL.md
@@ -203,11 +203,23 @@ schema, unit, provenance, and request readback.
 
 The regional v1 wire schema has no calibration record, so the editor does not
 invent one or widen that schema. `unvalidated` is explicit presentation/source
-qualification and is included in the source digest. This first slice is
-session-only: current web workspace file commands do not persist model inputs,
-and neither client claims request export/import, execution, result playback,
-or measured-course calibration. Those capabilities require separate contracts
-and acceptance evidence.
+qualification and is included in the source digest. PyQt6 native Open/Save As
+and React browser import/download persist only the canonical regional request;
+workspace model-input persistence remains a separate contract. Import is
+transactional and accepts only the editor producer/provider v1, fixed qualified
+axis/geometry, and editor row capacity. An unchanged import retains the exact
+request and provenance; editing rebinds the draft digest. Browser downloads
+cannot promise a native path, atomic replacement, or recent-file access.
+Wire numbers are bounded to the shared cross-runtime safe range. Native
+precedence entry preserves every nonnegative integer through
+9,007,199,254,740,991 exactly, so a qualified import cannot be silently narrowed
+before validation or Save As.
+Native import reads one binary handle with a one-byte overflow sentinel before
+strict UTF-8 decoding, so mutable files cannot bypass the 1 MiB allocation cap
+between a metadata check and content parsing.
+Neither client claims execution, result playback, or measured-course
+calibration. Those capabilities require separate contracts and acceptance
+evidence.
 
 Skid and roll distances are accumulated separately from centre speed relative
 to the moving surface. Collinear constant-acceleration segments use the exact
diff --git a/src/rate_of_closure/AGENT_HANDOFF.md b/src/rate_of_closure/AGENT_HANDOFF.md
index 52a1d5c5f..d858b22d7 100644
--- a/src/rate_of_closure/AGENT_HANDOFF.md
+++ b/src/rate_of_closure/AGENT_HANDOFF.md
@@ -1,5 +1,38 @@
 # AGENT_HANDOFF — rate_of_closure
 
+## 2026-08-11 PR #4342 current-parent propagation candidate
+
+This local no-publish merge keeps `feat/4274-regional-plan-io` over
+`feat/4274-regional-surface-ui`. Exact current child
+`c1f47f2ef68b3db102da5416aaac17a40f675207` is the first parent and exact
+reviewed local #4339 candidate `db335937afc4b587d235eb705e315f577519c5e6`
+is the second parent. Child-owned canonical request import/export, bounded UTF-8,
+native atomic save, browser-qualified download, tests, and limitations remain
+intact while inheriting current editor, wire, regional physics, and complete
+ground ancestry.
+
+The default shared canonical encoder continues to reject floats and integers
+outside JavaScript's safe range. Only the capability-observation facade selects
+a separately named extended finite-float policy. It shares the same recursive
+encoder, retains safe-range integer checks, emits exact exponent-free `1e20`
+and `1e21` tokens matching TypeScript, and rejects non-finite values.
+
+Local qualification passes all `909` Rate-of-Closure Python tests, all `110`
+React files / `686` tests, the complete Cargo workspace, `47` focused Python
+compatibility/regional-I/O tests, and `12` focused React capability tests.
+Pinned Ruff 0.14.10 check/format passes `17` changed Python files; pinned MyPy
+1.13 and Bandit pass `12` changed production files. TypeScript, zero-warning
+ESLint, the 199-module Vite build, Rust format/warning-denied clippy, 400- and
+500-LOC gates, manifest/docs/blocking-gate/assertion/minimum-test governance,
+marker scans, and diff checks pass. One untouched manual-delivery UI test timed
+out in the first concurrent full run, then passed alone and in the single
+complete rerun.
+
+The candidate is not published or released. Execution/playback, result
+interchange, measured calibration, persistence, changing geometry or velocity,
+TypeScript/compiled regional physics, downstream parity, protected exact-head
+evidence, review, approval, dependency integration, and release remain open.
+
 ## 2026-08-11 PR #4339 current-parent propagation candidate
 
 This local no-publish merge keeps `feat/4274-regional-surface-ui` over
@@ -28,184 +61,16 @@ interchange, measured calibration, persistence, changing geometry or velocity,
 TypeScript/compiled regional physics, downstream parity, protected exact-head
 evidence, review, approval, dependency integration, and release remain open.
 
-## 2026-08-11 PR #4339 structural helper repair candidate
-
-This local follow-up descends exact independently reviewed reconciliation
-candidate `eedad6d23b517eb4c99d4ba9000ff6555101099f`. Independent review found
-one publication blocker: the regional PyQt number-input helper accepted six
-parameters, above the campaign's mandatory four-parameter limit. RED-first
-tests now bind the helper's signature and all observable widget settings. A
-frozen `NumberInputSpec` validates finite positive step and finite ordered
-bounds before widget construction; the three-parameter helper preserves field
-names, values, units, ranges, increments, tooltips, and presentation order.
-
-Eight focused Python/PyQt tests and 25 focused React tests pass; the complete
-React suite passes all 108 files / 672 tests. Pinned Ruff check/format, MyPy on
-all six affected production modules, Bandit medium/high screening, TypeScript,
-zero-warning ESLint, documentation/manifest governance, changed-production,
-test-assertion, minimum-test, 500-LOC, and diff gates pass. Mandatory maxima
-are 400 module lines, 42 function lines, and three parameters. A broader
-Rate-of-Closure Python run reached the ten-minute workstation lifecycle bound
-without yielding an assertion failure; it is not counted as passing evidence.
-The monorepo-wide lane separately exposed unrelated baseline failures and
-xdist worker termination. No schema, digest, regional physics, UI behavior,
-PR base, protected evidence, or release boundary changes.
-
-## 2026-08-11 PR #4339 current-parent ancestry candidate
-
-The clean `feat/4274-regional-surface-ui` worktree normally merges exact
-published child `cbb9c0a6bdc6a50f59f7a661139b9d53e1892980` with exact newly
-published #4335 parent `9e01ccc3e891cc45907293751a192624195a77a5`,
-without changing #4339's base. The accessible PyQt6/React editors and strict
-invalidation behavior remain intact alongside the parent's wire, resolver,
-regional-physics, and complete ground ancestry. Execution/playback, result
-interchange, persistence, compiled physics, downstream parity, protected
-evidence, integration, and release remain open.
-
-## 2026-08-10 PR #4339 stale-validation invalidation follow-up
-
-Rendered exact-head browser QA found that a validated one-overlay readback
-remained visible after adding a second overlay. The broad PyQt6 suite also
-found missing hover hints on the new controls. PyQt6 and React now clear both
-canonical readback and prior validation state whenever any identity, interval,
-material, or overlay-row draft value changes; PyQt6 exposes the explicit
-"Changes not validated" pending state, and every PyQt6 interactive has a
-specific engineering-context tooltip. RED-first regression tests cover the
-dynamic-row path on both surfaces. Final local evidence is 870 Python/PyQt
-tests and 672 React tests passing; Ruff, format, MyPy, TypeScript, zero-warning
-ESLint, the 198-module production build, manifest/docs/file-size governance,
-and diff checks are clean. The known polynomial-generator empty-legend warning
-is unrelated. This changes no wire schema, provenance digest, physics, or
-persistence boundary; fresh protected CI is still required before merge.
-
-## 2026-08-11 Regional wire receives current regional-surface parent
-
-Open draft PR `#4335` retains `feat/4271-regional-wire-contract` over
-`feat/4271-regional-surface-transitions`. Exact remote regional-wire child
-`9e01ccc3e891cc45907293751a192624195a77a5` is merged first and exact current
-regional-surface parent `5ac3c1f233de30dd42eeac57073cbdb149bde652`
-second by an ordinary no-ff merge. Child-owned wire production, schemas,
-tests, canonical fixture evidence, resolver binding, and limitations remain
-intact while inheriting current regional, skid/roll, impact/bounce,
-flight-transfer, and strict-ground ancestry.
-
-Focused local qualification passes all `132` Python ground/regional-wire tests,
-`9` focused React regional/ground-contract tests, and all `137` `tools-core`
-Rust tests. TypeScript, zero-warning ESLint, the 190-module Vite build, Rust
-format and warning-denied clippy, pinned Ruff 0.14.10 across six changed Python
-files, pinned MyPy 1.13 and Bandit medium/high across four production files,
-the 500-LOC changed-file gate, manifest validation plus eight contracts,
-documentation governance, exact wire-tree preservation, history/marker, and
-diff checks are green.
-
-This candidate remains unpublished and `not_released`. Changing geometry or
-velocity, internal transition-ledger export, regional UI, compiled or
-TypeScript regional physics, UpstreamDrift parity, protected exact-head CI,
-review, approval, dependency integration, and release remain open.
-
-## 2026-08-11 Regional surfaces receive current skid/roll parent
-
-Open draft PR `#4332` retains
-`feat/4271-regional-surface-transitions` over
-`feat/4271-ground-skid-roll`. Exact remote regional-surface child
-`04ccf08dd990de1cd056a3420e67772773a4be2e` is merged first and exact current
-skid/roll parent `3f861c1faaa7455faee92eaa7f813174667208e0` second by an ordinary no-ff
-merge. Child-owned regional transition production, tests, fixtures, contracts,
-evidence, and limitations remain intact while inheriting current skid/roll,
-impact/bounce, flight-transfer, and strict-ground ancestry.
-
-Focused local qualification passes all `121` Python ground/regional tests,
-`20` focused React ground-contract/transfer tests, and all `137` `tools-core`
-Rust tests. TypeScript, zero-warning ESLint, the 189-module Vite build, Rust
-format and warning-denied clippy, pinned Ruff 0.14.10 across 14 changed Python
-files, pinned MyPy 1.13 and Bandit medium/high across 12 production files, the
-500-LOC changed-file gate, manifest validation plus eight contracts,
-documentation governance, exact child-tree preservation, history/marker, and
-diff checks are green.
-
-This candidate remains unpublished and `not_released`. Changing normals,
-discontinuous geometry or surface velocity, regional UI, compiled regional
-physics, UpstreamDrift parity, protected exact-head CI, review, approval,
-dependency integration, and release remain open.
-
-## 2026-08-11 Skid/roll receives current impact/bounce parent
-
-Open draft PR `#4304` retains `feat/4271-ground-skid-roll` over
-`feat/4270-ground-impact-bounce`. Exact remote skid/roll child
-`0ea6740965068542e9d8c7449e06ec07d88969e0` is merged first and exact current
-impact/bounce parent `62beb3e1aa951645fd556a53f9cbae4bb46c47e5`
-second by an ordinary no-ff merge. Child-owned skid/roll production, contracts,
-tests, fixture, and limitations remain intact while inheriting current impact,
-flight-transfer, and strict-ground ancestry.
-
-Local verification passes `154` focused ground/flight tests, all eight manifest
-tests, and all `137` `tools-core` Rust tests. Pinned Ruff 0.14.10 covers the 18
-PR-delta Python files; pinned MyPy 1.13 and Bandit medium/high cover 12
-production files. Rust format/clippy, manifest, docs, exact child-tree,
-history/marker, and diff gates are green.
-
-This local model slice remains `not_released`. Regional/changing-normal
-surfaces, deformation, torsional damping, roll-to-skid transitions, UI,
-compiled physics, publication, protected exact-head CI, review, approval,
-downstream integration, and release remain open.
-
-## PR #4302 corrected-flight-transfer propagation
-
-Exact corrected #4288 parent
-`247215422a6d4b677552955b4923bc609a553259` is incorporated into
-`feat/4270-ground-impact-bounce` by the normal merge containing this handoff.
-PR #4302 keeps base `feat/4269-flight-ground-transfer`; neither branch was
-rebased, retargeted, rewritten, or force-pushed. The child retains passive
-impact, deterministic repeated hops, exact contact, capture-to-skid, bounded
-cancellation/failure, and airborne evidence while inheriting corrected flight
-transfer, deterministic workspace timestamps, and canonical `swing_sim`
-identity.
-
-This remains a partial `not_released` slice. Issue #4271 retains skid, roll,
-rest, total distance, and final ground results; terrain deformation, UI,
-TypeScript bounce physics, compiled bounce runtimes, and UpstreamDrift adapters
-remain excluded. Protected CI, independent review, and normal stack collapse
-remain open.
 
-Merged-tree validation is `987` Python tests, `106` React files / `661` tests,
-and the complete `tools-core` Rust suite at `137` tests (`111` unit, `20`
-transfer, `6` wire). The affected cross-version suite is also `146` tests on
-real CPython 3.10.20. The 189-module Vite production build, TypeScript,
-zero-warning ESLint, Ruff check/format across 90 files, pinned mypy 1.13 across
-all 17 ground and nine transfer production modules, Rust workspace format plus
-warning-denied `tools-core` clippy, campaign-manifest validator plus eight
-contracts, documentation governance, 11-file 500-LOC budget, marker scan, and
-diff checks are clean. Hosted checks and review apply to the new exact merge
-head only.
-
-
-## PR #4288 corrected-ground propagation
-
-Exact corrected #4285 parent
-`788aa547651a3685a363ea401824a5d81477bafb` is incorporated into
-`feat/4269-flight-ground-transfer` by the normal merge containing this handoff.
-PR #4288 keeps base `feat/4268-ground-contract`; neither branch was rebased,
-retargeted, rewritten, or force-pushed. The child retains its qualified
-cross-runtime terminal-state/contact transfer and inherits the corrected UTC
-parser plus canonical `swing_sim` import identity.
-
-This is ancestry propagation, not bounce, skid, roll, terrain response, total
-distance, or UI delivery. Protected CI, independent review, and normal stack
-collapse remain open.
-
-Merged-tree validation is `951` Python tests, `106` React files / `661` tests,
-and `26` focused Rust transfer/wire tests. The affected cross-version suite is
-also `110` tests on real CPython 3.10.20. The 189-module Vite production build,
-TypeScript, zero-warning ESLint, Ruff check/format across 82 files, pinned mypy
-1.13 across the six-file transfer and 13-file parent production namespaces,
-Rust workspace format plus warning-denied `tools-core` clippy, campaign-manifest
-validator plus eight contracts, documentation governance, 13-file 500-LOC
-budget, marker scan, and diff checks are clean. Hosted checks and review apply
-to the new exact merge head only.
+## 2026-08-11 PR #4342 append-only preservation repair
 
-
-> **Update this file with every PR and every push to main.**
-> Last updated: 2026-08-11
+Independent audit of local merge `e7fedfb18de1550eed3484ed2fc99d0baaecdca1`
+found that three older parent ancestry sections had been summarized instead of
+retained verbatim. This docs-only follow-up restores the full parent text below
+without changing production code, tests, schemas, manifest state, PR base, or
+release claims. The exact parent sections and SPEC rows are now governed by
+byte-for-byte comparison against
+`d21741e312b849a63f73cabf351a15d9de80fb94`.
 
 ## 2026-08-11 PR #4335 current-parent ancestry candidate
 
@@ -279,55 +144,6 @@ now records open PR #4332 at its still-published child head and open PR #4304
 at its exact newly published parent head; neither record claims protected or
 main-release evidence.
 
-## 2026-08-11 Ground-impact bounce receives current flight-transfer parent
-
-Open draft PR `#4302` retains `feat/4270-ground-impact-bounce` over
-`feat/4269-flight-ground-transfer`. Exact remote bounce child
-`cf6e72bad98e5f36f782254942e6895b8b71e670` is merged first and exact current
-transfer parent `c39297d37a17d5d8d3520ed62c5563cdcc609cab` second by an
-ordinary no-ff merge. The child-owned impact and repeated-bounce contracts,
-tests, fixture, and explicit limitations remain intact while inheriting current
-strict-ground and cross-runtime flight-transfer ancestry.
-
-Three inherited PR-delta Python files receive only their pinned Ruff 0.14.10
-layout. Analysis, persistence, and contract assertion behavior are unchanged.
-
-Local verification passes `1129` Rate/ground/flight Python tests, `25` focused
-React transfer tests, and `137` Rust tests. TypeScript, zero-warning ESLint, the
-189-module Vite build, Rust format/clippy, pinned Ruff 0.14.10, pinned MyPy
-1.13, Bandit medium/high, manifest parsing, history preservation, and diff
-checks are green.
-
-This is a local `not_released` candidate. Skid, roll, rest, total distance,
-final ground results, UI, TypeScript bounce physics, and compiled bounce
-runtimes remain outside this slice. Publication, protected exact-head CI,
-review, approval, dependency integration, and release remain open.
-
-## PR #4285 workspace timestamp propagation
-
-Exact corrected #4282 parent
-`5f77af4add23547a21cc3fabce98ae9ad4260427` is incorporated into
-`feat/4268-ground-contract` by the normal merge containing this handoff. PR
-#4285 keeps base `feat/4197-capability-observer`; neither branch was rebased,
-retargeted, rewritten, or force-pushed. The child retains its strict
-flight-to-ground contract and inherits the deterministic Python 3.10-3.12 UTC
-parser plus complete variation, scalar, wind, capability, and release ancestry.
-
-The preceding exact-head Python 3.12 CI lane loaded embedded
-`src.shared.python.swing_sim` tests and canonical `shared.python.swing_sim`
-imports as distinct package trees, producing ground/impact collection errors.
-The shared alias registry now coalesces that package root. A subprocess identity
-contract failed before the fix and now passes together with both affected
-public-API suites. The file-size job was cancelled in checkout before its
-budget step and is not a code failure.
-
-Merged-tree evidence is `915` Rate/ground/impact Python tests, all `12` shared
-alias architecture tests, `28` focused tests on CPython 3.12, `28` direct
-workspace compatibility tests on real CPython 3.10.20, and `104` React files /
-`642` tests. The production build, TypeScript, zero-warning ESLint, Ruff,
-format, pinned mypy 1.13, campaign manifest and eight contracts, docs
-governance, 500-LOC budget, marker scan, and diff checks pass.
-
 ## 2026-08-11 PR #4304 current-parent ancestry candidate
 
 The clean dedicated `feat/4271-ground-skid-roll` worktree first fast-forwarded
@@ -350,6 +166,180 @@ the official changed-file 500-LOC budget, and mandatory production limits of
 local merge is a PUBLISH candidate for independent review only; protected
 exact-head CI, approval, normal child propagation, and publication remain open.
 
+
+## 2026-08-11 PR #4342 current-parent reconciliation candidate
+
+The clean `feat/4274-regional-plan-io` worktree normally merges exact
+published PR #4342 child `8e1c7ccd99a7c4886c5fb9ccc7e4d94a6d7e3833`
+with exact newly published PR #4339 parent
+`d21741e312b849a63f73cabf351a15d9de80fb94` as its second parent. PR #4342
+continues to target `feat/4274-regional-surface-ui`; neither branch is rebased,
+retargeted, rewritten, force-pushed, or published by this reconciliation.
+The merge preserves strict regional request import/export together with the
+parent's frozen engineering-number specification and bounded helper.
+
+The child widget module now delegates its canonical-precision controls to that
+shared three-parameter helper while retaining the inclusive cross-runtime safe
+number bounds and eleven-decimal presentation required by its I/O contract.
+Merged-tree qualification is 53 focused Python/PyQt/shared-ground tests and 14
+focused React tests passing. Pinned Ruff 0.14.10 check/format, MyPy 1.13 over
+all ten affected production modules, Bandit medium/high screening, TypeScript,
+zero-warning ESLint, and the 199-module production build pass. Campaign
+manifest validation and all eight manifest tests, documentation and tool
+manifest governance, changed-Python, minimum-test, exact-diff assertion,
+500-LOC, parent/child diff, and whitespace gates pass. Mandatory structural
+maxima are 396 module lines, 40 function lines, and three parameters.
+The local merge is not protected or release evidence. Independent review,
+ordinary fast-forward publication, fresh protected CI/approval, dependency
+integration, and release remain required.
+
+## 2026-08-11 PR #4339 structural helper repair candidate
+
+This parent follow-up descends exact independently reviewed reconciliation
+candidate `eedad6d23b517eb4c99d4ba9000ff6555101099f`. Independent review found
+one publication blocker: the regional PyQt number-input helper accepted six
+parameters, above the campaign's mandatory four-parameter limit. RED-first
+tests bind the helper's signature and all observable widget settings. A frozen
+`NumberInputSpec` validates finite positive step and finite ordered bounds
+before widget construction; the three-parameter helper preserves field names,
+values, units, ranges, increments, tooltips, and presentation order.
+
+Eight focused Python/PyQt tests and 25 focused React tests pass; the complete
+React suite passes all 108 files / 672 tests. Pinned Ruff check/format, MyPy,
+Bandit medium/high screening, TypeScript, zero-warning ESLint,
+documentation/manifest governance, changed-production, test-assertion,
+minimum-test, 500-LOC, and diff gates pass. Mandatory maxima are 400 module
+lines, 42 function lines, and three parameters. No schema, digest, regional
+physics, UI behavior, PR base, protected evidence, or release boundary changes.
+
+## 2026-08-11 PR #4339 current-parent ancestry candidate
+
+The clean `feat/4274-regional-surface-ui` worktree normally merged exact
+published child `cbb9c0a6bdc6a50f59f7a661139b9d53e1892980` with exact published
+#4335 parent `9e01ccc3e891cc45907293751a192624195a77a5`, while preserving
+#4339's `feat/4271-regional-wire-contract` base. Production UI, contract,
+resolver, regional physics, and ground ancestry merged without conflict; only
+SPEC, manifest, and handoffs were reconciled. Independent review, ordinary
+publication, protected exact-head CI/approval, dependency integration, and
+release remained open.
+
+## 2026-08-11 PR #4342 delta-MyPy follow-up
+
+Protected CI on exact PR #4342 head
+`cffe349ac0a8054f1d168cb36684fd00bc5f8a49` identified one redundant `bool`
+cast in the regional atomic-write adapter under the Linux delta-MyPy gate. The
+cast is removed; the typed helper's direct Boolean return is preserved. This
+changes no wire bytes, validation, file semantics, UI behavior, or physics.
+Focused persistence tests and the CI-equivalent pinned MyPy command pass
+locally; fresh protected CI/review and dependency ordering remain required.
+
+## 2026-08-11 regional request I/O protected publication
+
+Branch `feat/4274-regional-plan-io` is published normally as draft PR
+[#4342](https://github.com/D-sorganization/Tools/pull/4342), targeting exact
+PR #4339 branch `feat/4274-regional-surface-ui` at parent head
+`cbb9c0a6bdc6a50f59f7a661139b9d53e1892980`. Its reviewed implementation
+head is `d748e7a5ef3da5e6ce7737ff6829e0f14665fe97`; publication commits change
+no runtime behavior. Protected CI, independent review, issue #4274, parent
+ordering, integration, and release remain open.
+
+## 2026-08-11 bounded regional request read follow-up
+
+Independent review of local safe-number commit
+`10c394f6b1fd2927e7f3b1f96cc097cae6bfd380` found that native request import
+checked file size with `stat()` and then performed a separate unbounded text
+read. A concurrently grown or replaced file could therefore allocate beyond
+the 1 MiB wire cap before the strict parser rejected it.
+
+Native import now opens one binary handle, reads at most the wire cap plus one
+sentinel byte, rejects overflow, strictly decodes UTF-8, and only then delegates
+to the unchanged canonical v1 parser. Tests simulate content growth after the
+metadata check and reject invalid UTF-8 explicitly. The complete regional and
+atomic persistence set is 28 tests passing. This changes no schema, digest,
+physics, browser behavior, or write semantics. Static and governance evidence
+is recorded in this commit: Ruff, Ruff format, Black, focused MyPy, manifest,
+module-size, documentation-governance, and diff checks pass. The child remains
+local and unprotected.
+
+## 2026-08-11 regional request I/O safe-number follow-up
+
+Independent review of local commit `e39edf4b50b1fb9811b0032bec4758c7a08c9b74`
+found two cross-runtime representation gaps. The PyQt6 precedence spin box
+silently narrowed otherwise valid v1 integers above 1,000,000, and Python
+accepted integer-valued wire numbers above JavaScript's exact safe range.
+
+The native editor now uses an exact decimal integer entry for precedence and
+preserves every v1 value from zero through 9,007,199,254,740,991 without a
+floating-point conversion. Shared Python canonical-number and ground-record
+validation now reject magnitudes outside that same range, matching React while
+retaining all finite fractional values inside it. PyQt6 floating-point editors
+publish the same bounds. Regression coverage pins exact maximum-precedence
+open/save round-trip and matched native/browser refusal of unsafe material
+numbers. This changes no schema version, provenance algorithm, physics, or
+browser filesystem limitation. Final local evidence is 132 shared ground tests,
+21 regional editor/I/O tests, five atomic-workspace tests, 23 focused React
+tests, the whole-shell tooltip sweep, and eight manifest tests passing. Ruff,
+Ruff format, Black, focused MyPy, TypeScript, zero-warning ESLint, the 199-module
+production build, module and changed-file size budgets, documentation
+governance, and diff checks pass. The child remains local and unprotected.
+
+## 2026-08-10 PR #4339 stale-validation invalidation follow-up
+
+Rendered exact-head browser QA found that a validated one-overlay readback
+remained visible after adding a second overlay. The broad PyQt6 suite also
+found missing hover hints on the new controls. PyQt6 and React now clear both
+canonical readback and prior validation state whenever any identity, interval,
+material, or overlay-row draft value changes; PyQt6 exposes the explicit
+"Changes not validated" pending state, and every PyQt6 interactive has a
+specific engineering-context tooltip. RED-first regression tests cover the
+dynamic-row path on both surfaces. Final local evidence is 870 Python/PyQt
+tests and 672 React tests passing; Ruff, format, MyPy, TypeScript, zero-warning
+ESLint, the 198-module production build, manifest/docs/file-size governance,
+and diff checks are clean. The known polynomial-generator empty-legend warning
+is unrelated. This changes no wire schema, provenance digest, physics, or
+persistence boundary; fresh protected CI is still required before merge.
+
+> **Update this file with every PR and every push to main.**
+> Last updated: 2026-08-11
+
+## 2026-08-10 issue #4274 canonical regional request I/O local child
+
+Local branch `feat/4274-regional-plan-io` started directly from published PR
+#4339 head `a9ace5052bcd54b78f79b62d5d9ac26debedb4b1`, then fast-forwarded
+normally to corrected exact parent `cbb9c0a6bdc6a50f59f7a661139b9d53e1892980`.
+The parent's stale-validation invalidation and tooltip follow-up are preserved.
+This child has not been pushed, opened as a PR, reviewed, merged, or released.
+
+The matched Ground Surfaces editors now import and export the canonical
+`ground-regional-material-plan-request/v1` document. PyQt6 uses native Open
+and Save As dialogs plus a shared flush/fsync/atomic-replace UTF-8 writer;
+cancel is a no-op, failed reads and writes preserve the last-known-good editor
+or file, and the recent path advances only after success. React validates a
+bounded browser File before committing state and downloads exact canonical
+bytes while revoking its object URL. Browser filesystem limitations are
+explicit, and workspace persistence remains separate.
+
+Import accepts only editor producer/provider v1 evidence, the fixed target
+frame origin/downrange axis, and one to eight rows. It never relabels external
+evidence. An unchanged import round-trips the exact request and provenance;
+after any edit, validation binds a fresh draft digest. No measured calibration
+is synthesized and neither path executes physics.
+
+RED-first coverage now includes exact-byte round-trip, native cancel and
+atomic rollback, corruption, duplicate keys, oversize input, unsupported
+producer/axis qualification, stale-provenance rebinding, transactional PyQt6
+population, browser rollback, URL cleanup, and pre-allocation size rejection.
+Focused evidence is 25 Python/PyQt tests and 12 React tests passing. With
+adjacent wire and manifest coverage, 44 Python and 16 React tests pass.
+Focused MyPy, Ruff, TypeScript, zero-warning ESLint, the 199-module production
+build, release-manifest validation, and module budgets are clean.
+
+Issue #4274 remains open. Measured calibration, workspace model-input
+persistence, execution/playback, result interchange, visualization, changing
+geometry, compiled regional physics, downstream parity, protected CI/review,
+integration of this child into the protected stack, and release remain explicit
+gaps.
+
 ## 2026-08-10 issue #4274 matched regional surface editor local child
 
 Local branch `feat/4274-regional-surface-ui` started from published PR #4335
@@ -521,420 +511,6 @@ fixture bytes. There is no physics, numerical, schema, or API change. SPEC
 a finding-free local scan of the affected file, documentation governance, the
 `370`-line source-size check, and diff gates pass; fresh protected CI/review
 remain required after a normal guarded push.
-## 2026-08-11 Flight-ground transfer receives current strict-ground parent
-
-Open draft PR `#4288` retains `feat/4269-flight-ground-transfer` over
-`feat/4268-ground-contract`. Exact remote child
-`df716ac740beb9d47dced66f867dc3dde1fa5367` is merged first and exact current
-parent `584e4e53a7990053a346938c4e45538786887ec7` second by an ordinary
-no-ff merge. All transfer production/tests/fixtures and Python, TypeScript,
-Rust, PyO3, and WASM contracts remain intact. Only append-only handoff/SPEC
-records conflict, and both dated histories are retained. Focused verification
-passes 85 Python, 25 React, and 137 Rust tests; pinned Ruff 0.14.10, MyPy 1.13,
-Bandit, TypeScript, exact-delta ESLint, the production web build, Rust format,
-docs, strict manifest, diff, exact transfer-tree, and append-only preservation
-gates pass. Publication, protected exact-head CI, review, approval, and release
-  remain open.
-
-## 2026-08-11 PR #4288 append-only history preservation repair
-
-This documentation-only follow-up to exact local candidate
-`1eb4188ff1a87a3d8e9a55aeca68194ca04e0bb4` restores every omitted parent
-handoff section and specification row and repairs the interleaved paragraph
-below. Production code, tests, schemas, the campaign manifest, child-first
-merge topology, and #4288's `feat/4268-ground-contract` base are unchanged.
-Publication, protected exact-head CI, approval, dependency integration, and
-release remain open.
-
-## 2026-08-11 PR #4288 current-ground-parent ancestry candidate
-
-The clean `feat/4269-flight-ground-transfer` reconciliation starts from exact
-published child `afb1bdcfc6701caaf1f7bc3497a6a37dd9698c14` and normally merges
-exact newly published #4285 parent
-`e09ab96280cab363b36bd0e7db4a3cd064dc2527` second without changing #4288's
-base. The child's bounded `flightIntegrator.ts` facade is preserved instead of
-restoring the parent's superseded inline loop, while kinetics resolves to the
-parent's split validated implementation. Seven child-owned Python files receive
-pinned Ruff 0.14.10 normalization with no AST change; the preserved TypeScript
-facade uses standard line endings. Bounce, skid, roll, terrain response, total
-distance, UI integration, protected evidence, and release remain open.
-
-> Update this append-only handoff with every implementation commit and every
-> push to `main`. Last updated: 2026-08-11.
-
-## 2026-08-11 Ground contract receives current capability carrier
-
-Open draft PR `#4285` retains branch `feat/4268-ground-contract` and base
-`feat/4197-capability-observer`. Exact remote ground-contract child
-`e09ab96280cab363b36bd0e7db4a3cd064dc2527` is merged first and exact current
-capability-observer parent `d0b827667e61f0583e1dad0b3cbbca6819624d3c`
-second by an ordinary no-ff merge. All strict flight-to-ground production,
-tests, schemas, fixtures, dependencies, and fail-closed limitations are
-preserved. Only append-only handoff/SPEC files conflict; their dated records
-from both parents remain. Publication, protected exact-head CI, review,
-approval, release, and downstream #4288 propagation remain open. Focused
-verification passes 214 Python/PyQt tests, pinned Ruff 0.14.10 across 27
-PR-delta Python files, pinned MyPy 1.13 across 20 production files, and Bandit's
-medium/high threshold. Docs, strict manifest, diff, exact ground-tree, and
-both-parent append-only preservation gates pass.
-
-## 2026-08-11 Merged capability-observer carrier receives current wind workflow
-
-PR `#4283` is merged, while its former head branch
-`feat/4197-capability-observer` remains the base of open draft PR `#4285`.
-Exact remote/local carrier child
-`b2baf6ef615b0d756a86ac4ca7eef2fc583210ee` is merged first and exact published
-#4282 parent `3faf9e9493c9ba37168d9f65ab10cafeedc2a72f` second by a normal
-no-ff merge. Since the carrier is already the first parent of #4282, the source
-merge is conflict-free: the complete capability-observer implementation and
-history remain present while the exact current wind workflow and manifest tree
-are inherited. Only handoff/SPEC carrier records differ from #4282. All 86
-focused Python/PyQt capability, wind, and manifest tests and all 86 tests in 12
-focused React capability/wind suites pass. React type-check, zero-warning lint,
-and the production build pass. Docs, manifest, and exact-parent diff gates pass;
-changed-Python static gates are not applicable to the documentation-only exact
-delta. This local audit does not authorize publication; review,
-protected CI, and
-later dependency propagation into #4285 remain open.
-
-## 2026-08-11 Wind workflow receives newly published scalar-adapter parent
-
-PR `#4282` keeps branch `feat/4199-wind-workflow` and base
-`feat/4199-wind-scalar-adapter`. Exact published child
-`b2baf6ef615b0d756a86ac4ca7eef2fc583210ee` is merged first and exact current
-parent `9321c1d2e091b8c7e5a4a83aa9ad726290e7fb5a` second. Implementation and test
-paths merge automatically: responsive PyQt6/React wind execution,
-progress/cancellation, controls, scatter, persistence/export, consolidated
-capability workflow, and strict campaign release authority remain intact while
-the parent contributes current scalar-adapter, variation, workspace,
-launch-monitor/D-plane, split-kinetics, and Qt repairs. Only append-only
-handoffs and SPEC conflict textually; both histories are preserved. All 94
-focused Python/PyQt wind, responsive-layout, navigation, playback, and manifest
-tests pass, as do all 44 tests in eight focused React wind/responsive suites.
-Docs governance and strict campaign-manifest generation/validation pass.
-Pinned Ruff checks/formats all 57 changed Python files; pinned MyPy accepts 38
-production files; pinned Bandit finds no medium/high issues in 37 scanned source
-files; the 500-LOC, changed-Python policy, test-contract/assertion, diff, React
-type-check/lint, and production-build gates pass. Review, ordinary publication,
-protected exact-head CI, unresolved threads, dependency order, and release
-remain open.
-
-## 2026-08-11 Wind workflow receives current scalar-adapter parent
-
-PR `#4282` remains on `feat/4199-wind-workflow`, based on
-`feat/4199-wind-scalar-adapter`. Exact remote child
-`29e15d6ff631f7f30afcc745be783f2e716d7dcf` normally merges exact published
-parent `a7dce5f89b483303938f518b74b4028a1c68ba81` second, preserving child-first
-history and the existing base. The responsive PyQt6/React workflow and
-consolidated capability stack inherit the parent's split kinetics facade,
-Ground/Tee parity, and 252-line public/213-line private wind-adapter repair.
-
-The only source conflict is an obsolete formatting-only edit to the pre-split
-kinetics monolith and resolves wholly to the validated parent facade. Both
-append-only documentation histories remain present. Seven surviving
-child-owned paths receive only repository-pinned Ruff formatting. The exact
-skipped-import MyPy profile also replaces one now-unused cache-return ignore
-with an explicit `KineticsSeries | None` cast; the cached object and runtime
-behavior are unchanged. Protected exact-head CI, independent review,
-unresolved-thread checks, publication, and dependency order remain release
-gates.
-
-Local evidence is 74 focused Python/PyQt wind, kinetics, and Ground/Tee tests;
-29 focused React tests; and 8 campaign-manifest tests. TypeScript, focused
-zero-warning ESLint, Ruff check/format, pinned MyPy 1.13, Bandit's medium/high
-threshold, strict wind module/signature/function budgets, manifest,
-documentation, policy, SPEC, and diff gates pass.
-
-## 2026-08-11 Ground contract receives current workflow/base ancestry
-
-Draft PR `#4285` keeps branch `feat/4268-ground-contract` and base
-`feat/4197-capability-observer`. Exact published child
-`7a38b9838d743e051a6900620c4d6e754582aa89` is merged first with exact current
-base/workflow head `b2baf6ef615b0d756a86ac4ca7eef2fc583210ee` second. The base contains
-merged #4283 head `9bbb98e16e435a0d4c74153b909f2ebfefbbce7a`; no rebase, retarget,
-force-push, parent rewrite, or publication is used.
-
-The child's strict UI-neutral `flight-to-ground/v1` schemas, canonical fixture,
-migrations, legacy-result adapter, explicit dependency, enum compatibility,
-and fail-closed validation remain authoritative. The parent contributes the
-current responsive wind/capability ancestry, Ground/Tee parity, and split
-kinetics modules. Its facade wholly resolves the sole formatting-only source
-conflict; both append-only documentation histories remain present.
-
-The ground branch's contract-checked scalar extraction repair remains intact.
-Because it makes the plotting catalog part of the effective delta, the former
-459-line registry now delegates immutable ordered rows to focused private
-scalar/series modules. The public catalog facade is 61 lines; every changed
-production module is at most 400 lines, every function at most 50 lines, and
-every signature at most four non-receiver parameters. Stable keys, display
-order, units, categories, extractor behavior, and public objects do not change.
-
-Complete local regression evidence is 1,539 Rate/shared-swing Python tests
-passed with one explicit optional Rust-wheel skip; 106 React files / 648 tests,
-TypeScript, zero-warning ESLint, and the 188-module Vite build passed; and all
-12 `swing-core` Rust tests passed. The focused ground, Ground/Tee parity,
-kinetics, wind adapter, compatibility, and manifest set passes 146 tests; 57
-additional plotting/catalog/variation/kinetics regressions pin the structural
-split. Pinned Ruff 0.14.10, MyPy 1.13, Bandit 1.7.7, manifest, documentation,
-SPEC, policy, diff, and structural gates are required again on the committed
-candidate before publication.
-
-This is an unpublished local candidate. It does not claim bounce, skid, roll,
-terrain, total distance, ground UI, Rust/WASM ground delivery, protected CI,
-approval, or release. Independent review and fresh protected checks are
-required before normal publication and later #4288 propagation.
-
-## 2026-08-10 Repaired scalar-adapter propagation into wind workflow
-
-Draft PR `#4282` remains on `feat/4199-wind-workflow`, based on
-`feat/4199-wind-scalar-adapter`. It now normally incorporates repaired parent
-head `d6fb04e07c2a625412e9208b07103acdc42c621b` after that head's quality gate
-passed. The merge had no wind-workflow production/test conflict and used no
-rebase, retarget, force-push, or history rewrite. Twenty-five focused tests and
-the documentation, size, and whitespace gates pass locally; protected CI,
-review, and later-stack propagation remain outstanding.
-
-## 2026-08-10 Exact-head manifest typing repair
-
-PR #4282 exact head `aa6eeffb0395f7ed7954f2315b1c625cada552d8`
-failed hosted quality-gate run `31395741841` only at pinned-mypy
-`scripts/rate_campaign_manifest.py:335:5 [no-any-return]`. The CI profile skips
-import analysis, so the Pydantic validator result becomes `Any`; an explicit
-local `CampaignManifest` annotation now retains the validated return contract
-without changing parsing, validation, or error behavior. Ruff and format had
-already passed on the failed head. Publish this repair normally, then propagate
-its new exact head into #4285; do not retry the obsolete run.
-
-## 2026-08-10 Wind workflow receives exact scalar-adapter parent
-
-PR `#4282` keeps branch `feat/4199-wind-workflow`, base
-`feat/4199-wind-scalar-adapter`, and original child first parent
-`5f77af4add23547a21cc3fabce98ae9ad4260427`. Exact parent
-`4a793c4c3f19aad43a3c215800b266be487ace49` merges normally second. The
-responsive PyQt6/React wind strategy workflow, consolidated capability stack,
-and campaign release authority remain additive with the parent's complete
-variation, workspace, compatibility, and strict scalar-adapter history.
-
-The parent's extracted navigation-state tuple is the single authority and now
-includes the child-owned `capability_optimization` ID. This preserves safe
-legacy visibility migration and prevents a persisted pre-capability workspace
-from hiding the new tab. Runtime-only UI and capability invariants now raise
-explicit errors under optimized Python, and the safe fixed-argument builder
-subprocess is annotated for Bandit. The skipped-import manifest return repair
-is recorded above. SPEC 1.14.15 is the current exact-head repair entry.
-Verification passes 73 focused Python tests, the complete 1,657-test
-Rate/shared-swing/golf-club matrix with two explicit optional build123d skips,
-all 643 React tests across 105 files plus type-check/lint/build, 12 Rust tests,
-31 real-CPython-3.10 compatibility tests, and 43 post-review invariant/GUI/
-manifest tests. Ruff/format, pinned MyPy 1.13 across 37 changed production
-files, Bandit, exact-parent size, manifest, docs, minimum-test, assertions,
-CI-policy secrets, compilation, and diff gates pass. Independent re-review
-found no actionable findings after 94 Python workflow/navigation tests, 20
-manifest/GUI tests, 61 React workflow tests, and independent static checks.
-Protected CI/review and dependency-ordered propagation remain release gates.
-
-## Current continuation
-
-The active local continuation is integrated directly on the existing PR #4282
-carrier `feat/4199-wind-workflow`, starting from original published head
-`5f77af4add23547a21cc3fabce98ae9ad4260427` and normally incorporating exact
-published #4281 parent `4a793c4c3f19aad43a3c215800b266be487ace49`.
-The base remains `feat/4199-wind-scalar-adapter`; no branch was rebased,
-retargeted, force-pushed, or published by this continuation. It adds strict
-cross-runtime capability parsing, reliable signed decimal entry, complete
-ranked diagnostics and result exports, quantitative React scatter annotations,
-package-safe static-web release entrypoints, and the strict
-`rate-of-closure-campaign/v1` release authority. It also carries the corrected
-parent's Python 3.10 compatibility, variation-export, scalar-ensemble, and
-wind-adapter history. Normal publication does not establish a protected
-release; CI, review, and downstream evidence remain required.
-
-The direct web launcher dynamically loads the root bootstrap module without a
-launcher-local `sys.path` mutation; its subprocess delegation test and the
-changed-production-Python policy guard cover this release entrypoint.
-
-Use these authorities together:
-
-- `docs/release/rate_of_closure_campaign.v1.json` — machine state;
-- `scripts/rate_campaign_manifest.py` — validation and generated JSON Schema;
-- `docs/release/RATE_OF_CLOSURE_CAMPAIGN_MANIFEST.md` — update procedure;
-- `docs/development/RATE_OF_CLOSURE_CAMPAIGN_HANDOFF.md` — detailed history;
-- `docs/specs/*.md` — scientific and interoperability contracts.
-
-The manifest is intentionally normalized: programs reference shared carrier
-and test-evidence records. It fails on missing primary issues, undeclared
-references, duplicate IDs, malformed SHAs, absent repository evidence,
-placeholders, and contradictory release claims.
-
-## Carrier state
-
-The capability stack was collapsed top-down on 2026-08-09:
-
-1. #4294 Shot Optimizer UI merged into `feat/4197-capability-flight-evaluator`;
-2. #4289 evaluator merged into `feat/4197-capability-observer`;
-3. #4283 observer merged into `feat/4199-wind-workflow`.
-
-These were feature-parent merges, not protected releases. Current carrier
-#4282 descends from #4281 → #4280 → #4279 → #4203 → #4202 → wedge and
-variation parents. Exact published #4281 parent
-`4a793c4c3f19aad43a3c215800b266be487ace49` is incorporated locally. Preserve
-that dependency order and use normal merges only.
-
-Outer platform PR #4119 targets `main` and still needs reconciliation. PR #4133
-impact-interval dynamics merged into a historical feature parent after that
-parent had already propagated, so its source and tests must be explicitly
-reconciled before #4130 can close.
-
-## Implemented on the feature stack
-
-- PyQt6 and React variation workspaces: selectable input/contact/impact/shot
-  scatter, matrix, landing dispersion, every-trial 3D swing arcs, quiet zones,
-  linked selection, bounded raw rows, and lossless exports.
-- Matched toolstrip/workspace controls, module visibility, independent plots,
-  legend placement, zoom/autofit, replay, loop, and granular playback rate.
-- Launch-monitor convention, D-plane, spatial target, inverse flight, wind,
-  interactive flight playback, and capability-optimization contracts/workflows.
-- Wedge kinematics, clearance, turf proxy, impact visualization, and
-  forgiveness-analysis slices.
-
-## Current limitations
-
-- No complete campaign program has a qualified `main` release SHA.
-- Capability v1 is still-air carry to first ground crossing; it excludes wind,
-  bounce, roll, and total distance in that evaluator.
-- Ground contracts and flight transfer are open carriers; qualified bounce,
-  skid, roll, profiles, total-distance optimization, UI, and Rust/WASM parity
-  remain.
-- UpstreamDrift React has no native Rate route, and the Tools source pin and
-  resolver behavior are not a qualified immutable parity boundary.
-- Camera tracking/snap views (#4284), complete persistence adapters, high-DPI,
-  keyboard, reduced-motion, and visual-regression matrices remain open.
-- Local/rendered evidence is not protected CI or installed-package evidence.
-
-## Most recent evidence
-
-- Corrected-parent propagation: 62 focused wind/scalar/variation and
-  compatibility tests pass on Python 3.11 and real CPython 3.10.20; 8 React
-  files / 35 tests, TypeScript, and focused zero-warning ESLint pass. The real
-  3.10 run found and now guards #4282's child-owned capability-observation
-  `StrEnum` boundary. Ruff check/format passes 15 focused files, pinned mypy
-  1.13 passes 10 production modules, and nine campaign manifest/parity
-  contracts pass.
-- Composed local continuation: 828 Rate Python/PyQt tests and 104 React files /
-  642 tests passed; TypeScript, zero-warning ESLint, and the 188-module Vite
-  production build passed. The manifest, schema JSON, Ruff, targeted mypy, and
-  nine manifest/parity contracts pass on implementation head `2c1a77baa`.
-  Hosted CI has not run on this composition.
-- Hosted quality-gate run `31340032608` subsequently found mypy 1.13
-  `no-any-return` findings at the Pydantic loader and Qt elapsed-timer adapter
-  under `--follow-imports=skip`; both boundaries now narrow their return values
-  explicitly. The exact Python 3.12/mypy 1.13 delta is clean across 54 files;
-  Ruff, 62 focused regression tests, and eight campaign-manifest tests also
-  pass locally.
-- Consolidated capability head `c1827bbdc`: 1,426 Python Rate/shared-swing
-  tests and 624 React tests passed locally.
-- Variation continuation `d71b0ea01`: 890 Python/PyQt/shared-swing tests and
-  545 React tests passed locally, with one optional Rust skip.
-- Toolstrip head `c36ca36e9`: the same 890/545 local suite totals passed.
-- Hosted #4282 head `3186a265b1`: `swing_core` built, but cached plugin loading
-  failed before parity collection; this is failed evidence, not a model pass.
-- CI-isolation head `18fe89201`: focused workflow regression passes locally;
-  the hosted non-skipped parity run is still required.
-
-## Validation
-
-```powershell
-python scripts/rate_campaign_manifest.py
-python -m pytest tests/rate_of_closure/test_campaign_release_manifest.py -q
-$env:QT_QPA_PLATFORM='offscreen'
-$env:PYTHONPATH=(Resolve-Path 'src').Path
-python -m pytest tests/rate_of_closure src/shared/python/swing_sim -q
-cd src/rate_of_closure/web
-npm test -- --run
-npm run type-check
-npm run lint
-npm run build
-```
-
-## 2026-08-11 Wind workflow receives current scalar-adapter parent
-
-PR `#4282` remains on `feat/4199-wind-workflow`, based on
-`feat/4199-wind-scalar-adapter`. Exact remote child
-`29e15d6ff631f7f30afcc745be783f2e716d7dcf` normally merges exact published
-parent `a7dce5f89b483303938f518b74b4028a1c68ba81` second, preserving child-first
-history and the existing base. The responsive PyQt6/React workflow and
-consolidated capability stack inherit the parent's split kinetics facade,
-Ground/Tee parity, and 252-line public/213-line private wind-adapter repair.
-
-The only source conflict is an obsolete formatting-only edit to the pre-split
-kinetics monolith and resolves wholly to the validated parent facade. Both
-append-only documentation histories remain present. Seven surviving
-child-owned paths receive only repository-pinned Ruff formatting. The exact
-skipped-import MyPy profile also replaces one now-unused cache-return ignore
-with an explicit `KineticsSeries | None` cast; the cached object and runtime
-behavior are unchanged. Protected exact-head CI, independent review,
-unresolved-thread checks, publication, and dependency order remain release
-gates.
-
-Local evidence is 74 focused Python/PyQt wind, kinetics, and Ground/Tee tests;
-29 focused React tests; and 8 campaign-manifest tests. TypeScript, focused
-zero-warning ESLint, Ruff check/format, pinned MyPy 1.13, Bandit's medium/high
-threshold, strict wind module/signature/function budgets, manifest,
-documentation, policy, SPEC, and diff gates pass.
-
-## 2026-08-10 Final repaired workflow ancestry in strict ground contract
-
-Draft PR `#4285` remains on `feat/4268-ground-contract`, based on
-`feat/4197-capability-observer`. Its original head
-`e5bcbd1096d3be1f621a805c9d9f3fd321e375a5` normally incorporates exact
-quality-green #4282 head `1e82f15026786ea0b08f78f4c001590ddce9ff39`
-second. Ground-contract source and tests had no conflict. Protected CI and
-review must pass before the new exact head propagates normally into #4288.
-
-## 2026-08-10 Final repaired strict-contract ancestry in flight transfer
-
-Draft PR `#4288` remains on `feat/4269-flight-ground-transfer`, based on
-`feat/4268-ground-contract`. Its original head
-`108a841b1378c992defd3c7b7ee263d41a6c8b24` normally incorporates exact
-quality-green #4285 head `a93edb4bfd6a8dc9334122cd2ae660983d5bf424`
-second. Transfer source and tests had no conflict. Fresh protected CI and review
-must pass before any release claim.
-
-## PR #4288 exact repaired-ground propagation
-
-Original child `247215422a6d4b677552955b4923bc609a553259` normally incorporates
-exact repaired #4285 parent `e5bcbd1096d3be1f621a805c9d9f3fd321e375a5`
-second in the merge on `feat/4269-flight-ground-transfer`. PR #4288
-keeps base `feat/4268-ground-contract`; neither branch was rebased, retargeted,
-rewritten, or force-pushed. The child retains its qualified cross-runtime
-terminal-state/contact transfer and inherits the corrected UTC parser,
-canonical `swing_sim` import identity, hosted-mypy manifest repair, and the
-complete variation, wind, capability, and campaign ancestry.
-
-This is ancestry propagation, not bounce, skid, roll, terrain response, total
-distance, or UI delivery. Protected CI, independent review, and normal stack
-collapse remain open.
-
-Exact composed-tree verification is 1,080 Python tests passed with six explicit
-optional installed-`tools_core` wheel skips, 107 React files / 662 tests,
-26 direct Rust transfer/wire tests, TypeScript, zero-warning ESLint, and the
-189-module production build. Ruff check/format passes all 28 changed Python
-files; pinned mypy 1.13 passes all 21 changed production files. Campaign
-manifest, documentation governance, and diff checks pass. The missing local
-wheel remains an explicit installed-package release boundary and is not
-misreported as accelerated parity evidence.
-
-## 2026-08-10 Exact repaired #4282 propagation into PR #4285
-
-Original child `788aa547651a3685a363ea401824a5d81477bafb`
-normally incorporates exact repaired #4282 parent
-`686016196a2f895058b8a566dff103a0fd32cd10` second on
-`feat/4268-ground-contract`. PR #4285 keeps base
-`feat/4197-capability-observer`; neither branch was rebased, retargeted,
-rewritten, or force-pushed. The parent contains merged capability PR #4283,
-its exact observer head, and the exact-hosted-mypy manifest repair. The child
-retains its strict flight-to-ground contract while acquiring the latest
-deterministic variation, workspace, wind, capability, and release ancestry.
 
 ## 2026-08-09 PR #4304 corrected implementation evidence
 
@@ -1027,45 +603,6 @@ regions, changing normals, deformation/grass response, torsional spin damping,
 roll-to-skid, PyQt6/React UI, compiled runtimes, UpstreamDrift consumers,
 protected review, parent integration, and main release remain open.
 
-## 2026-08-10 Exact repaired #4282 propagation into PR #4285 (continued)
-
-The inherited ground descendant passes 1,703 Python tests with two optional `build123d`
-skips, 643 React tests across 105 files plus type-check/lint/build, 12 Rust
-tests, and 77 real-CPython-3.10 ground and compatibility tests. Ruff/format pass
-78 changed Python files; pinned mypy and Bandit pass 52 changed production
-files. Manifest, docs, minimum-test, assertions, 500-LOC, changed-file secrets,
-Python 3.10 compilation, and diff checks are clean.
-
-The repaired parent explicitly types the Pydantic manifest return. The composed
-tree also normalizes contract-checked scalar plot values with `float`, replaces
-the command-state `assert` with a runtime invariant error, and documents the
-alias finder's intentional optional-probe `B112` security boundary. Protected
-current-head CI and required review remain release gates. The exact repaired
-ground head is now incorporated into #4288; normal propagation of the reviewed
-#4288 head into #4298 is the next ancestry gate.
-
-## 2026-08-09 Ground parent reconciled with corrected wind carrier
-
-Draft PR #4285 keeps base `feat/4197-capability-observer` and normally
-incorporates exact corrected wind carrier
-`bb101cedd555d07d493aae998b46050c68660cdd`. No branch was rebased,
-retargeted, force-pushed, or published. The strict ground contract remains
-UI-neutral and fail closed; the merge adds no bounce, skid, roll, terrain
-profile, total distance, or presentation claim.
-
-This branch is now the carrier-reconciled ground parent. PR #4288 must merge
-the resulting exact #4285 head normally before its transfer logic is retested
-or published. That descendant merge will carry both the ground contract and
-the explicit observer-to-wind ancestry reconciliation without changing either
-PR base.
-
-Focused evidence is 89 ground/compatibility/wind tests on Python 3.11 and 89
-on real CPython 3.10.20. Ruff check/format passes 34 focused files; pinned mypy
-1.13 passes 23 production modules; the inherited campaign manifest validates
-and its nine manifest/parity tests pass. Ground production modules remain
-within the 400-line/50-line structural budgets and contain no placeholder
-markers.
-
 ## 2026-08-09 PR #4302 pinned-MyPy current-head correction
 
 Hosted quality-gate run `31350134551` exposed four deterministic MyPy 1.13
@@ -1879,1190 +1416,3 @@ Publish this as a normal follow-up commit on
 `feat/4270-ground-impact-bounce`, with the three handoffs in the same commit.
 Keep #4270 and #4267 open; do not bypass protected CI or propagate into child
 branches until the parent push is verified.
-## Inherited current-parent handoff history
-
- Python/TypeScript SHA-256 parity digests as high-entropy strings. These are
- expected deterministic test outputs, not secrets. Use only exact inline
- `pragma: allowlist secret` annotations on the four digest constants; preserve
- the repository baseline and scanner scope. Commit the annotations and this
- handoff together on the parent capability-observer branch, then re-run the
- protected stack without force-pushing or bypassing the gate. Parent commit
- `49612946138b1021f80c9f8d2a4d06f1610825db` is now propagated into this child
- by a normal merge commit.
-## Inherited #4281 handoff history
-
-## 2026-08-11 Wind scalar adapter receives current variation parent
-
-PR `#4281` keeps branch `feat/4199-wind-scalar-adapter` and base
-`feat/4144-variation-export-continuation`. Exact published child
-`a7dce5f89b483303938f518b74b4028a1c68ba81` is merged first and exact current
-parent `e6c7460a01082631565fb9ed48aa32538bd7772c` second. Implementation and test
-paths merge automatically: Python/TypeScript scalar-ensemble provenance,
-availability, deterministic-scenario, wind-strategy adapter, and strict
-module-budget behavior remain intact while the parent contributes current
-variation visualization/export, workspace, launch-monitor/D-plane ancestry,
-split kinetics, and behavior-preserving Qt return narrowing. Only append-only
-handoffs and SPEC conflict textually; both histories are preserved. All 77
-focused wind/variation Python tests and all 20 tests in the five focused React
-wind/variation suites pass before the merge commit. Pinned Ruff checks/formats
-all five changed Python files; pinned MyPy and Bandit accept all three production
-files; both 500- and strict 400-LOC, docs, changed-Python policy,
-test-contract/assertion, diff, React type-check/lint, and production-build gates
-pass. The later campaign-manifest artifact/checker/test exist on neither side of
-this pre-manifest stack and are not applicable. Review, ordinary publication,
-protected exact-head CI, unresolved threads, dependency order, and release
-remain open.
-
-## 2026-08-11 Wind scalar-adapter module-budget repair
-
-The 403-line `variation/wind_strategy_plot_adapter.py` exceeded the mandatory
-400-line ceiling. Request/analysis validation now lives in the private
-`variation/_wind_strategy_plot_validation.py` collaborator, leaving a 252-line
-public adapter and a 213-line validator. A frozen, slotted scenario-expectation
-object keeps the private scenario validator at two cohesive parameters; every
-production function is at or below four parameters and 50 lines. The public
-imports, dataset schema,
-variable/stage/category/cohort order, composite row identities, row values,
-typed availability, attributes, validation messages, and no-flight-rerun
-boundary are unchanged.
-
-The existing 14-test Python adapter suite and 16 focused React wind tests cover
-the behavior-preserving extraction; Ruff/format, pinned MyPy 1.13, Bandit, and
-the explicit strict line budget pass locally. Protected exact-head CI,
-independent review, unresolved-thread checks, ordinary publication, and stack
-order remain release gates.
-
-## 2026-08-11 Wind scalar adapter receives reviewed variation parent
-
-PR `#4281` stays on `feat/4199-wind-scalar-adapter`, based on
-`feat/4144-variation-export-continuation`. Exact remote child
-`247046d55afcad3e6cd4f8029f854856c427f59c` normally merges exact reviewed
-parent `3337945699966b63cb5cd8e52d7c3b194315e911` in child-first order,
-preserving both histories and every commit. Python/TypeScript scalar-ensemble
-provenance, typed availability, deterministic-scenario rules, and wind-strategy
-plot-adapter behavior remain unchanged. The parent supplies selected-scatter
-export, linked variation views, the kinetics split, Ground/Tee parity,
-workspace/toolstrip behavior, and protected Ruff normalization. The only
-feature-code conflict referenced obsolete monolithic kinetics and is resolved
-to the validated parent `pendulum.sample(...)` facade. This local merge still
-requires independent review, fresh exact-head protected CI, unresolved-thread
-checks, dependency order, and ordinary publication.
-## 2026-08-11 Variation export receives current workspace parent
-
-PR `#4280` keeps branch `feat/4144-variation-export-continuation` and base
-`feat/4218-toolstrip-workspace`. Exact published child
-`3337945699966b63cb5cd8e52d7c3b194315e911` is merged first and exact current
-parent `efbca84095b617b4018732f7802c2da3f0525387` second. Implementation and test
-paths merge automatically: selected-scatter CSV parity, typed unavailable
-outcomes, bounded accessible tables, linked selection, and all-trial arc
-analysis remain intact while the parent contributes current workspace,
-launch-monitor/D-plane ancestry, split kinetics, and behavior-preserving Qt
-return narrowing. Only append-only handoffs and SPEC conflict textually; both
-histories are preserved. All 56 focused Python variation tests and all 10 tests
-in the three focused React variation suites pass before the merge commit.
-Pinned Ruff checks/formats all five changed Python files; pinned MyPy and Bandit
-accept all four production files; docs, 500-LOC, changed-Python policy,
-test-contract/assertion, diff, React type-check/lint, and production-build gates
-pass. The later campaign-manifest artifact/checker/test exist on neither side of
-this pre-manifest stack and are not applicable. Review, ordinary publication,
-protected exact-head CI, unresolved threads, dependency order, and release
-remain open.
-## 2026-08-11 Variation export receives reviewed workspace parent
-
-PR `#4280` stays on `feat/4144-variation-export-continuation`, based on
-`feat/4218-toolstrip-workspace`. Exact remote child
-`668ba96746f79f7a12e8092161bd610054197f58` normally merges exact reviewed
-parent `ccd0e026c580c93038fdf5c59d5d452a85ba27a0` in child-first order,
-preserving both histories and every commit. Selected-scatter CSV parity, typed
-unavailable outcomes, bounded accessible tables, linked selection, and
-all-trial arc analysis remain unchanged. The parent supplies its validated
-kinetics split, Ground/Tee parity contracts, protected Ruff normalization, and
-workspace/toolstrip implementation. The only feature-code conflict referenced
-obsolete monolithic kinetics source and is resolved to the current
-`pendulum.sample(...)` façade. Seven duplicate child automation edits do not
-match protected Ruff 0.14.10 output and are normalized to the exact reviewed
-parent blobs while their commit remains in history. This local reconciliation
-still requires independent review, exact-head protected CI, unresolved-thread
-checks, dependency order, and ordinary publication.
-## 2026-08-11 Workspace child receives current #4203 parent
-
-PR `#4279` keeps branch `feat/4218-toolstrip-workspace` and base
-`feat/4181-launch-monitor-registry`. Exact published child
-`ccd0e026c580c93038fdf5c59d5d452a85ba27a0` is merged first and exact current
-parent `7abce9ad767fe8311da66a1e5998b892ea3ca9de` second. Implementation and
-test paths merge automatically: the child workspace/toolstrip, visibility,
-navigation, playback, and independent-plot behavior remain intact while the
-parent contributes current launch-monitor/D-plane ancestry, split kinetics,
-and behavior-preserving Qt return narrowing. Only append-only handoffs and
-SPEC conflict textually; both histories are preserved. All 142 exact
-PR-delta Python tests and all 32 tests in the eight changed React suites pass
-before the merge commit. Pinned Ruff checks/formats all 27 changed Python
-files; pinned MyPy and Bandit accept all 18 production files; docs, 500-LOC,
-changed-Python policy, test-contract/assertion, diff, React type-check/lint, and
-production-build gates pass. The later campaign-manifest artifact/checker/test
-exist on neither side of this pre-manifest stack and are not applicable.
-Review, ordinary publication, protected exact-head CI, unresolved threads,
-dependency order,
-and release remain open.
-## 2026-08-11 Remote workspace history reconciled locally
-
-PR `#4279` retains branch `feat/4218-toolstrip-workspace` and base
-`feat/4181-launch-monitor-registry`. Exact local head
-`0b22c401a26c31441a599d8d9b39de123706e7ea` ordinarily merges divergent
-remote head `61fe2d556a5413e525d958612ccfd57e65b8d5a2`, preserving every commit in
-both histories. The remote automation commit is formatting-only: 15 of 23
-paths were already identical through the parent, while its seven unique edits
-are normalized back to protected Ruff 0.14.10 output. Its one content conflict
-referenced the obsolete pre-split kinetics implementation. The reconciled tree
-keeps the current `pendulum.sample(...)` façade path and therefore preserves
-the kinetics split, physics, frames, units, imports, schemas, and UI behavior. Workspace/toolstrip,
-visibility, navigation, playback, and independent plot behavior are unchanged.
-Publication remains blocked on independent review, fresh exact-head protected
-CI, unresolved-thread checks, and dependency order.
-
-## 2026-08-11 Workspace child receives hosted MyPy repair
-
-PR `#4279` retains branch `feat/4218-toolstrip-workspace` and base
-`feat/4181-launch-monitor-registry`. Exact child
-`7806a16f58e1c6999d32f0127a187fbb21f839a1` normally merges exact published
-parent `3796b49e40b677fbac4e05739f8be49f905df2cb`. No feature-code conflict
-exists: the inherited production delta is four static `numpy.ndarray` casts,
-while workspace/toolstrip behavior and runtime arrays, physics, frames, units,
-imports, and UI behavior remain unchanged. This local merge does not satisfy
-fresh protected CI, required review, unresolved-thread, dependency, or release
-gates.
-
-## 2026-08-11 workspace child receives latest repaired parent
-
-PR `#4279` retains branch `feat/4218-toolstrip-workspace` and base
-`feat/4181-launch-monitor-registry`. A normal merge incorporates exact parent
-`0216a547aa79727091a2939b96e779e8ddbd7304` into exact child
-`61b7f48b5aeb7d57246b4963da3df086e79cbe15`. There are no feature-code
-conflicts: workspace/toolstrip, visibility, navigation, playback, and plot
-behavior remain unchanged while the parent kinetics façade/dynamics/series
-split and pinned formatting are inherited. This local merge does not satisfy
-fresh protected CI, required review, unresolved-thread, dependency, or release
-gates.
-
-## 2026-08-11 Append-only SPEC preservation repair
-
-Independent review found that the local #4203 reconciliation omitted four
-exact D-plane parent SPEC rows dated 2026-08-10 (versions 1.13.11, 1.13.9,
-1.13.7, and 1.13.6). They are restored verbatim in a documentation-only
-follow-up. Production behavior, tests, ordered parents, base, and local quality
-evidence are unchanged; re-review, normal publication, and protected CI remain
-open.
-
-## 2026-08-11 Current D-plane parent reconciliation
-
-PR #4203's published child `217e36dc93d30f79826847f958fbcd10805e58ed`
-is being normally merged with exact current D-plane base
-`f3363aa88868f6a5c7e9ccfc682a9eca014e86c1`. The base and source branch stay
-unchanged. The only conflict is the parent's formatting of an older monolithic
-kinetics expression: the split facade keeps its typed pendulum accessor and
-inherits the parent's geometry explanation. No launch-monitor, D-plane, or
-kinetics behavior is intentionally changed. Four Qt-stub `Any` returns exposed
-by the exact MyPy 1.13 delta are also narrowed to their declared `bool`/`str`
-contracts without changing values. Focused/full validation, independent
-review, normal publication, and protected CI remain open.
-
-## 2026-08-11 Exact-head format completion
-
-Protected CI on #4203 head `7d69a545ae555679f0318940e67c1786626d6794`
-failed only Ruff formatting. The pinned 0.14.10 reproduction found eleven
-noncompliant changed Python files: four inherited files plus seven altered by
-the automated pre-commit repair. The pending repair intentionally changes no
-launch-monitor, kinetics, simulation, or UI behavior and requires independent
-review plus fresh protected CI before integration.
-
-## 2026-08-11 Hosted MyPy NumPy-return repair
-
-CI Standard run `31477542889`, job `93734652129`, found four
-`no-any-return` errors after the kinetics façade extraction at exact PR #4203
-head `0216a547aa79727091a2939b96e779e8ddbd7304`. The NumPy stubs widen
-`linalg.norm`, `concatenate`, and matrix projection results to `Any` even
-though the runtime operations return arrays. The private series/dynamics
-modules and public façade now use explicit `cast(np.ndarray, ...)` boundaries.
-These casts are static only: array identity, values, dtype, shape, physics,
-frames, units, imports, and UI behavior are unchanged. Verification must cover
-the complete PR changed-source MyPy set and both focused and full Rate tests.
-Current local evidence is 102 changed-source MyPy files, 28 focused tests, and
-all 701 Rate tests passing; complete-delta Ruff/format and Bandit are also
-green.
-
-## 2026-08-11 Kinetics façade extraction
-
-The strict current-head delta reproduces a protected size failure that the
-PR-base comparison hides: `kinetics.py` is changed versus `HEAD~1` and is 646
-LOC, above the ungrandfathered 500-LOC maximum. The public module is now a
-222-LOC run adapter/facade; its pure double-pendulum dynamics are in the
-205-LOC `_kinetics_dynamics.py`, and the immutable `KineticsSeries`/DbC
-contract is in the 131-LOC `_kinetics_series.py`. Existing public imports and
-the `_reaction_forces` seam are the same implementation objects. A RED-first
-identity test plus the existing physics, energy, force, parity-fixture,
-presentation, and PyQt tests protect behavior. No physics, frame, SI-unit,
-schema, UI, or stack-order change is intended; the commit remains local only.
-Focused kinetics/presentation/PyQt verification is 28 passing tests; the
-complete Rate-of-Closure Python regression suite is 701 passing tests.
-
-## 2026-08-11 #4203 pinned-Ruff formatting repair
-
-No material handoff change: current-head CI Standard run `31468208320`, job
-`93705508050`, identified eight files that differ from repository-pinned Ruff
-0.14.10 output. This commit changes formatting only; Rate physics, behavior,
-frames, persistence schemas, PyQt/React contracts, stack bases, and dependency
-order remain unchanged. Fresh protected CI and review are still required.
-
-## 2026-08-10 Ground/Tee parity child receives repaired parent
-
-Ready PR `#4325` keeps branch `feat/4143-tee-parity-fixture` and base
-`feat/4181-launch-monitor-registry`. It normally merges exact repaired parent
-head `12dd76a8dbcc106c4683f2f2e53076f8dc6f1b76` without any production/test
-code conflict or history rewrite. Shared parity and rendered-evidence
-contracts remain intact; fresh protected CI, review, dependency order, and
-#4143 release remain open.
-
-## 2026-08-10 #4143 Python/React Golden Ball-Setup Parity
-
-The isolated `feat/4143-tee-parity-fixture` child begins at exact draft PR
-#4203 head `31cbc007d4c85b5479b7cd0fb0969124eab2af67`. One versioned JSON fixture now
-drives both Python and React ball-support tests with explicit SI units and the
-ground-plane-to-ball-bottom height reference. The shared cases cover Driver
-and non-Driver defaults, user overrides, Ground zero effective height,
-derived ball-center geometry and serialization, negative/NaN/infinite height
-rejection, and backward-compatible migration of a legacy run without
-`ball_setup`.
-
-Evidence is 18 passing Python tee/parity tests and 24 passing React
-tee/persistence/parity tests, plus green TypeScript, ESLint, Vite production
-build, Ruff check, and Ruff format. Production model and UI code are unchanged.
-Recorded visual evidence now adds 1600 x 1200 Playwright Driver/Tee and
-rerun-Ground captures plus 1400 x 900 hidden-window PyQt captures. The browser
-run asserts control state, diagram geometry, and zero console/page errors; the
-desktop run asserts canonical center/artist state, nonblank output, and
-different Ground/Tee digests without pixel absolutes. The exact artifacts and
-versioned manifests are in
-`C:\Users\diete\AppData\Local\Temp\rate-4143-visual-evidence-8050eeba`.
-Do not close #4143: protected current-head CI/review and release to `main`
-remain. This exact parent predates the strict campaign manifest on a divergent
-branch, so no downstream manifest was copied into this bounded child.
-
-## 2026-08-10 Wind scalar child receives second repaired parent
-
-PR `#4281` retains branch `feat/4199-wind-scalar-adapter` and base
-`feat/4144-variation-export-continuation`. It normally merges exact repaired
-parent `b90e5021a59e2081415b51ef29fbed06377bc201` without scalar-adapter code
-conflict or history rewrite. Fresh child CI, review, and dependency order
-remain required.
-Post-reconciliation evidence is 25 focused D-plane/impact tests plus docs,
-changed-file-size, and whitespace gates.
-
-## 2026-08-10 Variation export child receives exact workspace parent
-
-PR `#4280` retains branch `feat/4144-variation-export-continuation`, base
-`feat/4218-toolstrip-workspace`, and original child first parent
-`f90836e342efc8be624739802375af2876d11e5f`. Exact parent
-`6717e9e09d507dbc24bedb36177f1cdf0b4fd90b` merges normally as the second
-parent. All variation visualization/export source merged without conflict:
-selected scatter CSV, typed unavailable rows, accessible bounded raw tables,
-linked trials, focused PyQt widgets, and React parity remain additive with the
-parent workspace/toolstrip/playback/plot/navigation repairs. SPEC 1.14.12 is
-the unique combined child entry. Staged review exposed a stale linked-trial
-selection when a smaller rerun replaced the active result. Corrected PyQt
-views clear result-local selection before repopulating and validate all public
-setter indices; React clears on result identity change and shares only bounded
-selections with linked views. Thirty-seven focused variation tests and the
-complete Rate/shared-swing/golf-club matrix pass 1,528 tests with two explicit
-optional build123d skips; all 546 React tests across 90 files pass;
-TypeScript, ESLint, and production build are green. Exact-parent
-Ruff/Ruff-format/Black, pinned MyPy 1.13 across four Python production modules,
-Bandit, file-size, docs, minimum-test, assertion, detect-secrets, diff, real
-CPython 3.10.20 compilation, and 30 compatibility regressions pass. There is
-no Rust delta from the exact parent. Independent staged re-review found no
-actionable findings, including the first replacement-render React boundary.
-Protected CI/review remain release gates after an ordinary guarded push.
-
-## 2026-08-10 Variation child receives second repaired workspace parent
-
-PR `#4280` retains branch `feat/4144-variation-export-continuation` and base
-`feat/4218-toolstrip-workspace`. It normally merges exact repaired parent head
-`61b7f48b5aeb7d57246b4963da3df086e79cbe15` without feature-code conflict or
-history rewrite. Variation/export behavior remains unchanged; fresh child CI,
-review, and dependency order remain required.
-Post-reconciliation evidence is 25 focused D-plane/impact tests plus docs,
-changed-file-size, and whitespace gates.
-
-## 2026-08-10 Wind scalar child receives exact variation parent
-
-PR `#4281` keeps branch `feat/4199-wind-scalar-adapter`, base
-`feat/4144-variation-export-continuation`, and original child first parent
-`cf52529b1e68479321bb93b1be3d59c77f782008`. Exact parent
-`8bcc49fc4e16e5e43be0b7f0f03c3017d5b79d0c` merges normally second. The
-child's scalar-ensemble wire contract, provenance-preserving wind rows,
-availability accounting, and plot adapter remain additive with the parent's
-variation visualization/export and result-local selection repairs. Python and
-TypeScript reject wind scenarios unless their per-trial provenance is exact,
-shear/turbulence are zero, seed is zero, and gusts are empty. SPEC 1.14.13 is
-the unique combined child entry. Twenty-seven focused Python tests, the
-1,549-test complete Rate/shared-swing/golf-club matrix with two explicit
-optional build123d skips, and all 556 React tests across 92 files pass.
-TypeScript, ESLint, production build, Ruff/format, pinned MyPy, Bandit, size,
-docs, minimum-test, assertions, detect-secrets, diff, and CPython 3.10 compile
-gates pass. Independent re-review found no actionable findings after checking
-the exact provenance derivation and all five deterministic-scenario
-regressions; protected CI/review remain release gates.
-
-## 2026-08-10 Variation child timestamp propagation
-
-The #4280 continuation now carries exact parent
-`05383d333b6fd87eaf5e37305476f50b505c2c2e` through a normal merge while
-preserving its original base and complete variation export/accessibility
-implementation. The inherited workspace parser now has one strict UTC grammar
-and consistent zero- through six-digit fractional-second behavior across
-Python 3.10-3.12. The child variation behavior is unchanged.
-
-SPEC 1.14.10 records the parent compatibility repair and 1.14.11 records the
-child variation delta. Keep this handoff, the root and campaign handoffs, and
-the reconciled SPEC in the merge commit. The combined evidence is `778 passed`
-for the Rate suite, `27 passed` on real CPython 3.10.20, and `1 file / 8 tests`
-for focused React variation, with TypeScript, focused ESLint, Ruff, format, and
-pinned mypy 1.13 clean. Normal publication and dependency-ordered propagation
-remain separate gates.
-
-## 2026-08-10 Workspace child receives second repaired parent
-
-PR `#4279` retains branch `feat/4218-toolstrip-workspace` and base
-`feat/4181-launch-monitor-registry`. It normally merges exact repaired parent
-head `12dd76a8dbcc106c4683f2f2e53076f8dc6f1b76`; there is no feature-code
-conflict, rebase, retarget, or history rewrite. The inherited explicit ndarray
-boundaries do not alter numerical behavior. Fresh child CI, review, and
-dependency order remain required.
-Verification after reconciliation is 25 focused D-plane/impact tests plus
-docs-governance, changed-file-size, and whitespace gates.
-
-## 2026-08-10 Workspace child receives exact repaired parent
-
-PR `#4279` keeps branch `feat/4218-toolstrip-workspace` and base
-`feat/4181-launch-monitor-registry`. Its original head
-`05383d333b6fd87eaf5e37305476f50b505c2c2e` normally merges exact parent
-`31cbc007d4c85b5479b7cd0fb0969124eab2af67` in that order. The combined
-shell retains the child's top toolstrip, persistent module visibility,
-reorderable workspace, granular playback, path trail, and independent plot
-controls. Parent repairs remain additive: the workspace mixin reuses the
-canonical navigation constants, and the simulation controls reuse the single
-persisted `ImpactLayerControls` checkbox map without duplicating UI state.
-
-Focused and broad verification is green: 1,339 Python tests pass with six
-explicit optional/Rust-wheel skips; 545 React tests pass, as do type-check,
-lint, and production build. Forty navigation/simulation GUI tests pass after
-the final formatting change. Exact-parent Ruff/Ruff-format/Black, pinned MyPy
-1.13 for 18 production modules, Bandit, file-size, docs, minimum-test,
-assertion, detect-secrets, diff, CPython 3.10.20 compilation/import, and 30
-compatibility regressions pass. There is no Rust delta from the exact parent.
-The merge commit must keep this file, the root handoff, campaign handoff, and
-`SPEC.md` together. Independent staged review found no actionable findings
-after 76 additional focused PyQt/navigation/workspace tests; exact-head
-protected CI and required review remain release gates after a guarded ordinary
-push.
-
-## 2026-08-10 Workspace fractional-timestamp compatibility repair
-
-The Python 3.10 lane on descendant PR #4281 revealed that the strict workspace
-ordering test reached `datetime.fromisoformat` with a one-digit fractional
-second. Python 3.10 rejects that spelling before comparing timestamps, although
-newer supported interpreters accept it. PR #4279 is the earliest owning branch.
-The workspace validator now enforces one anchored UTC grammar and parses every
-losslessly representable zero- to six-digit fraction consistently, rejecting
-greater precision rather than allowing interpreter-dependent truncation. It
-does not modify the persisted spelling, UTC requirement, schema, or ordering
-rule.
-
-Evidence is `778 passed` for the full Rate suite, `45 passed` across the
-compatibility and workspace document tests on the local interpreter, and
-`27 passed` under real CPython 3.10.20. Ruff, format, pinned mypy 1.13, docs
-governance, and the 400-line budget pass. Keep this handoff, the root handoff,
-campaign handoff, and SPEC entry in the same implementation commit; verify the
-exact remote head before a normal push and propagate through later descendants.
-
-## 2026-08-09 Universal Variation Visualization Continuation
-
-Branch `feat/4144-variation-export-continuation` remains based on
-`feat/4218-toolstrip-workspace` and now contains exact corrected parent
-`3f67ed466fefc8991db9c4409f921f25e1c37142` through the normal merge containing
-this handoff. The core PyQt6/React variation workspace has selectable
-input/contact/impact/shot scatter, a scatter matrix with marginals, landing
-dispersion, and an all-trial 3D swing-arc overlay with reference trace,
-principal spread, RMS variability, quiet zones, filtering, and linked trial
-selection.
-
-This continuation closes an export/accessibility parity gap: both clients now
-export the complete selected scatter axes as CSV, including every trial,
-typed outcome, and explicit unavailable values. PyQt now also exposes the raw
-selected-axis rows in a bounded read-only table. Shared table population is
-factored into `variation_trial_table.py`; the scalar scatter view is isolated
-in `variation_scatter_view.py`. Changed production modules remain below 400
-lines and changed functions remain at or below 50 lines.
-
-The continuation is published as draft PR #4280. Current exact local evidence
-on implementation commit `b3b09215e` is 890
-Python/PyQt/shared swing tests passed with one expected optional-Rust skip and
-15 existing warnings, and 89 React test files / 545 tests passed. Ruff,
-formatting, Black, targeted mypy, TypeScript, zero-warning ESLint, the 166-module
-Vite build, and `git diff --check` pass. This is local evidence only; no PR,
-protected CI, review, merge, or default-branch release has yet been established.
-
-The current local continuation normally incorporates exact corrected parent
-`3f67ed466fefc8991db9c4409f921f25e1c37142` while preserving #4280's
-`feat/4218-toolstrip-workspace` base. Its current focused evidence is `46`
-Python 3.11 tests and `19` real-Python-3.10.20 tests with PyQt6 installed,
-including the merged compatibility and variation widget contracts. React
-passes `1 file / 8 tests`, TypeScript, and focused zero-warning ESLint. Ruff,
-format, pinned mypy 1.13, documentation governance, ancestry/SPEC assertions,
-and diff checks pass. SPEC 1.14.11 is the unique child entry for the
-scatter-export and accessible-row-table source change. This propagation is
-local only and has not been pushed.
-
-Literal audit of "every many-trial simulation" found two dependent gaps that
-must remain explicit:
-
-- wind-strategy uncertainty retains paired per-strategy outcomes in both
-  runtimes but has no PyQt/React workflow and does not feed the universal
-  scalar plot facade;
-- the capability optimizer stores aggregate statistics but discards the
-  individual evaluation rows needed for honest scatter plots.
-
-Issue #4199 already owns the wind UI/scatter requirement. Its adapter must use
-composite row identity `(strategy_id, trial_index)`, retain completed,
-nonconverged, and invalid cohorts, and report that impact variables are
-unavailable because this runner begins at prescribed launch. Do not coerce it
-into the impact-specific cohort enum or fabricate impact/landing values.
-
-Branch `feat/4199-wind-scalar-adapter` is published as
-[draft PR #4281](https://github.com/D-sorganization/Tools/pull/4281), stacked
-on exact corrected draft PR #4280 head
-`38ed58cab96842a3007e76a855db83ee2452b8fd`. That parent is incorporated by a
-normal local merge while preserving base
-`feat/4144-variation-export-continuation`; no branch was rebased, retargeted,
-force-pushed, or published by this continuation.
-Implementation commit `4a28114aa` supplies that first UI-neutral model slice.
-Python and React share the exact
-snake-case `scalar-ensemble/v1` wire structure: structured provenance,
-labeled stages/categories/cohorts, unit-bearing variables, RFC3986 composite
-row identity, immutable nullable raw rows, and overall/per-cohort x/y/paired
-availability. The wind adapter validates deterministic request/analysis trial
-agreement, retains every actual and perfect-information status, and exposes
-true/estimated wind, launch/aim, target, landing, miss, cost, and information
-delta without invoking the flight solver. Impact variables remain honestly
-absent because this analysis begins at launch.
-
-The original broad local gates remain green: 906 Python/PyQt/shared-swing tests
-passed with one expected optional-Rust skip and 15 existing warnings; 91 React
-files / 555 tests passed. After corrected-parent propagation, focused evidence
-is `49 passed` on Python 3.11 and `49 passed` on real CPython 3.10.20 with
-PyQt6/scientific dependencies present, plus `2 React files / 10 tests`,
-TypeScript, and zero-warning focused ESLint. Ruff check/format passes all four
-child Python files; pinned mypy 1.13 passes both production modules and the
-wind-adapter test after its status helper received an exact `Literal` type.
-The function budget passed at that historical head, but the 403-line wind
-adapter did not satisfy the mandatory below-400 module ceiling; the 2026-08-11
-repair above is the first point at which that module claim is true. SPEC
-1.14.11 uniquely records the child change above parent 1.14.10. This
-contract/adapter does not complete #4199: the next slice still needs background
-execution, progress and cancellation, PyQt/React scatter/strategy UI,
-persistence, and export wiring.
-
-## 2026-08-07 Ground Model and Four-Surface Parity Revalidation
-
-The user's requested rolling-ground and site-wide parity programs already
-exist as [epic #4267](https://github.com/D-sorganization/Tools/issues/4267)
-with children #4268-#4276 and
-[epic #4260](https://github.com/D-sorganization/Tools/issues/4260) with
-children #4261-#4266. Do not create duplicate epics. A fresh repository and
-live-GitHub audit was recorded in
-[the ground comment](https://github.com/D-sorganization/Tools/issues/4267#issuecomment-5222725556)
-and [the parity comment](https://github.com/D-sorganization/Tools/issues/4260#issuecomment-5222726010).
-
-There is no qualified landing/bounce/roll implementation yet. The current
-airborne solvers stop at the relative launch plane and the shared
-`TrajectoryPoint` omits terminal angular velocity. For a teed shot, translating
-that relative trajectory into the course frame can leave the terminal ball
-center at tee elevation. Complete #4268 and #4269 before implementing or
-presenting bounce, roll, or total distance.
-
-Reuse the Tools `GroundModelResult` fail-closed boundary, putting skid/roll
-limiting cases, and turf provenance/variation/cancellation patterns. Adapt
-UpstreamDrift's split terrain material, elevation, normal, and region contracts
-through a one-way versioned DTO. Do not qualify Upstream's scalar landing
-helper, heuristic putting spin relaxation, legacy duplicate terrain model, or
-Rust `(1-friction)` contact law as production high-speed landing physics.
-
-The parity matrix must keep Rate and Upstream's separate products distinct:
-standalone Rate PyQt6/React, Upstream's Rate PyQt provider/React route, Shot
-Tracer PyQt6/React, and the legacy ball-flight GUI. Current Upstream `main`
-has no native Rate React route, and its Tools source resolvers disagree about
-vendor-first versus sibling-first precedence. A launcher tile cannot satisfy a
-calculation-parity row; #4261, #4262, and #4264 must make the runtime source,
-exact Tools pin, and support state machine-verifiable.
-
-## 2026-08-09 Workspace Python 3.10 compatibility completion
-
-The #4279 continuation now contains exact corrected parent
-`08a2fdd8ce6bbc8fbb8f121927a677d4addb6b11` through normal local merge
-`a340fabefa443d47325c5538f342683b38c01ade`, without rebasing, retargeting,
-force-pushing, or publication. The workspace command and view enums use the
-shared `StrEnum` runtime compatibility contract with native enum typing kept
-under `TYPE_CHECKING`; `_workspace_validation` uses the shared `UTC` value.
-Stable identifiers, enum values, UTC serialization, schemas, and behavior are
-unchanged.
-
-The combined regression inspects the runtime branch for all nine inherited
-and child string-enum modules plus both UTC modules and directly executes the
-three child workspace modules. Focused Python 3.11 evidence is `126 passed`.
-Real CPython 3.10.20 evidence is `14 passed` plus successful dotted imports of
-10 Rate/shared targets. Ruff check/format is clean, and pinned mypy 1.13 passes
-all 11 affected production files under the changed-file CI contract. Keep the
-root, tool, campaign handoffs and SPEC 1.14.9 entry in the implementation
-commit; protected CI and review remain required after a normal push.
-
-## 2026-08-09 Workspace stack parent propagation
-
-Draft PR #4279 keeps base `feat/4181-launch-monitor-registry`. Exact published
-parent head `08a2fdd8ce6bbc8fbb8f121927a677d4addb6b11` is carried by a normal
-merge without rebasing, retargeting, force-pushing, or publishing this local
-continuation. Source changes applied cleanly; overlapping handoff/specification
-history was reconciled monotonically. The result retains the parent's facade
-and Python 3.10 repairs plus the child's complete File/View/Tools registry,
-strict workspace document, matched PyQt6/React toolstrips, module
-visibility/order, playback, and independent plot controls.
-
-The reconciled ancestry passes `126` focused Python/PyQt tests and `8 React
-files / 32 tests`. Ruff check/format passes all 28 relevant Python files, and
-CI-pinned mypy 1.13 passes the 18 changed production files and both facade
-tests. Mypy caught an untyped Qt `isChecked()` result in the child's
-`legend_visible()` boundary; the method now returns an explicit `bool` without
-changing runtime behavior. The affected simulation GUI rerun is `29 passed`;
-documentation governance and diff checks pass. Keep this handoff, the root
-handoff, the campaign handoff, and `SPEC.md` together in the merge commit.
-
-## 2026-08-10 Second D-Plane Parent Repair Propagation
-
-Draft PR `#4203` keeps base `feat/4189-dplane` and normally merges exact
-repaired parent head `7d8d2f06dc797021d01939691e58f8425b652b33` without
-rewriting either branch. The inherited change makes the two private D-plane
-NumPy return boundaries explicit for pinned MyPy while preserving numerical
-semantics, frames, schemas, and UI behavior. The parent quality gate is green;
-child protected CI, review, dependency order, and #4189 closure remain open.
-Post-reconciliation evidence is 25 focused D-plane/impact tests plus docs,
-changed-file-size, and whitespace gates. Local pinned-type evidence remains the
-parent's green hosted quality gate: Windows MyPy 1.15 encounters incompatible
-installed NumPy stub syntax and WSL is unavailable with `E_FAIL`, so neither
-local attempt is represented as green.
-
-## 2026-08-10 Launch-Registry Child Propagation
-
-Draft PR `#4203` (`feat/4181-launch-monitor-registry`) retains its
-`feat/4189-dplane` base and normally merges original child head
-`08a2fdd8ce6bbc8fbb8f121927a677d4addb6b11` with exact parent head
-`b443fdbed7064c5db0320106013c8413e3e24356`.
-
-The reconciliation preserves the child branch's responsive
-`SimulationViewControlsMixin` and delegates its persisted D-plane layer
-checkboxes to the parent's `ImpactLayerControls` helper. The compatibility
-mapping and helper mapping are the same object, which prevents duplicate state
-while retaining existing UI automation and persistence behavior. Both PyQt
-modules remain below the 500-line limit.
-
-The original child already exceeded the protected budget in swing sources,
-the plotting catalog, and the PyQt main window. Focused modules now own the
-triple-pendulum model, immutable plotting-variable contract, and versioned tab
-state. Compatibility imports remain the same objects, while the former
-monoliths fall to 282, 459, and 494 lines. Focused evidence is green across 36
-PyQt simulation/layout tests, 38 plotting/navigation tests, and 21 simulation
-source/export tests. Combined verification is green across 1,249 Python tests
-with six explicit skips, 521 React tests and all web gates, 12 Rust tests,
-real CPython 3.10 checks, scoped Ruff/Black/pinned MyPy, and repository
-governance/security checks. The changed-file 500-line gate passes all 107
-candidate files. A full-tree audit still reports untouched `kinetics.py` and
-`torque_profile_panel.py`; neither differs in this propagation. Independent
-staged review found no actionable findings after 95 additional focused tests.
-Current-head protected CI and required repository review remain release gates.
-
-## 2026-08-09 Python 3.10 enum import repair
-
-PR #4203 now owns the earliest fix for seven inherited Rate/shared swing
-modules that imported Python 3.11's `enum.StrEnum` directly. Runtime code uses
-the established shared compatibility helper while mypy sees the native enum
-only inside `TYPE_CHECKING`. This preserves wire values and enum semantics on
-supported interpreters and lets the repository's hosted Python 3.10 lane
-collect the modules. The focused evidence is 64 tests, Ruff/format, pinned
-mypy 1.13 across eight changed files, and a real CPython 3.10.20 source/runtime
-probe. The exact published #4203 head is
-`ab7de5a47977417e02926c3fbc7476002e82b690`; propagate it through the existing
-stack without changing bases.
-
-A follow-up scan found the torque-profile controller importing
-`datetime.UTC` directly. That parent-owned Python 3.11 boundary now uses
-`shared.python.compatibility.UTC`; persisted timestamps and torque-profile
-behavior are unchanged. Re-run the focused torque-profile UI tests and the
-real Python 3.10 compatibility probe before propagation.
-
-## 2026-08-09 Launch-registry parent CI repair
-
-PR #4203 exact-head run `31199764932` reached the Python test lanes but failed
-during Linux collection, before behavioral assertions. The two in-package
-flight/solver facade tests were collected as `src.shared...` modules while
-their absolute aliases crossed into the editable `shared...` namespace. They
-now import their sibling facade package relatively, preserving the pinned
-public API contract and production behavior. Reproduce with pytest
-`--import-mode=importlib`; keep the separate Rust missing-`libpython3.11`
-failure classified as runner infrastructure. Publish normally and propagate
-the repaired parent through the existing stack without changing PR bases.
-Verification is `12 passed` on Windows and `12 passed` on WSL Python 3.11
-under importlib collection; Ruff/format and exact mypy 1.13 pass for both
-changed modules. The dataclass metadata assertion remains active behind an
-explicit test-only `Any` introspection boundary.
-
-## 2026-08-10 PR #4202 D-plane ndarray typing repair
-
-On top of verified published head
-`b443fdbed7064c5db0320106013c8413e3e24356`, the two private D-plane ndarray
-helpers now bind NumPy results to explicit ndarray locals before returning
-them. This closes the exact MyPy 1.13 `no-any-return` findings from CI Standard
-run `31384810375`, job `93442745760`, without changing numerical operations,
-DbC validation, reference frames, persistence, rendering, or public contracts.
-
-The exact MyPy failure was reproduced before editing and is green after the
-repair. Twenty-four focused D-plane and impact integration tests pass, along
-with seven metadata/pre-push contract tests, scoped Ruff, Ruff format, Black,
-docs governance, minimum-test, module-size, changed-file-size, and diff checks.
-Three unrelated CI-workflow contract tests still expect later toolcache/env
-steps not present on this older branch; no workflow file is in scope. This
-local commit does not push, retarget, change draft state, or claim issue
-completion; protected CI, review, and parent-first release order remain
-mandatory.
-
-## 2026-08-10 D-Plane Child Propagation
-
-Draft PR `#4202` (`feat/4189-dplane`) retains its
-`feat/4162-wedge-impact-visualization` base and normally merges original child
-head `b4abec03bccfbdd87ddf91427159c5c2332c21dd` with exact parent head
-`6704a3e541a3e74c28b4a284530d1a21269dd340`. The inherited Python 3.10 UTC
-repair and AST guard remain intact beside the frame-explicit D-plane contract.
-
-Persisted D-plane layer controls now live in a focused helper so
-`simulation_view.py` again satisfies the protected 500-line budget while
-retaining the existing UI-automation compatibility seam. Combined verification
-is green: 93 focused and 825 scoped Python tests (two optional `build123d`
-skips), 360 React tests and all web gates, real CPython 3.10.20
-compilation/UTC, scoped Ruff/Black/MyPy, and repository governance gates. The
-exact parent's 12 unchanged `swing-core` tests remain applicable because this
-child has no Rust delta. The 17-error broad MyPy Qt/NumPy baseline in 11
-untouched files remains separate. Protected CI and required review remain
-release gates.
-
-## 2026-08-10 Impact-Visualization Child Propagation
-
-Draft PR `#4179` (`feat/4162-wedge-impact-visualization`) retains its
-`feat/4166-wedge-turf-physics` base and normally merges original child head
-`0eb804e70887c788421332369e42792411aff55a` with exact parent head
-`bfa83aedc88ead380babc73a699377d98b971006`. The inherited Python 3.10 UTC
-repair and AST guard remain intact beside the exact-event scene contract.
-
-Combined verification is green: 58 focused and 739 scoped Python tests (two
-optional `build123d` skips), 347 React tests and all web gates, real CPython
-3.10.20 compilation/UTC, scoped Ruff/Black/MyPy, and repository governance
-gates. The exact parent's 12 unchanged `swing-core` tests remain applicable
-because this child has no Rust delta. The 17-error broad MyPy Qt/NumPy baseline
-in 11 untouched files remains separate. Protected CI and required review
-remain release gates.
-
-## 2026-08-10 Turf-Physics Child Propagation
-
-Draft PR `#4178` (`feat/4166-wedge-turf-physics`) retains its
-`feat/4161-wedge-ground-clearance` base and normally merges original child
-head `aaae3f73e17dbfaad5cca1dc6f49559b3aebe9d5` with exact parent head
-`9ea93e92563280ec34bca682ad44d7409edd7a02`. The inherited Python 3.10 UTC
-repair and AST guard remain intact beside the provenance-gated turf model.
-
-Combined verification is green: 56 focused and 732 scoped Python tests (two
-optional CAD-dependency skips), real CPython 3.10.20 checks, scoped
-Ruff/Black/MyPy, and repository governance gates. The unchanged TypeScript and
-Rust surfaces retain the exact parent's green 345 React and 12 Rust test
-evidence. The 17-error broad MyPy Qt/NumPy baseline in 11 untouched files
-remains separate. Protected CI and required review remain release gates.
-
-## 2026-08-10 Ground-Clearance Child Propagation
-
-Draft PR `#4174` (`feat/4161-wedge-ground-clearance`) keeps its
-`feat/4163-impact-inspector` base and normally merges original child head
-`880a6465fc872cf3d6650283db154ddc41793a31` with exact parent head
-`9ddaff3b6bca542fd7a2befc7d7b0ae53910a60a`. The inherited Python 3.10 UTC
-repair and AST guard remain intact beside the ground-clearance analysis.
-
-Combined verification is green: 56 focused and 703 scoped Python tests (two
-optional `build123d` skips), 345 React tests and all web gates, 12 Rust tests,
-real CPython 3.10.20 compile/UTC checks, scoped Ruff/Black/MyPy, and repository
-governance checks. The existing 17-error broad MyPy Qt/NumPy baseline across
-11 untouched files is documented, not expanded. Current-head protected CI and
-required review remain pending.
-
-## 2026-08-10 Python 3.10 Repair Propagation
-
-Draft child PR `#4173` (`feat/4163-impact-inspector`) retains its
-`feat/4144-variation-visualizations` base and normally merges original child
-head `3c43955aaeb3964ff8c3ef2748d626baae518b76` with exact parent head
-`22b66b560652b78de84141344c4ddd9a92a83b26`. This carries the shared
-Python 3.10-compatible UTC export and the source-wide AST guard into the wedge
-impact inspector without changing the persistence schema or user-visible
-timestamp format.
-
-Combined-stack verification is green across 63 focused Python tests, all 562
-Rate tests, all 334 React tests, TypeScript/ESLint/Vite gates, 12 `swing-core`
-tests, real CPython 3.10.20 compile/UTC checks, Ruff/Black, focused pinned MyPy
-1.13, and repository governance checks. The broad MyPy sweep retains 17
-pre-existing Qt/NumPy typing findings in 11 untouched files. The PR must remain
-draft until its exact-head protected checks complete and required review
-approves. Do not retarget, rewrite, force-push, admin-merge, or count
-infrastructure failures as passing evidence.
-
-## 2026-08-05 Advanced Wedge Impact Visualization
-
-Branch `feat/4162-wedge-impact-visualization` extends issue #4162 on top of the
-validated turf stack. It corrects the impact adapter to evaluate pose, twist,
-and articulated wrist geometry at the exact event time rather than silently
-using the nearest retained sample. The new versioned
-`rate-of-closure.impact-scene/v1` contract carries complete scene geometry,
-velocity components, metric equations, frames, assumptions, availability, and
-screw-axis data without placing physics in either UI.
-
-PyQt6 adds an exact-event Impact Inspector layer, locked physical axis scaling,
-isometric/face-on/down-the-line cameras, and 300-DPI PNG, vector SVG, and strict
-JSON export. React adds an orbitable and keyboard-controllable impact still,
-the same named cameras and velocity toggles, visibly expandable engineering
-metric definitions, and device-resolution PNG, true-primitive SVG, and JSON
-exports. The web mirror now retains and shortest-arc interpolates the canonical
-head rotation; the older limitation note saying it lacked full head pose is no
-longer accurate for this branch.
-
-Scientific boundaries remain explicit: articulated sources do not yet have an
-independent torsional head state; the scene is rigid-body instantaneous
-kinematics; illustrative turf profiles cannot support optimal-bounce or
-forgiveness claims; and turf force is not replayed into the retained swing.
-
-Current-head verification: all 576 Rate Python/PyQt tests passed (one existing
-polynomial-generator legend warning); all 347 React/model tests passed; the
-production Vite build, TypeScript, ESLint, Ruff, formatting, changed-module
-strict mypy, and protected module-size budget passed. Headless Chrome visual QA
-at 1600×1400 exercised named views, a vector toggle, keyboard orbit, and an
-expanded metric definition with zero console exceptions/log errors. The new
-web branch is running at `http://localhost:5260/`; the current PyQt process was
-also launched successfully and remained responsive.
-
-## 2026-08-05 Wedge Impact Inspector Integration
-
-Draft PR #4173 (`feat/4163-impact-inspector`) integrates the draft variation
-branch with the shared golf-club stack through wedge kinematics PR #4172. It
-adds the first bounded implementation slice for wedge epic #4158 /
-impact-inspector issue #4163:
-
-- Canonical inspection time and event label on every `SimulationRun`: impact
-  for hits, closest approach for misses.
-- Exact `Jump to Impact` / `Jump to Closest Approach` controls in PyQt6 and
-  React, with playback paused before the jump.
-- `simulation/impact_kinematics.py`, which adapts retained pose/twist/contact
-  data to `shared.python.golf_club.WedgeKinematicState` and preserves geometry
-  provenance and model limitations.
-- Engineering readouts in both clients for contact/reference AoA, remove-shaft
-  counterfactual, shaft rotation and vertical velocity, face-normal rate,
-  leading-edge/arc rate where available, and screw-axis distance.
-- Restored manual angular velocity in the React simulation path; the previous
-  hard-coded zero made all closure and shaft metrics false zeros.
-- Deterministic midpoint tie-breaking for a flat maximum-speed plateau, so the
-  manual source's automatic event is its documented square pose at 30 ms.
-
-Physics boundary: articulated pendulum runs expose the measured wrist-to-head
-shaft line but have no shaft-twist degree of freedom. The inspector reports
-that absence rather than inventing torsional motion. The web mirror still does
-not retain full head pose; its readout declares that limitation until WASM or a
-canonical backend replaces the temporary TypeScript mirror.
-
-Current-head release evidence: 1,006 Python/PyQt/shared-swing tests passed with
-five optional Rust-wheel parity skips; 334 React/model tests and 12 swing-core
-Rust tests passed; Vite production build, TypeScript, ESLint, Ruff, formatting,
-and mypy across 165 source files passed. A focused post-refactor PyQt run passed
-46 tests. Browser QA verified the 1,307 deg/s manual fixture at 30 ms and a
-1.430 s closest-approach miss with zero console warnings/errors. Native-window
-QA confirmed the control and readout are visible in the standalone PyQt6 app.
-
-## Status Note
-
-`src/rate_of_closure` and `src/shared/python/swing_sim` do **not exist on
-`main` yet** — they land with PR #4119. This doc describes the tool as it
-exists on the in-flight branch stack (`feat/impact-simulation-platform` →
-`feat/investigation-suite` → `feat/course-showcase`) so the next agent has
-full context the moment #4119 merges. If you're reading this on a fresh
-`main` checkout and don't see `src/rate_of_closure/`, check out one of those
-branches or wait for #4119 to land.
-
-## Where This Tool Is Headed
-
-Rate of Closure started as a single-page "closure rate" calculator (twist
-model: GC-path vs impact-point-path gap, °/ft). Epics #4103 → #4120 → #4125 →
-#4130 are growing it into a full swing → impact → ball-flight simulator with
-PyQt6 + web parity and eventual public GitHub Pages distribution. Read
-`src/rate_of_closure/README.md` (frame conventions, Cheetham dossier sourcing,
-run instructions) before touching physics code.
-
-## The #4119 → #4124 → #4129 PR Stack
-
-Each PR consolidates a whole epic's stacked feature branches into one PR to
-keep self-hosted CI load down (see `CLAUDE.md` — these are big diffs).
-
-- **#4119** `feat/impact-simulation-platform` (epic #4103, open, auto-merge
-  armed). Base tool (twist model, PyQt6 + React/Vite web, Cheetham dossier
-  data) + STL clubheads/club library/inertial model + `swing_sim` shared
-  package (`src/shared/python/swing_sim/`: swing sources, impact model ported
-  from UpstreamDrift with 3 physics fixes, gear effect, 7 literature
-  ball-flight models, goal-driven multi-start solver) + `swing-core` Rust
-  crate (pyo3 + wasm) + app integration (simulation session, impact-time
-  scrubber, screw-axis overlay via the rotation_converter adapter, video
-  controls, CSV/JSON export, solver panel). 404 pytest + 72 vitest + 111
-  cargo tests, all local gates green. Supersedes #4092/#4097/#4098/#4112-4118.
-
-- **#4124** `feat/investigation-suite` (epic #4120, open, **draft state —
-  do not merge yet**, stacked on #4119). Adds `rate_of_closure/plotting/`
-  (40-variable data catalog, `PlotSpec`, Plots tab + custom-plot wizard),
-  scale-separated Strike/Swing/Flight viewers + standalone Flight Explorer,
-  `swing_sim/variation/` (seeded Monte Carlo/NoiseSpec engine, dispersion +
-  Spearman sensitivity, Variation tab), and V4: glossary (60 terms),
-  cold-user help system, "Derivation & Traceability" → "Calculation
-  Description" rename, full-model derivations, hover-hint completeness sweep.
-  Supersedes draft PRs #4121/#4122/#4123. 566 pytest + 125 vitest passing.
-
-- **#4129** `feat/course-showcase` (epic #4125, open, **draft state — do
-  not merge yet**, stacked on #4124). Merges `feat/realistic-heads` (H1:
-  parametric club-type geometry, volumetric COG via divergence theorem),
-  `feat/swing-kinetics` (H2: joint torque/force plots + 3D overlays from
-  pendulum inverse dynamics), `feat/putting-vertical` (H3: `swing_sim`
-  putting module + app Putting tab), then adds on top: H7 course scene
-  (`ui/course.py` / `course_scene.py`) + target optimization
-  (`swing_sim/solver/targets.py`, `TargetRegion`, `ImpactGoal.target_region`),
-  and H6 showcase styling (`ui/pyqt6/app_style.py`, UpstreamDrift launcher
-  visual language) + yards-default distance units. 413 pytest + 309 swing_sim
-  - 174 vitest passing. H4 (AffineDrift putting research content) and H5
-    (public release-management repo) are cross-repo and tracked in #4125
-    directly, not in this PR.
-
-**Do not merge #4124 or #4129 before their base merges** — SPEC.md sections
-were unioned assuming sequential merge order; merging out of order will
-produce conflicting/duplicate changelog rows.
-
-## swing_sim Packages
-
-`src/shared/python/swing_sim/` (introduced by #4119, home for physics shared
-with UpstreamDrift via the established shared-module arrow):
-
-- `swing_sim/flight/` — 7 literature ball-flight models (drag/lift/spin
-  decay), citations in registry metadata.
-- `swing_sim/impact/` — impact model ported from UpstreamDrift (offset-drop
-  fix, opt-in 3×3 MOI tensor, inverted friction spin axis fix), gear effect,
-  `SpringDamperImpactModel` (Kelvin-Voigt contact force history, 1e-7s steps)
-  — this is the contact-force law epic #4130 will extend for the full
-  contact-interval integration rather than duplicate.
-- `swing_sim/solver/` — goal-driven multi-start least-squares solver;
-  `targets.py` (added in #4129) adds `TargetRegion` for green/fairway
-  optimization goals.
-- `swing_sim/variation/` (added in #4124) — namespaced variable registry,
-  `NoiseSpec`/`VariationPlan`, seeded parallel N-run Monte Carlo engine,
-  dispersion/OAT sensitivity/Spearman/landing-ellipse stats.
-
-Epic #4130 (Impact-Interval Club Dynamics) will add `impact_interval/` to
-this same package — its home is explicitly `swing_sim` so UpstreamDrift
-reaches it via vendor, per that epic's F2 phase description. Not started yet
-(foundation-only epic, no PR).
-
-## How #4103/#4120/#4125/#4130 Relate to This Tool
-
-All four are rate_of_closure epics specifically (unlike the wider-monorepo
-epics tracked in the root `AGENT_HANDOFF.md`, e.g. SCADA #4085-#4089).
-#4103 is the foundation platform; #4120, #4125 are sequential feature waves
-stacked directly on its PR; #4130 is a physics-depth epic that extends the
-impact model #4103 introduced (contact-interval integration replacing the
-instantaneous-impulse approximation) — foundation phase only so far.
-
-## Web Mirror + GitHub Pages
-
-The web mirror (`src/rate_of_closure/web/`, React/Vite/TS) is pinned
-test-for-test against the PyQt6 model today (TS mirrors hand-written, not yet
-WASM — that swap is explicitly deferred to Phase 7 of #4103). It builds to a
-static bundle (`npm run build`) and carries Tauri scripts for desktop
-packaging, same as other web tools in the repo.
-
-**There is no automated GitHub Pages CI deploy for this tool yet.** No
-`.github/workflows/*.yml` references `rate_of_closure` or Pages deploy
-actions as of this writing. The only Pages-publishing precedent in the repo
-is `src/web_applications/unit_converter/unit-converter-app/DEPLOYMENT.md`'s
-manual branch-folder publish (Settings → Pages → select branch/folder).
-Phase 7 of #4103 ("GitHub Pages mirror updated (public share link), parity
-tests as deploy gates") owns building a real workflow — do not improvise one
-in an unrelated PR.
-
-## Must-Read Architecture Pointers
-
-1. `src/rate_of_closure/README.md` — frame conventions, unit conventions,
-   dossier sourcing, run/build instructions.
-2. `src/rate_of_closure/model.py` (base twist physics, no Qt) once #4119
-   lands.
-3. `src/shared/python/swing_sim/impact/` — the contact-force law shared
-   with (and about to be extended by) epic #4130.
-4. `rust_core/swing-core/` — pendulum EOM + plane projection, pyo3 + wasm
-   targets, follows the `tools-core` feature-contract pattern.
-5. `.github/workflows/maturin-swing-core.yml` (added by #4119) — Rust wheel
-   build for this crate.
-
-## Gate Commands (this tool)
-
-```bash
-python3 -m pytest tests/rate_of_closure src/shared/python/swing_sim -n auto --timeout=60
-cd src/rate_of_closure/web && npm run test && npm run build && npx tsc --noEmit && npx eslint .
-cargo test -p swing-core
-python3 -m ruff check src/rate_of_closure src/shared/python/swing_sim
-python3 -m mypy src/rate_of_closure src/shared/python/swing_sim
-```
-
-Run Ruff format/check and pinned delta mypy on every changed Python file. Do not
-rewrite parents, force-push, bypass checks, infer unavailable values, or close
-#4201 until exact protected merge, installed package, downstream pin, science,
-performance, accessibility, documentation, and rollback evidence all exist.
-
-## 2026-08-11 parent sections preserved during #4282 reconciliation
-
-## 2026-08-11 Workspace child receives hosted MyPy repair
-
-PR `#4279` retains branch `feat/4218-toolstrip-workspace` and base
-`feat/4181-launch-monitor-registry`. Exact child
-`7806a16f58e1c6999d32f0127a187fbb21f839a1` normally merges exact published
-parent `3796b49e40b677fbac4e05739f8be49f905df2cb`. No feature-code conflict
-exists: the inherited production delta is four static `numpy.ndarray` casts,
-while workspace/toolstrip behavior and runtime arrays, physics, frames, units,
-imports, and UI behavior remain unchanged. This local merge does not satisfy
-fresh protected CI, required review, unresolved-thread, dependency, or release
-gates.
-
-## 2026-08-11 workspace child receives latest repaired parent
-
-PR `#4279` retains branch `feat/4218-toolstrip-workspace` and base
-`feat/4181-launch-monitor-registry`. A normal merge incorporates exact parent
-`0216a547aa79727091a2939b96e779e8ddbd7304` into exact child
-`61b7f48b5aeb7d57246b4963da3df086e79cbe15`. There are no feature-code
-conflicts: workspace/toolstrip, visibility, navigation, playback, and plot
-behavior remain unchanged while the parent kinetics façade/dynamics/series
-split and pinned formatting are inherited. This local merge does not satisfy
-fresh protected CI, required review, unresolved-thread, dependency, or release
-gates.
-
-## 2026-08-11 Append-only SPEC preservation repair
-
-Independent review found that the local #4203 reconciliation omitted four
-exact D-plane parent SPEC rows dated 2026-08-10 (versions 1.13.11, 1.13.9,
-1.13.7, and 1.13.6). They are restored verbatim in a documentation-only
-follow-up. Production behavior, tests, ordered parents, base, and local quality
-evidence are unchanged; re-review, normal publication, and protected CI remain
-open.
-
-## 2026-08-11 Current D-plane parent reconciliation
-
-PR #4203's published child `217e36dc93d30f79826847f958fbcd10805e58ed`
-is being normally merged with exact current D-plane base
-`f3363aa88868f6a5c7e9ccfc682a9eca014e86c1`. The base and source branch stay
-unchanged. The only conflict is the parent's formatting of an older monolithic
-kinetics expression: the split facade keeps its typed pendulum accessor and
-inherits the parent's geometry explanation. No launch-monitor, D-plane, or
-kinetics behavior is intentionally changed. Four Qt-stub `Any` returns exposed
-by the exact MyPy 1.13 delta are also narrowed to their declared `bool`/`str`
-contracts without changing values. Focused/full validation, independent
-review, normal publication, and protected CI remain open.
-
-## 2026-08-11 Exact-head format completion
-
-Protected CI on #4203 head `7d69a545ae555679f0318940e67c1786626d6794`
-failed only Ruff formatting. The pinned 0.14.10 reproduction found eleven
-noncompliant changed Python files: four inherited files plus seven altered by
-the automated pre-commit repair. The pending repair intentionally changes no
-launch-monitor, kinetics, simulation, or UI behavior and requires independent
-review plus fresh protected CI before integration.
-
-## 2026-08-11 Hosted MyPy NumPy-return repair
-
-CI Standard run `31477542889`, job `93734652129`, found four
-`no-any-return` errors after the kinetics façade extraction at exact PR #4203
-head `0216a547aa79727091a2939b96e779e8ddbd7304`. The NumPy stubs widen
-`linalg.norm`, `concatenate`, and matrix projection results to `Any` even
-though the runtime operations return arrays. The private series/dynamics
-modules and public façade now use explicit `cast(np.ndarray, ...)` boundaries.
-These casts are static only: array identity, values, dtype, shape, physics,
-frames, units, imports, and UI behavior are unchanged. Verification must cover
-the complete PR changed-source MyPy set and both focused and full Rate tests.
-Current local evidence is 102 changed-source MyPy files, 28 focused tests, and
-all 701 Rate tests passing; complete-delta Ruff/format and Bandit are also
-green.
-
-## 2026-08-11 Kinetics façade extraction
-
-The strict current-head delta reproduces a protected size failure that the
-PR-base comparison hides: `kinetics.py` is changed versus `HEAD~1` and is 646
-LOC, above the ungrandfathered 500-LOC maximum. The public module is now a
-222-LOC run adapter/facade; its pure double-pendulum dynamics are in the
-205-LOC `_kinetics_dynamics.py`, and the immutable `KineticsSeries`/DbC
-contract is in the 131-LOC `_kinetics_series.py`. Existing public imports and
-the `_reaction_forces` seam are the same implementation objects. A RED-first
-identity test plus the existing physics, energy, force, parity-fixture,
-presentation, and PyQt tests protect behavior. No physics, frame, SI-unit,
-schema, UI, or stack-order change is intended; the commit remains local only.
-Focused kinetics/presentation/PyQt verification is 28 passing tests; the
-complete Rate-of-Closure Python regression suite is 701 passing tests.
-
-## 2026-08-11 #4203 pinned-Ruff formatting repair
-
-No material handoff change: current-head CI Standard run `31468208320`, job
-`93705508050`, identified eight files that differ from repository-pinned Ruff
-0.14.10 output. This commit changes formatting only; Rate physics, behavior,
-frames, persistence schemas, PyQt/React contracts, stack bases, and dependency
-order remain unchanged. Fresh protected CI and review are still required.
-
-## 2026-08-10 Ground/Tee parity child receives repaired parent
-
-Ready PR `#4325` keeps branch `feat/4143-tee-parity-fixture` and base
-`feat/4181-launch-monitor-registry`. It normally merges exact repaired parent
-head `12dd76a8dbcc106c4683f2f2e53076f8dc6f1b76` without any production/test
-code conflict or history rewrite. Shared parity and rendered-evidence
-contracts remain intact; fresh protected CI, review, dependency order, and
-#4143 release remain open.
-
-## 2026-08-10 #4143 Python/React Golden Ball-Setup Parity
-
-The isolated `feat/4143-tee-parity-fixture` child begins at exact draft PR
-#4203 head `31cbc007d4c85b5479b7cd0fb0969124eab2af67`. One versioned JSON fixture now
-drives both Python and React ball-support tests with explicit SI units and the
-ground-plane-to-ball-bottom height reference. The shared cases cover Driver
-and non-Driver defaults, user overrides, Ground zero effective height,
-derived ball-center geometry and serialization, negative/NaN/infinite height
-rejection, and backward-compatible migration of a legacy run without
-`ball_setup`.
-
-Evidence is 18 passing Python tee/parity tests and 24 passing React
-tee/persistence/parity tests, plus green TypeScript, ESLint, Vite production
-build, Ruff check, and Ruff format. Production model and UI code are unchanged.
-Recorded visual evidence now adds 1600 x 1200 Playwright Driver/Tee and
-rerun-Ground captures plus 1400 x 900 hidden-window PyQt captures. The browser
-run asserts control state, diagram geometry, and zero console/page errors; the
-desktop run asserts canonical center/artist state, nonblank output, and
-different Ground/Tee digests without pixel absolutes. The exact artifacts and
-versioned manifests are in
-`C:\Users\diete\AppData\Local\Temp\rate-4143-visual-evidence-8050eeba`.
-Do not close #4143: protected current-head CI/review and release to `main`
-remain. This exact parent predates the strict campaign manifest on a divergent
-branch, so no downstream manifest was copied into this bounded child.
-
-## 2026-08-10 Wind scalar child receives second repaired parent
-
-PR `#4281` retains branch `feat/4199-wind-scalar-adapter` and base
-`feat/4144-variation-export-continuation`. It normally merges exact repaired
-parent `b90e5021a59e2081415b51ef29fbed06377bc201` without scalar-adapter code
-conflict or history rewrite. Fresh child CI, review, and dependency order
-remain required.
-Post-reconciliation evidence is 25 focused D-plane/impact tests plus docs,
-changed-file-size, and whitespace gates.
-
-## 2026-08-10 Variation export child receives exact workspace parent
-
-PR `#4280` retains branch `feat/4144-variation-export-continuation`, base
-`feat/4218-toolstrip-workspace`, and original child first parent
-`f90836e342efc8be624739802375af2876d11e5f`. Exact parent
-`6717e9e09d507dbc24bedb36177f1cdf0b4fd90b` merges normally as the second
-parent. All variation visualization/export source merged without conflict:
-selected scatter CSV, typed unavailable rows, accessible bounded raw tables,
-linked trials, focused PyQt widgets, and React parity remain additive with the
-parent workspace/toolstrip/playback/plot/navigation repairs. SPEC 1.14.12 is
-the unique combined child entry. Staged review exposed a stale linked-trial
-selection when a smaller rerun replaced the active result. Corrected PyQt
-views clear result-local selection before repopulating and validate all public
-setter indices; React clears on result identity change and shares only bounded
-selections with linked views. Thirty-seven focused variation tests and the
-complete Rate/shared-swing/golf-club matrix pass 1,528 tests with two explicit
-optional build123d skips; all 546 React tests across 90 files pass;
-TypeScript, ESLint, and production build are green. Exact-parent
-Ruff/Ruff-format/Black, pinned MyPy 1.13 across four Python production modules,
-Bandit, file-size, docs, minimum-test, assertion, detect-secrets, diff, real
-CPython 3.10.20 compilation, and 30 compatibility regressions pass. There is
-no Rust delta from the exact parent. Independent staged re-review found no
-actionable findings, including the first replacement-render React boundary.
-Protected CI/review remain release gates after an ordinary guarded push.
-
-## 2026-08-10 Variation child receives second repaired workspace parent
-
-PR `#4280` retains branch `feat/4144-variation-export-continuation` and base
-`feat/4218-toolstrip-workspace`. It normally merges exact repaired parent head
-`61b7f48b5aeb7d57246b4963da3df086e79cbe15` without feature-code conflict or
-history rewrite. Variation/export behavior remains unchanged; fresh child CI,
-review, and dependency order remain required.
-Post-reconciliation evidence is 25 focused D-plane/impact tests plus docs,
-changed-file-size, and whitespace gates.
-
-## 2026-08-10 Wind scalar child receives exact variation parent
-
-PR `#4281` keeps branch `feat/4199-wind-scalar-adapter`, base
-`feat/4144-variation-export-continuation`, and original child first parent
-`cf52529b1e68479321bb93b1be3d59c77f782008`. Exact parent
-`8bcc49fc4e16e5e43be0b7f0f03c3017d5b79d0c` merges normally second. The
-child's scalar-ensemble wire contract, provenance-preserving wind rows,
-availability accounting, and plot adapter remain additive with the parent's
-variation visualization/export and result-local selection repairs. Python and
-TypeScript reject wind scenarios unless their per-trial provenance is exact,
-shear/turbulence are zero, seed is zero, and gusts are empty. SPEC 1.14.13 is
-the unique combined child entry. Twenty-seven focused Python tests, the
-1,549-test complete Rate/shared-swing/golf-club matrix with two explicit
-optional build123d skips, and all 556 React tests across 92 files pass.
-TypeScript, ESLint, production build, Ruff/format, pinned MyPy, Bandit, size,
-docs, minimum-test, assertions, detect-secrets, diff, and CPython 3.10 compile
-gates pass. Independent re-review found no actionable findings after checking
-the exact provenance derivation and all five deterministic-scenario
-regressions; protected CI/review remain release gates.
-
-## 2026-08-10 Variation child timestamp propagation
-
-The #4280 continuation now carries exact parent
-`05383d333b6fd87eaf5e37305476f50b505c2c2e` through a normal merge while
-preserving its original base and complete variation export/accessibility
-implementation. The inherit…4250 tokens truncated…7417e02926c3fbc7476002e82b690`; propagate it through the existing
-stack without changing bases.
-
-A follow-up scan found the torque-profile controller importing
-`datetime.UTC` directly. That parent-owned Python 3.11 boundary now uses
-`shared.python.compatibility.UTC`; persisted timestamps and torque-profile
-behavior are unchanged. Re-run the focused torque-profile UI tests and the
-real Python 3.10 compatibility probe before propagation.
-
-## Do-Not List
-
-- Do not duplicate the Kelvin-Voigt contact-force law — #4130 requires
-  `SpringDamperImpactModel` and the new `impact_interval/` package to share
-  one implementation (DRY, explicit in the epic's binding standards).
-- Do not exceed 500 LOC per file in `rate_of_closure`, `swing_sim`, or
-  `swing-core` — sub-package instead.
-- Do not hand-mirror physics into the TS web layer once WASM lands (Phase 7)
-  — that's the whole point of the wasm-pack build; today's hand-written TS
-  mirrors are a stopgap, not the target architecture.
-- Do not merge the stacked PRs out of order (see stack section above).
-- Do not invent citations in derivation docstrings or the AffineDrift
-  putting research content (H4) — sourced/verifiable only, dossier
-  discipline per epic #4125.
-
-## Roadmap (ordered)
-
-1. Merge #4119, then #4124, then #4129 in order.
-2. Start epic #4130 Phase F1 (formulation document) — six-DOF rigid-club
-   contact-interval derivation, validation program design.
-3. Phase 7 of #4103: WASM swap + real Pages CI deploy workflow.
-4. #4125 H4/H5: AffineDrift putting research content and the public
-   release-management repo (both cross-repo, not started).
diff --git a/src/rate_of_closure/application/__init__.py b/src/rate_of_closure/application/__init__.py
index 20fd60d88..cf3ca435d 100644
--- a/src/rate_of_closure/application/__init__.py
+++ b/src/rate_of_closure/application/__init__.py
@@ -6,6 +6,10 @@
     CommandAvailability,
     CommandUnavailableError,
 )
+from .regional_surface_plan_files import (
+    read_regional_surface_plan_request,
+    write_regional_surface_plan_request_atomic,
+)
 from .workspace_document import (
     WORKSPACE_SCHEMA,
     WORKSPACE_SCHEMA_VERSION,
@@ -30,7 +34,9 @@
     "WorkspaceLayout",
     "WorkspaceMetadata",
     "read_workspace",
+    "read_regional_surface_plan_request",
     "workspace_from_json",
     "workspace_to_json",
     "write_workspace_atomic",
+    "write_regional_surface_plan_request_atomic",
 ]
diff --git a/src/rate_of_closure/application/atomic_text_files.py b/src/rate_of_closure/application/atomic_text_files.py
new file mode 100644
index 000000000..852ce12b8
--- /dev/null
+++ b/src/rate_of_closure/application/atomic_text_files.py
@@ -0,0 +1,55 @@
+"""Small atomic UTF-8 text-file boundary shared by application documents."""
+
+from __future__ import annotations
+
+import os
+import tempfile
+from pathlib import Path
+
+
+def _destination_path(destination: str | Path, document_name: str) -> Path:
+    path = Path(destination)
+    if not path.name:
+        raise ValueError(f"{document_name} destination must name a file")
+    if not path.parent.is_dir():
+        raise FileNotFoundError(
+            f"{document_name} parent directory does not exist: {path.parent}"
+        )
+    if path.exists() and not path.is_file():
+        raise IsADirectoryError(f"{document_name} destination is not a file: {path}")
+    return path
+
+
+def write_utf8_text_atomic(
+    text: str,
+    destination: str | Path | None,
+    *,
+    document_name: str,
+) -> bool:
+    """Flush and atomically replace one UTF-8 text file.
+
+    ``None`` represents a cancelled chooser. Serialization must happen before
+    this function is called so a validation failure cannot touch the target.
+    """
+    if destination is None:
+        return False
+    if not isinstance(text, str):
+        raise TypeError("atomic file content must be text")
+    path = _destination_path(destination, document_name)
+    descriptor, temporary_name = tempfile.mkstemp(
+        prefix=f".{path.name}.", suffix=".tmp", dir=path.parent
+    )
+    temporary = Path(temporary_name)
+    try:
+        with os.fdopen(descriptor, "w", encoding="utf-8", newline="") as handle:
+            handle.write(text)
+            handle.flush()
+            os.fsync(handle.fileno())
+        os.replace(temporary, path)
+    except Exception:
+        temporary.unlink(missing_ok=True)
+        raise
+    return True
+
+
+__all__ = ["write_utf8_text_atomic"]
diff --git a/src/rate_of_closure/application/regional_surface_plan.py b/src/rate_of_closure/application/regional_surface_plan.py
index 34d5ecb89..1ffce376f 100644
--- a/src/rate_of_closure/application/regional_surface_plan.py
+++ b/src/rate_of_closure/application/regional_surface_plan.py
@@ -6,11 +6,13 @@
 from dataclasses import asdict, dataclass
 
 from shared.python.swing_sim.canonical_numeric_json import canonical_numeric_json
+from shared.python.swing_sim.ground.contract_types import GroundSurfaceProfile
 from shared.python.swing_sim.ground.regional_plan_records import (
     REGIONAL_PLAN_GEOMETRY_MODEL,
     REGIONAL_PLAN_LIMITATIONS,
     REGIONAL_PLAN_REQUEST_SCHEMA_VERSION,
     GroundRegionalMaterialPlanRequest,
+    GroundRegionalMaterialRegion,
 )
 from shared.python.swing_sim.ground.regional_plan_wire import (
     regional_material_plan_request_from_dict,
@@ -158,6 +160,99 @@ def _region_payload(region: RegionalOverlayDraft) -> dict[str, object]:
     }
 
 
+def _surface_draft(surface: GroundSurfaceProfile) -> SurfaceMaterialDraft:
+    """Project one already-validated wire surface into editable SI values."""
+    return SurfaceMaterialDraft(
+        surface.surface_id,
+        surface.normal_restitution,
+        surface.static_friction,
+        surface.kinetic_friction,
+        surface.rolling_resistance,
+        surface.firmness_pa,
+        surface.hardness_fraction,
+        surface.grass_height_m,
+        surface.compressibility_fraction,
+        surface.compression_damping_fraction,
+        surface.turf_density_kg_m3,
+        surface.moisture_fraction,
+    )
+
+
+def _assert_editor_surface(surface: GroundSurfaceProfile) -> None:
+    """Reject material evidence not authored under the editor v1 contract."""
+    if (
+        surface.provider_id != EDITOR_PROVIDER_ID
+        or surface.provider_version != EDITOR_PROVIDER_VERSION
+    ):
+        raise ValueError("surface is not qualified by the editor provider v1")
+
+
+def _overlay_draft(region: GroundRegionalMaterialRegion) -> RegionalOverlayDraft:
+    return RegionalOverlayDraft(
+        region.region_id,
+        region.precedence,
+        region.lower_coordinate_m,
+        region.upper_coordinate_m,
+        _surface_draft(region.surface),
+    )
+
+
+def editor_draft_from_regional_surface_plan_request(
+    request: GroundRegionalMaterialPlanRequest,
+) -> RegionalSurfacePlanDraft:
+    """Project one fully validated, editor-qualified v1 request into the editor.
+
+    The editor provider v1 contract is explicitly unvalidated. Requests from a
+    different producer, material provider, axis qualification, or row capacity
+    are therefore rejected rather than relabelled or coerced.
+    """
+    if type(request) is not GroundRegionalMaterialPlanRequest:
+        raise TypeError("request must be an exact GroundRegionalMaterialPlanRequest")
+    provenance = request.provenance
+    if (
+        provenance.producer != EDITOR_PROVIDER_ID
+        or provenance.producer_version != EDITOR_PROVIDER_VERSION
+    ):
+        raise ValueError("request is not qualified by the editor producer v1")
+    if request.axis_origin_m != (0.0, 0.0, 0.0) or request.axis_unit != (
+        1.0,
+        0.0,
+        0.0,
+    ):
+        raise ValueError("request uses an unsupported editor axis qualification")
+    if len(request.regions) > MAX_EDITOR_REGIONS:
+        raise ValueError(f"editor supports one to at most {MAX_EDITOR_REGIONS} regions")
+    _assert_editor_surface(request.base_surface)
+    for region in request.regions:
+        _assert_editor_surface(region.surface)
+    draft = RegionalSurfacePlanDraft(
+        request.request_id,
+        request.lower_coordinate_m,
+        request.upper_coordinate_m,
+        provenance.source_revision,
+        "unvalidated",
+        _surface_draft(request.base_surface),
+        tuple(_overlay_draft(region) for region in request.regions),
+    )
+    if provenance.input_sha256 != _source_digest(draft):
+        raise ValueError("editor provenance digest does not match the editable request")
+    return draft
+
+
+def regional_surface_plan_request_for_draft(
+    draft: RegionalSurfacePlanDraft,
+    imported_request: GroundRegionalMaterialPlanRequest | None = None,
+) -> GroundRegionalMaterialPlanRequest:
+    """Preserve an unchanged import exactly; otherwise bind fresh provenance."""
+    if imported_request is not None:
+        imported_draft = editor_draft_from_regional_surface_plan_request(
+            imported_request
+        )
+        if draft == imported_draft:
+            return imported_request
+    return validate_regional_surface_plan_draft(draft)
+
+
 def validate_regional_surface_plan_draft(
     draft: RegionalSurfacePlanDraft,
 ) -> GroundRegionalMaterialPlanRequest:
@@ -191,10 +286,14 @@ def validate_regional_surface_plan_draft(
 
 
 __all__ = [
+    "EDITOR_PROVIDER_ID",
+    "EDITOR_PROVIDER_VERSION",
     "MAX_EDITOR_REGIONS",
     "RegionalOverlayDraft",
     "RegionalSurfacePlanDraft",
     "SurfaceMaterialDraft",
+    "editor_draft_from_regional_surface_plan_request",
     "illustrative_regional_surface_plan_draft",
+    "regional_surface_plan_request_for_draft",
     "validate_regional_surface_plan_draft",
 ]
diff --git a/src/rate_of_closure/application/regional_surface_plan_files.py b/src/rate_of_closure/application/regional_surface_plan_files.py
new file mode 100644
index 000000000..14ff29643
--- /dev/null
+++ b/src/rate_of_closure/application/regional_surface_plan_files.py
@@ -0,0 +1,57 @@
+"""Strict native file adapter for canonical regional surface-plan requests."""
+
+from __future__ import annotations
+
+from pathlib import Path
+
+from shared.python.swing_sim.ground.regional_plan_records import (
+    MAX_REGIONAL_PLAN_WIRE_BYTES,
+    GroundRegionalMaterialPlanRequest,
+)
+from shared.python.swing_sim.ground.regional_plan_wire import (
+    regional_material_plan_request_from_json,
+)
+
+from .atomic_text_files import write_utf8_text_atomic
+
+
+def _read_bounded_utf8(path: Path) -> str:
+    """Read one immutable handle snapshot without allocating beyond the cap."""
+    with path.open("rb") as handle:
+        raw = handle.read(MAX_REGIONAL_PLAN_WIRE_BYTES + 1)
+    if len(raw) > MAX_REGIONAL_PLAN_WIRE_BYTES:
+        raise ValueError("regional material plan exceeds maximum wire size")
+    try:
+        return raw.decode("utf-8")
+    except UnicodeDecodeError as exc:
+        raise ValueError("regional material plan must be valid UTF-8") from exc
+
+
+def read_regional_surface_plan_request(
+    source: str | Path,
+) -> GroundRegionalMaterialPlanRequest:
+    """Read and completely validate one canonical request before returning it."""
+    path = Path(source)
+    if not path.is_file():
+        raise FileNotFoundError(f"regional surface plan does not exist: {path}")
+    return regional_material_plan_request_from_json(_read_bounded_utf8(path))
+
+
+def write_regional_surface_plan_request_atomic(
+    request: GroundRegionalMaterialPlanRequest,
+    destination: str | Path | None,
+) -> bool:
+    """Write exact canonical request bytes, or return false after cancellation."""
+    if destination is None:
+        return False
+    if type(request) is not GroundRegionalMaterialPlanRequest:
+        raise TypeError("request must be an exact GroundRegionalMaterialPlanRequest")
+    return write_utf8_text_atomic(
+        request.to_json(), destination, document_name="regional surface plan"
+    )
+
+
+__all__ = [
+    "read_regional_surface_plan_request",
+    "write_regional_surface_plan_request_atomic",
+]
diff --git a/src/rate_of_closure/application/workspace_files.py b/src/rate_of_closure/application/workspace_files.py
index 58b6f2ebb..fa627902d 100644
--- a/src/rate_of_closure/application/workspace_files.py
+++ b/src/rate_of_closure/application/workspace_files.py
@@ -2,10 +2,9 @@
 
 from __future__ import annotations
 
-import os
-import tempfile
 from pathlib import Path
 
+from .atomic_text_files import write_utf8_text_atomic
 from .workspace_document import (
     WorkspaceDocument,
     workspace_from_json,
@@ -13,18 +12,6 @@
 )
 
 
-def _destination_path(destination: str | Path) -> Path:
-    path = Path(destination)
-    if not path.name:
-        raise ValueError("workspace destination must name a file")
-    parent = path.parent
-    if not parent.is_dir():
-        raise FileNotFoundError(f"workspace parent directory does not exist: {parent}")
-    if path.exists() and not path.is_file():
-        raise IsADirectoryError(f"workspace destination is not a file: {path}")
-    return path
-
-
 def read_workspace(source: str | Path) -> WorkspaceDocument:
     """Read and fully validate one workspace before returning it.
 
@@ -59,21 +46,7 @@ def write_workspace_atomic(
     if destination is None:
         return False
     serialized = workspace_to_json(document)
-    path = _destination_path(destination)
-    descriptor, temporary_name = tempfile.mkstemp(
-        prefix=f".{path.name}.", suffix=".tmp", dir=path.parent
-    )
-    temporary = Path(temporary_name)
-    try:
-        with os.fdopen(descriptor, "w", encoding="utf-8", newline="\n") as handle:
-            handle.write(serialized)
-            handle.flush()
-            os.fsync(handle.fileno())
-        os.replace(temporary, path)
-    except Exception:
-        temporary.unlink(missing_ok=True)
-        raise
-    return True
+    return write_utf8_text_atomic(serialized, destination, document_name="workspace")
 
 
 __all__ = ["read_workspace", "write_workspace_atomic"]
diff --git a/src/rate_of_closure/ui/pyqt6/engineering_number_input.py b/src/rate_of_closure/ui/pyqt6/engineering_number_input.py
index ff0ed6d15..af130c734 100644
--- a/src/rate_of_closure/ui/pyqt6/engineering_number_input.py
+++ b/src/rate_of_closure/ui/pyqt6/engineering_number_input.py
@@ -16,6 +16,7 @@ class NumberInputSpec:
     step: float = 0.1
     minimum: float = -1e9
     maximum: float = 1e9
+    decimals: int = 6
 
     def __post_init__(self) -> None:
         """Reject unusable Qt ranges before a widget is constructed."""
@@ -25,6 +26,12 @@ def __post_init__(self) -> None:
             raise ValueError("minimum and maximum must be finite")
         if self.minimum > self.maximum:
             raise ValueError("minimum must not exceed maximum")
+        if (
+            isinstance(self.decimals, bool)
+            or not isinstance(self.decimals, int)
+            or not 0 <= self.decimals <= 323
+        ):
+            raise ValueError("decimals must be an integer from zero through 323")
 
 
 def engineering_number_input(
@@ -35,7 +42,7 @@ def engineering_number_input(
     """Create one consistently configured accessible SI number input."""
     field = QDoubleSpinBox()
     field.setAccessibleName(name)
-    field.setDecimals(6)
+    field.setDecimals(spec.decimals)
     field.setRange(spec.minimum, spec.maximum)
     field.setSingleStep(spec.step)
     field.setSuffix(spec.suffix)
diff --git a/src/rate_of_closure/ui/pyqt6/regional_surface_plan_io.py b/src/rate_of_closure/ui/pyqt6/regional_surface_plan_io.py
new file mode 100644
index 000000000..db39e5183
--- /dev/null
+++ b/src/rate_of_closure/ui/pyqt6/regional_surface_plan_io.py
@@ -0,0 +1,101 @@
+"""Native Open/Save As controller for the regional surface-plan editor."""
+
+from __future__ import annotations
+
+from pathlib import Path
+from typing import Protocol
+
+from PyQt6.QtWidgets import QFileDialog, QLabel, QPlainTextEdit, QWidget
+
+from rate_of_closure.application.regional_surface_plan_files import (
+    read_regional_surface_plan_request,
+    write_regional_surface_plan_request_atomic,
+)
+from shared.python.swing_sim.ground.regional_plan_records import (
+    GroundRegionalMaterialPlanRequest,
+)
+
+
+class RegionalSurfacePlanFileHost(Protocol):
+    """Minimal host surface needed by the native file controller."""
+
+    status_label: QLabel
+    readback: QPlainTextEdit
+
+    def current_request(self) -> GroundRegionalMaterialPlanRequest: ...
+
+    def apply_imported_request(
+        self, request: GroundRegionalMaterialPlanRequest
+    ) -> None: ...
+
+
+class RegionalSurfacePlanFileActions:
+    """Transactional native file commands with one successful recent path."""
+
+    def __init__(self, host: RegionalSurfacePlanFileHost, parent: QWidget) -> None:
+        self._host = host
+        self._parent = parent
+        self.recent_path: Path | None = None
+
+    def open(self) -> None:
+        """Choose, fully validate, and only then replace the editor draft."""
+        selected, _filter = QFileDialog.getOpenFileName(
+            self._parent,
+            "Open Regional Surface Plan",
+            self._initial_location(),
+            "JSON files (*.json)",
+        )
+        if not selected:
+            return
+        path = Path(selected)
+        try:
+            request = read_regional_surface_plan_request(path)
+            self._host.apply_imported_request(request)
+        except (OSError, TypeError, ValueError) as exc:
+            self._set_error(f"Open failed: {exc}")
+            return
+        self.recent_path = path
+        self._host.readback.setPlainText(request.to_json())
+        self._set_success(f"Opened {path.name}. No physics executed.")
+
+    def save_as(self) -> None:
+        """Validate before choosing a destination, then atomically replace it."""
+        try:
+            request = self._host.current_request()
+        except (TypeError, ValueError) as exc:
+            self._set_error(f"Save failed: {exc}")
+            return
+        selected, _filter = QFileDialog.getSaveFileName(
+            self._parent,
+            "Save Regional Surface Plan As",
+            self._initial_location("regional-surface-plan.json"),
+            "JSON files (*.json)",
+        )
+        if not selected:
+            return
+        path = Path(selected)
+        try:
+            write_regional_surface_plan_request_atomic(request, path)
+        except OSError as exc:
+            self._set_error(f"Save failed: {exc}")
+            return
+        self.recent_path = path
+        self._host.readback.setPlainText(request.to_json())
+        self._set_success(f"Saved {path.name} atomically. No physics executed.")
+
+    def _initial_location(self, filename: str = "") -> str:
+        if self.recent_path is None:
+            return filename
+        parent = self.recent_path.parent
+        return str(parent / filename) if filename else str(parent)
+
+    def _set_error(self, message: str) -> None:
+        self._host.status_label.setText(message)
+        self._host.status_label.setAccessibleName("Regional surface plan file error")
+
+    def _set_success(self, message: str) -> None:
+        self._host.status_label.setText(message)
+        self._host.status_label.setAccessibleName("Regional surface plan file success")
+
+
+__all__ = ["QFileDialog", "RegionalSurfacePlanFileActions"]
diff --git a/src/rate_of_closure/ui/pyqt6/regional_surface_plan_tab.py b/src/rate_of_closure/ui/pyqt6/regional_surface_plan_tab.py
index 26d8e7406..bcdc38ef3 100644
--- a/src/rate_of_closure/ui/pyqt6/regional_surface_plan_tab.py
+++ b/src/rate_of_closure/ui/pyqt6/regional_surface_plan_tab.py
@@ -2,13 +2,10 @@
 
 from __future__ import annotations
 
-from collections.abc import Callable
 from dataclasses import replace
 
-from PyQt6.QtCore import Qt
 from PyQt6.QtWidgets import (
     QComboBox,
-    QDoubleSpinBox,
     QFormLayout,
     QGroupBox,
     QHBoxLayout,
@@ -17,7 +14,6 @@
     QPlainTextEdit,
     QPushButton,
     QScrollArea,
-    QSpinBox,
     QVBoxLayout,
     QWidget,
 )
@@ -26,124 +22,23 @@
     MAX_EDITOR_REGIONS,
     RegionalOverlayDraft,
     RegionalSurfacePlanDraft,
-    SurfaceMaterialDraft,
+    editor_draft_from_regional_surface_plan_request,
     illustrative_regional_surface_plan_draft,
-    validate_regional_surface_plan_draft,
+    regional_surface_plan_request_for_draft,
 )
-from rate_of_closure.ui.pyqt6.engineering_number_input import (
-    NumberInputSpec,
+from rate_of_closure.ui.pyqt6.regional_surface_plan_io import (
+    RegionalSurfacePlanFileActions,
 )
-from rate_of_closure.ui.pyqt6.engineering_number_input import (
-    engineering_number_input as _number_input,
+from rate_of_closure.ui.pyqt6.regional_surface_plan_widgets import (
+    MaterialEditor,
+    RegionalOverlayRow,
+    coordinate_input,
 )
-
-_MATERIAL_FIELDS = (
-    ("normal_restitution", "Normal restitution", "", 0.01, 0.0, 1.0),
-    ("static_friction", "Static friction", "", 0.01, 0.0, 5.0),
-    ("kinetic_friction", "Kinetic friction", "", 0.01, 0.0, 5.0),
-    ("rolling_resistance", "Rolling resistance", "", 0.01, 0.0, 1.0),
-    ("firmness_pa", "Firmness", " Pa", 1_000.0, 0.001, 1e9),
-    ("hardness_fraction", "Hardness", " fraction", 0.01, 0.0, 1.0),
-    ("grass_height_m", "Grass height", " m", 0.001, 0.0, 1.0),
-    ("compressibility_fraction", "Compressibility", " fraction", 0.01, 0.0, 1.0),
-    (
-        "compression_damping_fraction",
-        "Compression damping",
-        " fraction",
-        0.01,
-        0.0,
-        1.0,
-    ),
-    ("turf_density_kg_m3", "Turf density", " kg/m³", 1.0, 0.0, 10_000.0),
-    ("moisture_fraction", "Moisture", " fraction", 0.01, 0.0, 1.0),
+from shared.python.swing_sim.ground.regional_plan_records import (
+    GroundRegionalMaterialPlanRequest,
 )
 
 
-class MaterialEditor(QGroupBox):
-    """Editable surface identity and full v1 material parameter collection."""
-
-    def __init__(self, title: str, value: SurfaceMaterialDraft) -> None:
-        super().__init__(title)
-        self.surface_id = QLineEdit(value.surface_id)
-        self.surface_id.setAccessibleName(f"{title} surface ID")
-        self.surface_id.setToolTip(
-            f"Stable identifier for {title.lower()}; included in validated readback."
-        )
-        self.fields: dict[str, QDoubleSpinBox] = {}
-        layout = QFormLayout(self)
-        layout.addRow("Surface ID", self.surface_id)
-        for name, label, suffix, step, minimum, maximum in _MATERIAL_FIELDS:
-            field = _number_input(
-                f"{title} {label}",
-                getattr(value, name),
-                NumberInputSpec(suffix, step, minimum, maximum),
-            )
-            self.fields[name] = field
-            layout.addRow(label, field)
-
-    def draft(self) -> SurfaceMaterialDraft:
-        """Read the current widgets without applying separate UI validation."""
-        values = {name: field.value() for name, field in self.fields.items()}
-        return SurfaceMaterialDraft(self.surface_id.text(), **values)
-
-
-class RegionalOverlayRow(QGroupBox):
-    """One removable bounded regional overlay row."""
-
-    def __init__(
-        self,
-        ordinal: int,
-        value: RegionalOverlayDraft,
-        remove: Callable[[RegionalOverlayRow], None],
-    ) -> None:
-        super().__init__(f"Regional overlay {ordinal}")
-        self.region_id = QLineEdit(value.region_id)
-        self.region_id.setToolTip(
-            "Stable overlay identifier; it must be unique within the regional plan."
-        )
-        self.precedence = QSpinBox()
-        self.precedence.setRange(0, 1_000_000)
-        self.precedence.setValue(value.precedence)
-        self.precedence.setToolTip(
-            "Overlay selection precedence. Higher values win when intervals overlap."
-        )
-        self.lower_coordinate = _number_input(
-            f"Overlay {ordinal} lower coordinate",
-            value.lower_coordinate_m,
-            NumberInputSpec(suffix=" m"),
-        )
-        self.upper_coordinate = _number_input(
-            f"Overlay {ordinal} upper coordinate",
-            value.upper_coordinate_m,
-            NumberInputSpec(suffix=" m"),
-        )
-        self.material = MaterialEditor(f"Overlay {ordinal} material", value.surface)
-        self.remove_button = QPushButton(f"Remove overlay {ordinal}")
-        self.remove_button.setToolTip(
-            "Remove this overlay from the unvalidated draft; one overlay is required."
-        )
-        self.remove_button.clicked.connect(lambda: remove(self))
-        form = QFormLayout()
-        form.addRow("Region ID", self.region_id)
-        form.addRow("Precedence", self.precedence)
-        form.addRow("Lower coordinate", self.lower_coordinate)
-        form.addRow("Upper coordinate", self.upper_coordinate)
-        layout = QVBoxLayout(self)
-        layout.addLayout(form)
-        layout.addWidget(self.material)
-        layout.addWidget(self.remove_button, alignment=Qt.AlignmentFlag.AlignRight)
-
-    def draft(self) -> RegionalOverlayDraft:
-        """Read one overlay for authoritative contract validation."""
-        return RegionalOverlayDraft(
-            self.region_id.text(),
-            self.precedence.value(),
-            self.lower_coordinate.value(),
-            self.upper_coordinate.value(),
-            self.material.draft(),
-        )
-
-
 class RegionalSurfacePlanTab(QWidget):
     """Session-only regional surface editor with strict canonical readback."""
 
@@ -151,6 +46,8 @@ def __init__(self, parent: QWidget | None = None) -> None:
         super().__init__(parent)
         self._rows: list[RegionalOverlayRow] = []
         self._initial = illustrative_regional_surface_plan_draft()
+        self._imported_request: GroundRegionalMaterialPlanRequest | None = None
+        self.file_actions = RegionalSurfacePlanFileActions(self, self)
         self._build_ui()
         self._connect_static_changes()
         self._append_row(self._initial.regions[0])
@@ -173,7 +70,7 @@ def _connect_static_changes(self) -> None:
     def _connect_row_changes(self, row: RegionalOverlayRow) -> None:
         """Connect one dynamic overlay to the shared invalidation boundary."""
         row.region_id.textChanged.connect(self._mark_dirty)
-        row.precedence.valueChanged.connect(self._mark_dirty)
+        row.precedence.textChanged.connect(self._mark_dirty)
         row.lower_coordinate.valueChanged.connect(self._mark_dirty)
         row.upper_coordinate.valueChanged.connect(self._mark_dirty)
         self._connect_material_changes(row.material)
@@ -242,8 +139,8 @@ def _identity_box(self) -> QGroupBox:
         )
         geometry.setWordWrap(True)
         session = QLabel(
-            "Session-only draft: current workspace file commands do not persist "
-            "model inputs."
+            "Open/Save As persists this canonical request only. Workspace "
+            "persistence remains a separate contract."
         )
         session.setWordWrap(True)
         layout = QFormLayout(box)
@@ -257,15 +154,13 @@ def _identity_box(self) -> QGroupBox:
     def _base_box(self) -> QGroupBox:
         """Create the base interval and complete base material editor."""
         box = QGroupBox("Base surface and domain")
-        self.domain_lower = _number_input(
+        self.domain_lower = coordinate_input(
             "Base domain lower coordinate",
             self._initial.lower_coordinate_m,
-            NumberInputSpec(suffix=" m"),
         )
-        self.domain_upper = _number_input(
+        self.domain_upper = coordinate_input(
             "Base domain upper coordinate",
             self._initial.upper_coordinate_m,
-            NumberInputSpec(suffix=" m"),
         )
         self.base_material = MaterialEditor("Base material", self._initial.base_surface)
         layout = QVBoxLayout(box)
@@ -291,7 +186,21 @@ def _action_row(self) -> QHBoxLayout:
             "Validate the complete draft and display its canonical SI request."
         )
         self.validate_button.clicked.connect(self.validate_plan)
+        self.open_button = QPushButton("Open JSON")
+        self.open_button.setAccessibleName("Open regional surface plan JSON")
+        self.open_button.setToolTip(
+            "Open and fully validate an editor-qualified canonical request."
+        )
+        self.open_button.clicked.connect(self.file_actions.open)
+        self.save_button = QPushButton("Save As JSON")
+        self.save_button.setAccessibleName("Save regional surface plan JSON as")
+        self.save_button.setToolTip(
+            "Atomically save the validated canonical request to a chosen file."
+        )
+        self.save_button.clicked.connect(self.file_actions.save_as)
         layout.addWidget(self.add_button)
+        layout.addWidget(self.open_button)
+        layout.addWidget(self.save_button)
         layout.addStretch(1)
         layout.addWidget(self.validate_button)
         return layout
@@ -362,10 +271,35 @@ def draft(self) -> RegionalSurfacePlanDraft:
             tuple(row.draft() for row in self._rows),
         )
 
+    def current_request(self) -> GroundRegionalMaterialPlanRequest:
+        """Return exact imported evidence unless the visible draft has changed."""
+        return regional_surface_plan_request_for_draft(
+            self.draft(), self._imported_request
+        )
+
+    def apply_imported_request(
+        self, request: GroundRegionalMaterialPlanRequest
+    ) -> None:
+        """Populate widgets only after complete strict editor qualification."""
+        draft = editor_draft_from_regional_surface_plan_request(request)
+        self.request_id.setText(draft.request_id)
+        self.source_revision.setText(draft.source_revision)
+        self.domain_lower.setValue(draft.lower_coordinate_m)
+        self.domain_upper.setValue(draft.upper_coordinate_m)
+        self.base_material.set_draft(draft.base_surface)
+        for row in self._rows:
+            self.rows_layout.removeWidget(row)
+            row.deleteLater()
+        self._rows.clear()
+        for region in draft.regions:
+            self._append_row(region)
+        self._update_row_actions()
+        self._imported_request = request
+
     def validate_plan(self) -> None:
         """Validate through the strict contract and render canonical readback."""
         try:
-            request = validate_regional_surface_plan_draft(self.draft())
+            request = self.current_request()
         except (TypeError, ValueError) as exc:
             self.status_label.setText(str(exc))
             self.status_label.setAccessibleName(
diff --git a/src/rate_of_closure/ui/pyqt6/regional_surface_plan_widgets.py b/src/rate_of_closure/ui/pyqt6/regional_surface_plan_widgets.py
new file mode 100644
index 000000000..33cbb0aab
--- /dev/null
+++ b/src/rate_of_closure/ui/pyqt6/regional_surface_plan_widgets.py
@@ -0,0 +1,231 @@
+"""Reusable PyQt6 inputs for the regional surface-plan editor."""
+
+from __future__ import annotations
+
+from collections.abc import Callable
+
+from PyQt6.QtCore import Qt
+from PyQt6.QtGui import QValidator
+from PyQt6.QtWidgets import (
+    QDoubleSpinBox,
+    QFormLayout,
+    QGroupBox,
+    QLineEdit,
+    QPushButton,
+    QVBoxLayout,
+)
+
+from rate_of_closure.application.regional_surface_plan import (
+    RegionalOverlayDraft,
+    SurfaceMaterialDraft,
+)
+from rate_of_closure.ui.pyqt6.engineering_number_input import (
+    NumberInputSpec,
+    engineering_number_input,
+)
+from shared.python.swing_sim.canonical_numeric_json import (
+    MAX_CANONICAL_SAFE_INTEGER,
+)
+
+_MAX_SAFE_FLOAT = float(MAX_CANONICAL_SAFE_INTEGER)
+_MAX_SAFE_INTEGER_TEXT = str(MAX_CANONICAL_SAFE_INTEGER)
+
+_MATERIAL_FIELDS = (
+    ("normal_restitution", "Normal restitution", "", 0.01, 0.0, 1.0),
+    ("static_friction", "Static friction", "", 0.01, 0.0, 5.0),
+    ("kinetic_friction", "Kinetic friction", "", 0.01, 0.0, 5.0),
+    ("rolling_resistance", "Rolling resistance", "", 0.01, 0.0, 1.0),
+    ("firmness_pa", "Firmness", " Pa", 1_000.0, 1e-11, _MAX_SAFE_FLOAT),
+    ("hardness_fraction", "Hardness", " fraction", 0.01, 0.0, 1.0),
+    ("grass_height_m", "Grass height", " m", 0.001, 0.0, _MAX_SAFE_FLOAT),
+    ("compressibility_fraction", "Compressibility", " fraction", 0.01, 0.0, 1.0),
+    (
+        "compression_damping_fraction",
+        "Compression damping",
+        " fraction",
+        0.01,
+        0.0,
+        1.0,
+    ),
+    (
+        "turf_density_kg_m3",
+        "Turf density",
+        " kg/m³",
+        1.0,
+        0.0,
+        _MAX_SAFE_FLOAT,
+    ),
+    ("moisture_fraction", "Moisture", " fraction", 0.01, 0.0, 1.0),
+)
+
+
+def number_input(
+    name: str,
+    value: float,
+    spec: NumberInputSpec | None = None,
+) -> QDoubleSpinBox:
+    """Create one canonical-precision accessible SI number input."""
+    resolved_spec = spec or NumberInputSpec(
+        minimum=-_MAX_SAFE_FLOAT,
+        maximum=_MAX_SAFE_FLOAT,
+        decimals=11,
+    )
+    return engineering_number_input(name, value, resolved_spec)
+
+
+def coordinate_input(name: str, value: float) -> QDoubleSpinBox:
+    """Create one canonical-precision coordinate editor."""
+    return number_input(
+        name,
+        value,
+        NumberInputSpec(
+            suffix=" m",
+            minimum=-_MAX_SAFE_FLOAT,
+            maximum=_MAX_SAFE_FLOAT,
+            decimals=11,
+        ),
+    )
+
+
+class _SafeIntegerValidator(QValidator):
+    """Accept one decimal integer in the shared JSON-safe nonnegative range."""
+
+    def validate(
+        self, input_text: str | None, position: int
+    ) -> tuple[QValidator.State, str, int]:
+        input_text = input_text or ""
+        if input_text == "":
+            state = QValidator.State.Intermediate
+        elif not input_text.isascii() or not input_text.isdecimal():
+            state = QValidator.State.Invalid
+        elif len(input_text) < len(_MAX_SAFE_INTEGER_TEXT):
+            state = QValidator.State.Acceptable
+        elif input_text <= _MAX_SAFE_INTEGER_TEXT:
+            state = QValidator.State.Acceptable
+        else:
+            state = QValidator.State.Invalid
+        return state, input_text, position
+
+
+class SafeIntegerEdit(QLineEdit):
+    """Exact integer editor that does not narrow the v1 safe-integer contract."""
+
+    def __init__(self, name: str, value: int) -> None:
+        super().__init__()
+        self._name = name
+        self.setAccessibleName(name)
+        self.setMaxLength(len(_MAX_SAFE_INTEGER_TEXT))
+        self.setValidator(_SafeIntegerValidator(self))
+        self.set_value(value)
+
+    def set_value(self, value: int) -> None:
+        """Set one already validated exact safe integer without float conversion."""
+        if isinstance(value, bool) or not isinstance(value, int):
+            raise TypeError(f"{self._name} must be an integer")
+        if not 0 <= value <= MAX_CANONICAL_SAFE_INTEGER:
+            raise ValueError(f"{self._name} must lie within cross-runtime safe range")
+        self.setText(str(value))
+
+    def value(self) -> int:
+        """Return the exact integer or reject an incomplete editor value."""
+        if not self.hasAcceptableInput():
+            raise ValueError(f"{self._name} must lie within cross-runtime safe range")
+        return int(self.text())
+
+
+class MaterialEditor(QGroupBox):
+    """Editable surface identity and full v1 material parameter collection."""
+
+    def __init__(self, title: str, value: SurfaceMaterialDraft) -> None:
+        super().__init__(title)
+        self.surface_id = QLineEdit(value.surface_id)
+        self.surface_id.setAccessibleName(f"{title} surface ID")
+        self.surface_id.setToolTip(
+            f"Stable identifier for {title.lower()}; included in validated readback."
+        )
+        self.fields: dict[str, QDoubleSpinBox] = {}
+        layout = QFormLayout(self)
+        layout.addRow("Surface ID", self.surface_id)
+        for name, label, suffix, step, minimum, maximum in _MATERIAL_FIELDS:
+            field = number_input(
+                f"{title} {label}",
+                getattr(value, name),
+                NumberInputSpec(suffix, step, minimum, maximum, decimals=11),
+            )
+            self.fields[name] = field
+            layout.addRow(label, field)
+
+    def draft(self) -> SurfaceMaterialDraft:
+        """Read the current widgets without applying separate UI validation."""
+        values = {name: field.value() for name, field in self.fields.items()}
+        return SurfaceMaterialDraft(self.surface_id.text(), **values)
+
+    def set_draft(self, value: SurfaceMaterialDraft) -> None:
+        """Replace all visible material values from one validated draft."""
+        self.surface_id.setText(value.surface_id)
+        for name, field in self.fields.items():
+            field.setValue(getattr(value, name))
+
+
+class RegionalOverlayRow(QGroupBox):
+    """One removable bounded regional overlay row."""
+
+    def __init__(
+        self,
+        ordinal: int,
+        value: RegionalOverlayDraft,
+        remove: Callable[[RegionalOverlayRow], None],
+    ) -> None:
+        super().__init__(f"Regional overlay {ordinal}")
+        self.region_id = QLineEdit(value.region_id)
+        self.region_id.setToolTip(
+            "Stable overlay identifier; it must be unique within the regional plan."
+        )
+        self.precedence = SafeIntegerEdit(
+            f"Overlay {ordinal} precedence", value.precedence
+        )
+        self.precedence.setToolTip(
+            "Overlay selection precedence. Higher values win when intervals overlap."
+        )
+        self.lower_coordinate = coordinate_input(
+            f"Overlay {ordinal} lower coordinate",
+            value.lower_coordinate_m,
+        )
+        self.upper_coordinate = coordinate_input(
+            f"Overlay {ordinal} upper coordinate",
+            value.upper_coordinate_m,
+        )
+        self.material = MaterialEditor(f"Overlay {ordinal} material", value.surface)
+        self.remove_button = QPushButton(f"Remove overlay {ordinal}")
+        self.remove_button.setToolTip(
+            "Remove this overlay from the unvalidated draft; one overlay is required."
+        )
+        self.remove_button.clicked.connect(lambda: remove(self))
+        form = QFormLayout()
+        form.addRow("Region ID", self.region_id)
+        form.addRow("Precedence", self.precedence)
+        form.addRow("Lower coordinate", self.lower_coordinate)
+        form.addRow("Upper coordinate", self.upper_coordinate)
+        layout = QVBoxLayout(self)
+        layout.addLayout(form)
+        layout.addWidget(self.material)
+        layout.addWidget(self.remove_button, alignment=Qt.AlignmentFlag.AlignRight)
+
+    def draft(self) -> RegionalOverlayDraft:
+        """Read one overlay for authoritative contract validation."""
+        return RegionalOverlayDraft(
+            self.region_id.text(),
+            self.precedence.value(),
+            self.lower_coordinate.value(),
+            self.upper_coordinate.value(),
+            self.material.draft(),
+        )
+
+
+__all__ = [
+    "MaterialEditor",
+    "RegionalOverlayRow",
+    "SafeIntegerEdit",
+    "coordinate_input",
+    "number_input",
+]
diff --git a/src/rate_of_closure/variation/canonical_numeric_json.py b/src/rate_of_closure/variation/canonical_numeric_json.py
index b031cc6f8..a78b6a39e 100644
--- a/src/rate_of_closure/variation/canonical_numeric_json.py
+++ b/src/rate_of_closure/variation/canonical_numeric_json.py
@@ -1,5 +1,7 @@
-"""Compatibility import for the shared canonical numeric JSON encoder."""
+"""Capability-observation facade for canonical numeric JSON encoding."""
 
-from shared.python.swing_sim.canonical_numeric_json import canonical_numeric_json
+from shared.python.swing_sim.canonical_numeric_json import (
+    canonical_numeric_json_extended_floats as canonical_numeric_json,
+)
 
 __all__ = ["canonical_numeric_json"]
diff --git a/src/rate_of_closure/web/src/components/RegionalSurfacePlanPanel.test.tsx b/src/rate_of_closure/web/src/components/RegionalSurfacePlanPanel.test.tsx
index 1356cb121..72d77dc51 100644
--- a/src/rate_of_closure/web/src/components/RegionalSurfacePlanPanel.test.tsx
+++ b/src/rate_of_closure/web/src/components/RegionalSurfacePlanPanel.test.tsx
@@ -1,13 +1,22 @@
-import { fireEvent, render, screen } from "@testing-library/react";
-import { describe, expect, it } from "vitest";
+import { fireEvent, render, screen, waitFor } from "@testing-library/react";
+import { afterEach, describe, expect, it, vi } from "vitest";
 
 import {
   MAX_REGIONAL_SURFACE_EDITOR_ROWS,
   buildGroundRegionalSurfacePlanRequest,
+  editorDraftFromGroundRegionalSurfacePlanRequest,
   illustrativeRegionalSurfacePlanDraft,
+  regionalSurfacePlanRequestForDraft,
 } from "../model/regionalSurfacePlan";
+import {
+  MAX_GROUND_REGIONAL_PLAN_WIRE_BYTES,
+  parseGroundRegionalMaterialPlanRequest,
+  stableGroundRegionalMaterialPlanJson,
+} from "../model/groundRegionalPlan";
 import { RegionalSurfacePlanPanel } from "./RegionalSurfacePlanPanel";
 
+afterEach(() => vi.restoreAllMocks());
+
 describe("RegionalSurfacePlanPanel", () => {
   it("matches Python provenance for the disclosed illustrative draft", () => {
     const request = buildGroundRegionalSurfacePlanRequest(
@@ -49,6 +58,32 @@ describe("RegionalSurfacePlanPanel", () => {
       .not.toBeInTheDocument();
   });
 
+  it("accepts the exact cross-runtime maximum precedence", () => {
+    render(<RegionalSurfacePlanPanel />);
+    fireEvent.change(screen.getByLabelText("Overlay 1 precedence"), {
+      target: { value: String(Number.MAX_SAFE_INTEGER) },
+    });
+
+    fireEvent.click(screen.getByRole("button", { name: "Validate surface plan" }));
+
+    expect(screen.queryByRole("alert")).not.toBeInTheDocument();
+    expect(screen.getByRole("status", { name: "Regional surface plan readback" }))
+      .toHaveTextContent("1 overlay(s)");
+  });
+
+  it("rejects integer-valued material numbers beyond the cross-runtime safe range", () => {
+    render(<RegionalSurfacePlanPanel />);
+    fireEvent.change(screen.getByLabelText("Base Firmness (Pa)"), {
+      target: { value: "10000000000000000" },
+    });
+
+    fireEvent.click(screen.getByRole("button", { name: "Validate surface plan" }));
+
+    expect(screen.getByRole("alert")).toHaveTextContent(/cross-runtime safe range/i);
+    expect(screen.queryByRole("status", { name: "Regional surface plan readback" }))
+      .not.toBeInTheDocument();
+  });
+
   it("links invalid interval fields to an accessible error without clearing input", () => {
     render(<RegionalSurfacePlanPanel />);
     const lower = screen.getByLabelText("Overlay 1 lower coordinate (m)");
@@ -84,6 +119,123 @@ describe("RegionalSurfacePlanPanel", () => {
       .toHaveLength(MAX_REGIONAL_SURFACE_EDITOR_ROWS);
     expect(add).toBeDisabled();
     expect(screen.queryByRole("button", { name: /run|play/i })).not.toBeInTheDocument();
-    expect(screen.getByText(/session-only draft/i)).toBeInTheDocument();
+    expect(screen.getByText(/workspace persistence remains separate/i)).toBeInTheDocument();
+  });
+
+  it("preserves an unchanged imported editor request exactly", () => {
+    const request = buildGroundRegionalSurfacePlanRequest({
+      ...illustrativeRegionalSurfacePlanDraft(), request_id: "imported-plan",
+    });
+
+    const imported = editorDraftFromGroundRegionalSurfacePlanRequest(request);
+
+    expect(regionalSurfacePlanRequestForDraft(imported, request)).toBe(request);
+    expect(stableGroundRegionalMaterialPlanJson(
+      regionalSurfacePlanRequestForDraft(imported, request),
+    )).toBe(stableGroundRegionalMaterialPlanJson(request));
+
+    const payload = JSON.parse(stableGroundRegionalMaterialPlanJson(request));
+    payload.provenance.input_sha256 = "0".repeat(64);
+    const mismatched = parseGroundRegionalMaterialPlanRequest(payload);
+    expect(() => editorDraftFromGroundRegionalSurfacePlanRequest(mismatched))
+      .toThrow(/digest does not match/i);
+  });
+
+  it("imports only after complete validation and retains the prior draft on error", async () => {
+    render(<RegionalSurfacePlanPanel />);
+    const request = buildGroundRegionalSurfacePlanRequest({
+      ...illustrativeRegionalSurfacePlanDraft(), request_id: "browser-opened-plan",
+    });
+    const input = screen.getByLabelText("Import regional surface plan JSON file");
+    const validText = stableGroundRegionalMaterialPlanJson(request);
+    const valid = {
+      name: "valid.json",
+      size: new TextEncoder().encode(validText).byteLength,
+      text: vi.fn().mockResolvedValue(validText),
+    };
+
+    fireEvent.change(input, { target: { files: [valid] } });
+    await waitFor(() => expect(screen.getByLabelText("Regional plan request ID"))
+      .toHaveValue("browser-opened-plan"));
+
+    const corrupt = {
+      name: "corrupt.json", size: 47,
+      text: vi.fn().mockResolvedValue('{"request_id":"one","request_id":"two"}'),
+    };
+    fireEvent.change(input, { target: { files: [corrupt] } });
+    await waitFor(() => expect(screen.getByRole("alert")).toHaveTextContent(/duplicate/i));
+    expect(screen.getByLabelText("Regional plan request ID"))
+      .toHaveValue("browser-opened-plan");
+
+    const invalidSyntax = {
+      name: "syntax.json", size: 9,
+      text: vi.fn().mockResolvedValue("{not-json"),
+    };
+    fireEvent.change(input, { target: { files: [invalidSyntax] } });
+    await waitFor(() => expect(screen.getByRole("alert")).toHaveTextContent(/invalid profile json/i));
+    expect(screen.getByLabelText("Regional plan request ID"))
+      .toHaveValue("browser-opened-plan");
+  });
+
+  it("rejects non-editor qualification without populating the editor", async () => {
+    render(<RegionalSurfacePlanPanel />);
+    const request = buildGroundRegionalSurfacePlanRequest(
+      illustrativeRegionalSurfacePlanDraft(),
+    );
+    const payload = JSON.parse(stableGroundRegionalMaterialPlanJson(request));
+    payload.provenance.producer = "external.course.authority";
+    const input = screen.getByLabelText("Import regional surface plan JSON file");
+    const file = {
+      name: "external.json", size: 100,
+      text: vi.fn().mockResolvedValue(JSON.stringify(payload)),
+    };
+
+    fireEvent.change(input, { target: { files: [file] } });
+
+    await waitFor(() => expect(screen.getByRole("alert")).toHaveTextContent(/editor producer/i));
+    expect(screen.getByLabelText("Regional plan request ID"))
+      .toHaveValue("illustrative-regional-plan");
+  });
+
+  it("rejects oversized browser files before allocating their text", async () => {
+    render(<RegionalSurfacePlanPanel />);
+    const input = screen.getByLabelText("Import regional surface plan JSON file");
+    const text = vi.fn().mockResolvedValue("{}");
+    const file = {
+      name: "oversize.json", size: MAX_GROUND_REGIONAL_PLAN_WIRE_BYTES + 1, text,
+    };
+
+    fireEvent.change(input, { target: { files: [file] } });
+
+    await waitFor(() => expect(screen.getByRole("alert")).toHaveTextContent(/maximum wire size/i));
+    expect(text).not.toHaveBeenCalled();
+    expect(screen.getByLabelText("Regional plan request ID"))
+      .toHaveValue("illustrative-regional-plan");
+  });
+
+  it("downloads canonical bytes and revokes its object URL", () => {
+    const createUrl = vi.fn(() => "blob:regional-plan");
+    const revokeUrl = vi.fn();
+    Object.defineProperty(URL, "createObjectURL", { configurable: true, value: createUrl });
+    Object.defineProperty(URL, "revokeObjectURL", { configurable: true, value: revokeUrl });
+    const click = vi.spyOn(HTMLAnchorElement.prototype, "click").mockImplementation(() => {});
+    render(<RegionalSurfacePlanPanel />);
+
+    fireEvent.click(screen.getByRole("button", { name: "Download regional surface plan JSON" }));
+
+    expect(createUrl).toHaveBeenCalledOnce();
+    expect(click).toHaveBeenCalledOnce();
+    expect(revokeUrl).toHaveBeenCalledWith("blob:regional-plan");
+  });
+
+  it("treats an empty browser chooser as cancellation", () => {
+    render(<RegionalSurfacePlanPanel />);
+    const input = screen.getByLabelText("Import regional surface plan JSON file");
+
+    fireEvent.change(input, { target: { files: [] } });
+
+    expect(screen.queryByRole("alert")).not.toBeInTheDocument();
+    expect(screen.getByLabelText("Regional plan request ID"))
+      .toHaveValue("illustrative-regional-plan");
   });
 });
diff --git a/src/rate_of_closure/web/src/components/RegionalSurfacePlanPanel.tsx b/src/rate_of_closure/web/src/components/RegionalSurfacePlanPanel.tsx
index 068fff039..6e8fc544c 100644
--- a/src/rate_of_closure/web/src/components/RegionalSurfacePlanPanel.tsx
+++ b/src/rate_of_closure/web/src/components/RegionalSurfacePlanPanel.tsx
@@ -1,13 +1,19 @@
-import { useState } from "react";
+import { type ChangeEvent, useRef, useState } from "react";
 
 import {
   MAX_REGIONAL_SURFACE_EDITOR_ROWS,
-  buildGroundRegionalSurfacePlanRequest,
+  editorDraftFromGroundRegionalSurfacePlanRequest,
   illustrativeRegionalSurfacePlanDraft,
+  regionalSurfacePlanRequestForDraft,
   type RegionalOverlayDraft,
   type RegionalSurfacePlanDraft,
   type SurfaceMaterialDraft,
 } from "../model/regionalSurfacePlan";
+import {
+  downloadRegionalSurfacePlanRequest,
+  readRegionalSurfacePlanFile,
+} from "../model/regionalSurfacePlanFiles";
+import type { GroundRegionalMaterialPlanRequest } from "../model/groundRegionalPlan";
 
 type NumericMaterialKey = Exclude<keyof SurfaceMaterialDraft, "surface_id">;
 
@@ -133,15 +139,20 @@ function nextOverlay(draft: RegionalSurfacePlanDraft): RegionalOverlayDraft {
 
 export function RegionalSurfacePlanPanel() {
   const [draft, setDraft] = useState(illustrativeRegionalSurfacePlanDraft);
+  const [importedRequest, setImportedRequest] =
+    useState<GroundRegionalMaterialPlanRequest | null>(null);
   const [error, setError] = useState<string | null>(null);
   const [readback, setReadback] = useState<string | null>(null);
+  const [fileStatus, setFileStatus] = useState<string | null>(null);
   const [validationAttempted, setValidationAttempted] = useState(false);
+  const fileInput = useRef<HTMLInputElement>(null);
   const updateDraft = (
     transform: (current: RegionalSurfacePlanDraft) => RegionalSurfacePlanDraft,
   ) => {
     setDraft(transform);
     setError(null);
     setReadback(null);
+    setFileStatus(null);
     setValidationAttempted(false);
   };
   const updateRegion = (index: number, value: RegionalOverlayDraft) =>
@@ -152,7 +163,7 @@ export function RegionalSurfacePlanPanel() {
   const validate = () => {
     setValidationAttempted(true);
     try {
-      const request = buildGroundRegionalSurfacePlanRequest(draft);
+      const request = regionalSurfacePlanRequestForDraft(draft, importedRequest);
       setError(null);
       setReadback(
         `${request.schema_version} · ${request.unit_system} · ` +
@@ -165,6 +176,35 @@ export function RegionalSurfacePlanPanel() {
       setError(caught instanceof Error ? caught.message : "Surface plan is invalid");
     }
   };
+  const importFile = async (event: ChangeEvent<HTMLInputElement>) => {
+    const file = event.currentTarget.files?.[0];
+    event.currentTarget.value = "";
+    if (file === undefined) return;
+    try {
+      const request = await readRegionalSurfacePlanFile(file);
+      const importedDraft = editorDraftFromGroundRegionalSurfacePlanRequest(request);
+      setDraft(importedDraft);
+      setImportedRequest(request);
+      setValidationAttempted(false);
+      setError(null);
+      setReadback(`${request.schema_version} · exact imported provenance ${request.provenance.input_sha256}`);
+      setFileStatus(`Imported ${file.name}. No physics executed.`);
+    } catch (caught) {
+      setError(caught instanceof Error ? caught.message : "Regional plan import failed");
+      setFileStatus("Import failed; the editor and prior validated readback were preserved.");
+    }
+  };
+  const download = () => {
+    try {
+      const request = regionalSurfacePlanRequestForDraft(draft, importedRequest);
+      downloadRegionalSurfacePlanRequest(request);
+      setError(null);
+      setFileStatus("Download prepared. Your browser controls the destination and overwrite behavior.");
+    } catch (caught) {
+      setError(caught instanceof Error ? caught.message : "Regional plan download failed");
+      setFileStatus("Download failed; no browser filesystem state was changed.");
+    }
+  };
   return (
     <section aria-labelledby="regional-surface-plan-title" className="space-y-4">
       <header>
@@ -181,7 +221,8 @@ export function RegionalSurfacePlanPanel() {
         This slice validates a material plan only; it does not run physics or playback.
       </div>
       <p className="text-xs text-slate-500">
-        Session-only draft: the current web workspace adapter does not persist model inputs.
+        Import/download persists this canonical request only; workspace persistence remains separate.
+        Browser downloads cannot promise a native path, atomic replacement, or recent-file access.
       </p>
       <fieldset className="rounded-xl border border-slate-700/80 bg-slate-900/60 p-4">
         <legend className="font-semibold text-slate-200">Plan identity and provenance</legend>
@@ -242,6 +283,9 @@ export function RegionalSurfacePlanPanel() {
           }))} />
       ))}
       <div className="flex flex-wrap gap-3">
+        <input ref={fileInput} type="file" accept=".json,application/json"
+          aria-label="Import regional surface plan JSON file" className="sr-only"
+          onChange={(event) => { void importFile(event); }} />
         <button type="button" aria-label="Add regional overlay"
           disabled={draft.regions.length >= MAX_REGIONAL_SURFACE_EDITOR_ROWS}
           onClick={() => updateDraft((current) => ({
@@ -250,6 +294,16 @@ export function RegionalSurfacePlanPanel() {
           className="rounded-md border border-sky-500/60 px-3 py-2 text-sm text-sky-200 disabled:opacity-40">
           Add overlay
         </button>
+        <button type="button" aria-label="Import regional surface plan JSON"
+          onClick={() => fileInput.current?.click()}
+          className="rounded-md border border-sky-500/60 px-3 py-2 text-sm text-sky-200">
+          Import JSON
+        </button>
+        <button type="button" aria-label="Download regional surface plan JSON"
+          onClick={download}
+          className="rounded-md border border-sky-500/60 px-3 py-2 text-sm text-sky-200">
+          Download JSON
+        </button>
         <button type="button" aria-label="Validate surface plan" onClick={validate}
           className="rounded-md bg-sky-500 px-3 py-2 text-sm font-semibold text-slate-950">
           Validate and preview
@@ -263,6 +317,8 @@ export function RegionalSurfacePlanPanel() {
         className="rounded-md border border-emerald-500/50 bg-emerald-500/10 p-3 text-sm text-emerald-100">
         Validated readback: {readback}
       </p>}
+      {fileStatus !== null && <p role="status" aria-label="Regional surface plan file status"
+        className="text-xs text-slate-400">{fileStatus}</p>}
     </section>
   );
 }
diff --git a/src/rate_of_closure/web/src/model/capabilityObservationEnsemble.test.ts b/src/rate_of_closure/web/src/model/capabilityObservationEnsemble.test.ts
index eb5b1311d..b5f399327 100644
--- a/src/rate_of_closure/web/src/model/capabilityObservationEnsemble.test.ts
+++ b/src/rate_of_closure/web/src/model/capabilityObservationEnsemble.test.ts
@@ -294,6 +294,7 @@ describe("capability observation wire parity", () => {
       row.parameters = [
         { parameterId: "half", unit: "1", nominalValue: 1.000000000005, perturbedValue: -1.000000000005 },
         { parameterId: "tiny", unit: "1", nominalValue: -0, perturbedValue: 0.0000001 },
+        { parameterId: "threshold", unit: "1", nominalValue: 1e20, perturbedValue: 1e-11 },
         { parameterId: "large", unit: "1", nominalValue: 1e21, perturbedValue: 1e-12 },
       ];
     });
@@ -302,6 +303,8 @@ describe("capability observation wire parity", () => {
     expect(json).toContain('"perturbed.half":-1.00000000001');
     expect(json).toContain('"nominal.tiny":0');
     expect(json).toContain('"perturbed.tiny":0.0000001');
+    expect(json).toContain('"nominal.threshold":100000000000000000000');
+    expect(json).toContain('"perturbed.threshold":0.00000000001');
     expect(json).toContain('"nominal.large":1000000000000000000000');
     expect(json).toContain('"perturbed.large":0');
     expect(json).not.toMatch(/"(?:nominal|perturbed)\.[^"]+":"/);
diff --git a/src/rate_of_closure/web/src/model/regionalSurfacePlan.ts b/src/rate_of_closure/web/src/model/regionalSurfacePlan.ts
index 33301feeb..a52cdc0d3 100644
--- a/src/rate_of_closure/web/src/model/regionalSurfacePlan.ts
+++ b/src/rate_of_closure/web/src/model/regionalSurfacePlan.ts
@@ -6,6 +6,7 @@ import {
   GROUND_REGIONAL_PLAN_REQUEST_VERSION,
   parseGroundRegionalMaterialPlanRequest,
   type GroundRegionalMaterialPlanRequest,
+  type GroundRegionalMaterialRegion,
 } from "./groundRegionalPlan";
 import { GROUND_TARGET_FRAME, type GroundSurfaceProfile } from "./flightGroundTypes";
 import { canonicalGroundJson } from "./flightGroundContract";
@@ -106,6 +107,70 @@ const surfacePayload = (surface: SurfaceMaterialDraft): GroundSurfaceProfile =>
   surface_velocity_m_s: [0, 0, 0],
 });
 
+const surfaceDraft = (surface: GroundSurfaceProfile): SurfaceMaterialDraft => ({
+  surface_id: surface.surface_id,
+  normal_restitution: surface.normal_restitution,
+  static_friction: surface.static_friction,
+  kinetic_friction: surface.kinetic_friction,
+  rolling_resistance: surface.rolling_resistance,
+  firmness_pa: surface.firmness_pa,
+  hardness_fraction: surface.hardness_fraction,
+  grass_height_m: surface.grass_height_m,
+  compressibility_fraction: surface.compressibility_fraction,
+  compression_damping_fraction: surface.compression_damping_fraction,
+  turf_density_kg_m3: surface.turf_density_kg_m3,
+  moisture_fraction: surface.moisture_fraction,
+});
+
+const assertEditorSurface = (surface: GroundSurfaceProfile): void => {
+  if (surface.provider_id !== REGIONAL_SURFACE_EDITOR_PROVIDER ||
+      surface.provider_version !== REGIONAL_SURFACE_EDITOR_VERSION) {
+    throw new RangeError("surface is not qualified by the editor provider v1");
+  }
+};
+
+const overlayDraft = (region: GroundRegionalMaterialRegion): RegionalOverlayDraft => ({
+  region_id: region.region_id,
+  precedence: region.precedence,
+  lower_coordinate_m: region.lower_coordinate_m,
+  upper_coordinate_m: region.upper_coordinate_m,
+  surface: surfaceDraft(region.surface),
+});
+
+/** Project a fully validated editor-v1 request without relabelling evidence. */
+export const editorDraftFromGroundRegionalSurfacePlanRequest = (
+  request: GroundRegionalMaterialPlanRequest,
+): RegionalSurfacePlanDraft => {
+  if (request.provenance.producer !== REGIONAL_SURFACE_EDITOR_PROVIDER ||
+      request.provenance.producer_version !== REGIONAL_SURFACE_EDITOR_VERSION) {
+    throw new RangeError("request is not qualified by the editor producer v1");
+  }
+  if (canonicalGroundJson(request.axis_origin_m) !== "[0,0,0]" ||
+      canonicalGroundJson(request.axis_unit) !== "[1,0,0]") {
+    throw new RangeError("request uses an unsupported editor axis qualification");
+  }
+  if (request.regions.length > MAX_REGIONAL_SURFACE_EDITOR_ROWS) {
+    throw new RangeError(
+      `editor supports one to at most ${MAX_REGIONAL_SURFACE_EDITOR_ROWS} regions`,
+    );
+  }
+  assertEditorSurface(request.base_surface);
+  request.regions.forEach((region) => assertEditorSurface(region.surface));
+  const draft: RegionalSurfacePlanDraft = {
+    request_id: request.request_id,
+    lower_coordinate_m: request.lower_coordinate_m,
+    upper_coordinate_m: request.upper_coordinate_m,
+    source_revision: request.provenance.source_revision,
+    calibration_kind: "unvalidated",
+    base_surface: surfaceDraft(request.base_surface),
+    regions: request.regions.map(overlayDraft),
+  };
+  if (request.provenance.input_sha256 !== sha256Text(canonicalGroundJson(draft))) {
+    throw new RangeError("editor provenance digest does not match the editable request");
+  }
+  return draft;
+};
+
 export const buildGroundRegionalSurfacePlanRequest = (
   draft: RegionalSurfacePlanDraft,
 ): GroundRegionalMaterialPlanRequest => {
@@ -145,3 +210,17 @@ export const buildGroundRegionalSurfacePlanRequest = (
     schema_version: GROUND_REGIONAL_PLAN_REQUEST_VERSION,
   });
 };
+
+/** Keep untouched imported evidence exact; bind fresh provenance after edits. */
+export const regionalSurfacePlanRequestForDraft = (
+  draft: RegionalSurfacePlanDraft,
+  importedRequest: GroundRegionalMaterialPlanRequest | null = null,
+): GroundRegionalMaterialPlanRequest => {
+  if (importedRequest !== null) {
+    const importedDraft = editorDraftFromGroundRegionalSurfacePlanRequest(importedRequest);
+    if (canonicalGroundJson(draft) === canonicalGroundJson(importedDraft)) {
+      return importedRequest;
+    }
+  }
+  return buildGroundRegionalSurfacePlanRequest(draft);
+};
diff --git a/src/rate_of_closure/web/src/model/regionalSurfacePlanFiles.ts b/src/rate_of_closure/web/src/model/regionalSurfacePlanFiles.ts
new file mode 100644
index 000000000..2ed7359b7
--- /dev/null
+++ b/src/rate_of_closure/web/src/model/regionalSurfacePlanFiles.ts
@@ -0,0 +1,40 @@
+/** Browser-only strict import/download boundary for regional plan requests. */
+
+import {
+  MAX_GROUND_REGIONAL_PLAN_WIRE_BYTES,
+  groundRegionalMaterialPlanRequestFromJson,
+  stableGroundRegionalMaterialPlanJson,
+  type GroundRegionalMaterialPlanRequest,
+} from "./groundRegionalPlan";
+
+export interface RegionalPlanFile {
+  readonly name: string;
+  readonly size: number;
+  text(): Promise<string>;
+}
+
+export const readRegionalSurfacePlanFile = async (
+  file: RegionalPlanFile,
+): Promise<GroundRegionalMaterialPlanRequest> => {
+  if (file.size > MAX_GROUND_REGIONAL_PLAN_WIRE_BYTES) {
+    throw new RangeError("regional material plan exceeds maximum wire size");
+  }
+  return groundRegionalMaterialPlanRequestFromJson(await file.text());
+};
+
+export const downloadRegionalSurfacePlanRequest = (
+  request: GroundRegionalMaterialPlanRequest,
+): void => {
+  const text = stableGroundRegionalMaterialPlanJson(request);
+  const url = URL.createObjectURL(
+    new Blob([text], { type: "application/json;charset=utf-8" }),
+  );
+  try {
+    const anchor = document.createElement("a");
+    anchor.href = url;
+    anchor.download = "regional-surface-plan.json";
+    anchor.click();
+  } finally {
+    URL.revokeObjectURL(url);
+  }
+};
diff --git a/src/shared/python/swing_sim/canonical_numeric_json.py b/src/shared/python/swing_sim/canonical_numeric_json.py
index 9934a77c1..25d6e69a6 100644
--- a/src/shared/python/swing_sim/canonical_numeric_json.py
+++ b/src/shared/python/swing_sim/canonical_numeric_json.py
@@ -9,7 +9,7 @@
 from typing import Any
 
 _FLOAT_QUANTUM = Decimal("0.00000000001")
-_MAX_SAFE_INTEGER = 9_007_199_254_740_991
+MAX_CANONICAL_SAFE_INTEGER = 9_007_199_254_740_991
 
 
 def _string_token(value: str) -> str:
@@ -18,9 +18,11 @@ def _string_token(value: str) -> str:
     return json.dumps(value, ensure_ascii=False)
 
 
-def _canonical_float_token(value: float) -> str:
+def _canonical_float_token(value: float, *, allow_extended_finite: bool = False) -> str:
     if not math.isfinite(value):
         raise ValueError("canonical JSON requires finite floats")
+    if not allow_extended_finite and abs(value) > MAX_CANONICAL_SAFE_INTEGER:
+        raise ValueError("canonical JSON number exceeds cross-runtime safe range")
     if value == 0 or value.is_integer():
         return "0" if value == 0 else str(int(value))
     rounded = Decimal.from_float(value).quantize(_FLOAT_QUANTUM, rounding=ROUND_HALF_UP)
@@ -34,16 +36,17 @@ def canonical_numeric_float(value: float) -> float:
     return float(_canonical_float_token(value))
 
 
-def canonical_numeric_json(value: Any) -> str:
-    """Serialize JSON-compatible data with stable fixed-point float tokens."""
+def _canonical_numeric_json(value: Any, *, allow_extended_finite: bool) -> str:
     if value is None:
         return "null"
     if isinstance(value, bool):
         return "true" if value else "false"
     if isinstance(value, float):
-        return _canonical_float_token(value)
+        return _canonical_float_token(
+            value, allow_extended_finite=allow_extended_finite
+        )
     if isinstance(value, int):
-        if abs(value) > _MAX_SAFE_INTEGER:
+        if abs(value) > MAX_CANONICAL_SAFE_INTEGER:
             raise ValueError("canonical JSON integer exceeds cross-runtime safe range")
         return str(value)
     if isinstance(value, str):
@@ -51,14 +54,37 @@ def canonical_numeric_json(value: Any) -> str:
     if isinstance(value, Mapping):
         if not all(isinstance(key, str) for key in value):
             raise TypeError("canonical JSON object keys must be strings")
-        items = (
-            f"{_string_token(key)}:{canonical_numeric_json(value[key])}"
-            for key in sorted(value)
-        )
+
+        def item_token(key: str) -> str:
+            token = _canonical_numeric_json(
+                value[key], allow_extended_finite=allow_extended_finite
+            )
+            return f"{_string_token(key)}:{token}"
+
+        items = (item_token(key) for key in sorted(value))
         return "{" + ",".join(items) + "}"
     if isinstance(value, (list, tuple)):
-        return "[" + ",".join(canonical_numeric_json(item) for item in value) + "]"
+        items = (
+            _canonical_numeric_json(item, allow_extended_finite=allow_extended_finite)
+            for item in value
+        )
+        return "[" + ",".join(items) + "]"
     raise TypeError(f"unsupported canonical JSON value: {type(value).__name__}")
 
 
-__all__ = ["canonical_numeric_float", "canonical_numeric_json"]
+def canonical_numeric_json(value: Any) -> str:
+    """Serialize JSON-compatible data within the shared runtime-safe range."""
+    return _canonical_numeric_json(value, allow_extended_finite=False)
+
+
+def canonical_numeric_json_extended_floats(value: Any) -> str:
+    """Serialize finite floats beyond the safe range while retaining safe integers."""
+    return _canonical_numeric_json(value, allow_extended_finite=True)
+
+
+__all__ = [
+    "MAX_CANONICAL_SAFE_INTEGER",
+    "canonical_numeric_float",
+    "canonical_numeric_json",
+    "canonical_numeric_json_extended_floats",
+]
diff --git a/src/shared/python/swing_sim/ground/contract_types.py b/src/shared/python/swing_sim/ground/contract_types.py
index 1844ddb68..deb2979e1 100644
--- a/src/shared/python/swing_sim/ground/contract_types.py
+++ b/src/shared/python/swing_sim/ground/contract_types.py
@@ -6,7 +6,10 @@
 from dataclasses import dataclass
 from typing import TYPE_CHECKING, Any, TypeAlias
 
-from shared.python.swing_sim.canonical_numeric_json import canonical_numeric_float
+from shared.python.swing_sim.canonical_numeric_json import (
+    MAX_CANONICAL_SAFE_INTEGER,
+    canonical_numeric_float,
+)
 
 if TYPE_CHECKING:
     from enum import StrEnum
@@ -18,7 +21,6 @@
 UNIT_SYSTEM_SI = "SI"
 Vector3: TypeAlias = tuple[float, float, float]
 _UNIT_TOLERANCE = 1e-9
-_MAX_SAFE_INTEGER = 9_007_199_254_740_991
 _MIN_CANONICAL_POSITIVE = 0.00000000001
 _TEXT_EDGE_WHITESPACE = " \t\r\n\f\v"
 
@@ -29,6 +31,8 @@ def _raw_finite(value: float, name: str) -> float:
     number = float(value)
     if not math.isfinite(number):
         raise ValueError(f"{name} must be finite")
+    if abs(number) > MAX_CANONICAL_SAFE_INTEGER:
+        raise ValueError(f"{name} must lie within the cross-runtime safe range")
     return number
 
 
@@ -67,10 +71,10 @@ def _integer(value: int | float, name: str, minimum: int = 0) -> int:
         normalized = int(value)
     else:
         raise ValueError(f"{name} must be an integer")
-    if not minimum <= normalized <= _MAX_SAFE_INTEGER:
+    if not minimum <= normalized <= MAX_CANONICAL_SAFE_INTEGER:
         raise ValueError(
             f"{name} must lie within cross-runtime safe range "
-            f"[{minimum}, {_MAX_SAFE_INTEGER}]"
+            f"[{minimum}, {MAX_CANONICAL_SAFE_INTEGER}]"
         )
     return normalized
 
diff --git a/src/shared/python/swing_sim/ground/tests/test_contract_api.py b/src/shared/python/swing_sim/ground/tests/test_contract_api.py
index 8c9815c96..fee25b250 100644
--- a/src/shared/python/swing_sim/ground/tests/test_contract_api.py
+++ b/src/shared/python/swing_sim/ground/tests/test_contract_api.py
@@ -178,6 +178,10 @@ def test_cross_runtime_integer_and_text_edges_fail_closed() -> None:
         replace(_request(), max_events=10**1000)
     with pytest.raises(ValueError, match="safe range"):
         canonical_numeric_json(unsafe_integer)
+    with pytest.raises(ValueError, match="safe range"):
+        canonical_numeric_json(float(unsafe_integer))
+    with pytest.raises(ValueError, match="safe range"):
+        replace(_surface(), firmness_pa=float(unsafe_integer))
     with pytest.raises(ValueError, match="surrogate"):
         replace(_request(), request_id="\ud800")
     with pytest.raises(ValueError, match="surrogate"):
diff --git a/tests/rate_of_closure/test_capability_observation_adapter.py b/tests/rate_of_closure/test_capability_observation_adapter.py
index 24c63c50c..440286866 100644
--- a/tests/rate_of_closure/test_capability_observation_adapter.py
+++ b/tests/rate_of_closure/test_capability_observation_adapter.py
@@ -262,6 +262,7 @@ def test_stable_wire_uses_canonical_numeric_tokens_for_every_float() -> None:
             CapabilitySampleParameter("positive_half", "1", 1.234567890125, 1e-12),
             CapabilitySampleParameter("negative_half", "1", -1.234567890125, -0.0),
             CapabilitySampleParameter("threshold", "1", 1e-11, 1e20),
+            CapabilitySampleParameter("large", "1", 1e21, 1e-12),
             CapabilitySampleParameter(
                 "half_away", "1", 1.000000000005, -1.000000000005
             ),
@@ -278,6 +279,8 @@ def test_stable_wire_uses_canonical_numeric_tokens_for_every_float() -> None:
     assert row_values["perturbed.negative_half"] == 0
     assert row_values["nominal.threshold"] == 0.00000000001
     assert row_values["perturbed.threshold"] == 100000000000000000000
+    assert row_values["nominal.large"] == 1000000000000000000000
+    assert row_values["perturbed.large"] == 0
     assert row_values["nominal.half_away"] == 1.00000000001
     assert row_values["perturbed.half_away"] == -1.00000000001
     assert all(
@@ -287,6 +290,8 @@ def test_stable_wire_uses_canonical_numeric_tokens_for_every_float() -> None:
     )
     assert '"nominal.threshold":0.00000000001' in payload
     assert '"perturbed.threshold":100000000000000000000' in payload
+    assert '"nominal.large":1000000000000000000000' in payload
+    assert "e+21" not in payload
 
 
 def test_parameter_labels_only_uppercase_initial_ascii_letters() -> None:
@@ -321,3 +326,8 @@ def test_parameter_labels_only_uppercase_initial_ascii_letters() -> None:
 def test_canonical_numeric_json_rejects_nonfinite_values(value: float) -> None:
     with pytest.raises(ValueError, match="finite floats"):
         canonical_numeric_json({"value": value})
+
+
+def test_capability_extended_float_policy_keeps_integer_safe_range() -> None:
+    with pytest.raises(ValueError, match="integer.*safe range"):
+        canonical_numeric_json({"value": 9_007_199_254_740_992})
diff --git a/tests/rate_of_closure/test_regional_surface_plan.py b/tests/rate_of_closure/test_regional_surface_plan.py
index 966ea2eb6..814c52eaa 100644
--- a/tests/rate_of_closure/test_regional_surface_plan.py
+++ b/tests/rate_of_closure/test_regional_surface_plan.py
@@ -4,6 +4,7 @@
 
 from dataclasses import replace
 from inspect import signature
+from pathlib import Path
 
 import pytest
 
@@ -12,10 +13,55 @@
     illustrative_regional_surface_plan_draft,
     validate_regional_surface_plan_draft,
 )
+from shared.python.swing_sim.canonical_numeric_json import (
+    MAX_CANONICAL_SAFE_INTEGER,
+)
 
 pytestmark = [pytest.mark.unit, pytest.mark.headless_safe]
 
 
+def test_pyqt_number_input_spec_preserves_behavior_and_structural_bound(qtbot) -> None:  # type: ignore[no-untyped-def]
+    pytest.importorskip("PyQt6")
+    pytest.importorskip("pytestqt")
+    from rate_of_closure.ui.pyqt6.engineering_number_input import (
+        NumberInputSpec,
+        engineering_number_input,
+    )
+
+    spec = NumberInputSpec(
+        suffix=" m",
+        step=0.25,
+        minimum=-2.0,
+        maximum=3.0,
+        decimals=11,
+    )
+    field = engineering_number_input("Test coordinate", 1.5, spec)
+    qtbot.addWidget(field)
+
+    assert NumberInputSpec.__dataclass_params__.frozen is True
+    assert len(signature(engineering_number_input).parameters) <= 4
+    assert field.accessibleName() == "Test coordinate"
+    assert field.value() == pytest.approx(1.5)
+    assert field.suffix() == " m"
+    assert field.singleStep() == pytest.approx(0.25)
+    assert field.minimum() == pytest.approx(-2.0)
+    assert field.maximum() == pytest.approx(3.0)
+    assert field.decimals() == 11
+    assert "Test coordinate" in field.toolTip()
+
+
+def test_pyqt_number_input_spec_rejects_invalid_configuration() -> None:
+    pytest.importorskip("PyQt6")
+    from rate_of_closure.ui.pyqt6.engineering_number_input import NumberInputSpec
+
+    with pytest.raises(ValueError, match="step must be finite and positive"):
+        NumberInputSpec(suffix="", step=0.0, minimum=0.0, maximum=1.0)
+    with pytest.raises(ValueError, match="minimum must not exceed maximum"):
+        NumberInputSpec(suffix="", step=0.1, minimum=2.0, maximum=1.0)
+    with pytest.raises(ValueError, match="decimals must be an integer"):
+        NumberInputSpec(decimals=True)
+
+
 def test_illustrative_draft_delegates_to_regional_wire_contract() -> None:
     draft = illustrative_regional_surface_plan_draft()
 
@@ -64,39 +110,6 @@ def test_editor_region_limit_fails_before_wire_construction() -> None:
         validate_regional_surface_plan_draft(replace(draft, regions=repeated))
 
 
-def test_pyqt_number_input_spec_preserves_behavior_and_structural_bound(qtbot) -> None:  # type: ignore[no-untyped-def]
-    pytest.importorskip("PyQt6")
-    pytest.importorskip("pytestqt")
-    from rate_of_closure.ui.pyqt6.regional_surface_plan_tab import (
-        NumberInputSpec,
-        _number_input,
-    )
-
-    spec = NumberInputSpec(suffix=" m", step=0.25, minimum=-2.0, maximum=3.0)
-    field = _number_input("Test coordinate", 1.5, spec)
-    qtbot.addWidget(field)
-
-    assert NumberInputSpec.__dataclass_params__.frozen is True
-    assert len(signature(_number_input).parameters) <= 4
-    assert field.accessibleName() == "Test coordinate"
-    assert field.value() == pytest.approx(1.5)
-    assert field.suffix() == " m"
-    assert field.singleStep() == pytest.approx(0.25)
-    assert field.minimum() == pytest.approx(-2.0)
-    assert field.maximum() == pytest.approx(3.0)
-    assert "Test coordinate" in field.toolTip()
-
-
-def test_pyqt_number_input_spec_rejects_invalid_configuration() -> None:
-    pytest.importorskip("PyQt6")
-    from rate_of_closure.ui.pyqt6.regional_surface_plan_tab import NumberInputSpec
-
-    with pytest.raises(ValueError, match="step must be finite and positive"):
-        NumberInputSpec(suffix="", step=0.0, minimum=0.0, maximum=1.0)
-    with pytest.raises(ValueError, match="minimum must not exceed maximum"):
-        NumberInputSpec(suffix="", step=0.1, minimum=2.0, maximum=1.0)
-
-
 def test_pyqt_editor_exposes_warnings_units_and_validated_readback(qtbot) -> None:  # type: ignore[no-untyped-def]
     pytest.importorskip("PyQt6")
     pytest.importorskip("pytestqt")
@@ -155,3 +168,162 @@ def test_pyqt_editor_invalidates_validated_readback_after_draft_change(qtbot) ->
 
     assert tab.readback.toPlainText() == ""
     assert tab.status_label.text() == "Changes not validated"
+
+
+def test_pyqt_open_applies_only_a_fully_valid_editor_request(
+    qtbot, tmp_path: Path, monkeypatch: pytest.MonkeyPatch
+) -> None:  # type: ignore[no-untyped-def]
+    pytest.importorskip("PyQt6")
+    from rate_of_closure.ui.pyqt6 import regional_surface_plan_io
+    from rate_of_closure.ui.pyqt6.regional_surface_plan_tab import (
+        RegionalSurfacePlanTab,
+    )
+
+    request = validate_regional_surface_plan_draft(
+        replace(illustrative_regional_surface_plan_draft(), request_id="opened-plan")
+    )
+    target = tmp_path / "opened.json"
+    target.write_text(request.to_json(), encoding="utf-8")
+    monkeypatch.setattr(
+        regional_surface_plan_io.QFileDialog,
+        "getOpenFileName",
+        lambda *_args: (str(target), "JSON files (*.json)"),
+    )
+    tab = RegionalSurfacePlanTab()
+    qtbot.addWidget(tab)
+
+    tab.open_button.click()
+
+    assert tab.request_id.text() == "opened-plan"
+    assert tab.file_actions.recent_path == target
+    assert "opened" in tab.status_label.text().lower()
+    assert tab.readback.toPlainText() == request.to_json()
+
+
+def test_pyqt_failed_open_rolls_back_editor_and_recent_path(
+    qtbot, tmp_path: Path, monkeypatch: pytest.MonkeyPatch
+) -> None:  # type: ignore[no-untyped-def]
+    pytest.importorskip("PyQt6")
+    from rate_of_closure.ui.pyqt6 import regional_surface_plan_io
+    from rate_of_closure.ui.pyqt6.regional_surface_plan_tab import (
+        RegionalSurfacePlanTab,
+    )
+
+    target = tmp_path / "corrupt.json"
+    target.write_text('{"request_id":"one","request_id":"two"}', encoding="utf-8")
+    monkeypatch.setattr(
+        regional_surface_plan_io.QFileDialog,
+        "getOpenFileName",
+        lambda *_args: (str(target), "JSON files (*.json)"),
+    )
+    tab = RegionalSurfacePlanTab()
+    qtbot.addWidget(tab)
+    before = tab.draft()
+
+    tab.open_button.click()
+
+    assert tab.draft() == before
+    assert tab.file_actions.recent_path is None
+    assert "open failed" in tab.status_label.text().lower()
+    assert tab.status_label.accessibleName() == "Regional surface plan file error"
+
+
+def test_pyqt_save_as_preserves_imported_request_bytes(
+    qtbot, tmp_path: Path, monkeypatch: pytest.MonkeyPatch
+) -> None:  # type: ignore[no-untyped-def]
+    pytest.importorskip("PyQt6")
+    from rate_of_closure.ui.pyqt6 import regional_surface_plan_io
+    from rate_of_closure.ui.pyqt6.regional_surface_plan_tab import (
+        RegionalSurfacePlanTab,
+    )
+
+    draft = illustrative_regional_surface_plan_draft()
+    request = validate_regional_surface_plan_draft(
+        replace(
+            draft,
+            request_id="saved-plan",
+            regions=(replace(draft.regions[0], precedence=MAX_CANONICAL_SAFE_INTEGER),),
+        )
+    )
+    source = tmp_path / "source.json"
+    destination = tmp_path / "copy.json"
+    source.write_text(request.to_json(), encoding="utf-8")
+    choices = iter(((str(source), "JSON files (*.json)"),))
+    monkeypatch.setattr(
+        regional_surface_plan_io.QFileDialog,
+        "getOpenFileName",
+        lambda *_args: next(choices),
+    )
+    monkeypatch.setattr(
+        regional_surface_plan_io.QFileDialog,
+        "getSaveFileName",
+        lambda *_args: (str(destination), "JSON files (*.json)"),
+    )
+    tab = RegionalSurfacePlanTab()
+    qtbot.addWidget(tab)
+
+    tab.open_button.click()
+    tab.save_button.click()
+
+    assert destination.read_bytes() == source.read_bytes()
+    assert tab.file_actions.recent_path == destination
+    assert tab.current_request() == request
+
+
+def test_pyqt_import_preserves_canonical_precision_and_large_si_values(
+    qtbot,
+) -> None:  # type: ignore[no-untyped-def]
+    pytest.importorskip("PyQt6")
+    from rate_of_closure.ui.pyqt6.regional_surface_plan_tab import (
+        RegionalSurfacePlanTab,
+    )
+
+    draft = illustrative_regional_surface_plan_draft()
+    request = validate_regional_surface_plan_draft(
+        replace(
+            draft,
+            upper_coordinate_m=1_000_000_000_000.0,
+            base_surface=replace(
+                draft.base_surface,
+                firmness_pa=0.00000000001,
+                grass_height_m=123.12345678901,
+                turf_density_kg_m3=20_000.12345678901,
+            ),
+        )
+    )
+    tab = RegionalSurfacePlanTab()
+    qtbot.addWidget(tab)
+
+    tab.apply_imported_request(request)
+
+    assert tab.current_request() is request
+
+
+def test_pyqt_cancelled_open_and_save_leave_status_and_recent_unchanged(
+    qtbot, monkeypatch: pytest.MonkeyPatch
+) -> None:  # type: ignore[no-untyped-def]
+    pytest.importorskip("PyQt6")
+    from rate_of_closure.ui.pyqt6 import regional_surface_plan_io
+    from rate_of_closure.ui.pyqt6.regional_surface_plan_tab import (
+        RegionalSurfacePlanTab,
+    )
+
+    monkeypatch.setattr(
+        regional_surface_plan_io.QFileDialog,
+        "getOpenFileName",
+        lambda *_args: ("", ""),
+    )
+    monkeypatch.setattr(
+        regional_surface_plan_io.QFileDialog,
+        "getSaveFileName",
+        lambda *_args: ("", ""),
+    )
+    tab = RegionalSurfacePlanTab()
+    qtbot.addWidget(tab)
+    status = tab.status_label.text()
+
+    tab.open_button.click()
+    tab.save_button.click()
+
+    assert tab.status_label.text() == status
+    assert tab.file_actions.recent_path is None
diff --git a/tests/rate_of_closure/test_regional_surface_plan_io.py b/tests/rate_of_closure/test_regional_surface_plan_io.py
new file mode 100644
index 000000000..0beb0c557
--- /dev/null
+++ b/tests/rate_of_closure/test_regional_surface_plan_io.py
@@ -0,0 +1,190 @@
+"""Canonical regional surface-plan import/export boundary tests."""
+
+from __future__ import annotations
+
+from dataclasses import replace
+from pathlib import Path
+
+import pytest
+
+from rate_of_closure.application.regional_surface_plan import (
+    EDITOR_PROVIDER_ID,
+    editor_draft_from_regional_surface_plan_request,
+    illustrative_regional_surface_plan_draft,
+    regional_surface_plan_request_for_draft,
+    validate_regional_surface_plan_draft,
+)
+from rate_of_closure.application.regional_surface_plan_files import (
+    read_regional_surface_plan_request,
+    write_regional_surface_plan_request_atomic,
+)
+from shared.python.swing_sim.ground.regional_plan_records import (
+    MAX_REGIONAL_PLAN_WIRE_BYTES,
+)
+from shared.python.swing_sim.ground.regional_plan_wire import (
+    regional_material_plan_request_from_dict,
+)
+
+pytestmark = [pytest.mark.unit, pytest.mark.headless_safe]
+
+
+def test_editor_request_file_round_trip_is_exact_and_deterministic(
+    tmp_path: Path,
+) -> None:
+    request = validate_regional_surface_plan_draft(
+        illustrative_regional_surface_plan_draft()
+    )
+    target = tmp_path / "regional-plan.json"
+
+    assert write_regional_surface_plan_request_atomic(request, target)
+
+    assert target.read_bytes() == request.to_json().encode("utf-8")
+    loaded = read_regional_surface_plan_request(target)
+    assert loaded == request
+    imported = editor_draft_from_regional_surface_plan_request(loaded)
+    assert regional_surface_plan_request_for_draft(imported, loaded) == request
+
+
+def test_cancelled_request_write_is_a_no_op(tmp_path: Path) -> None:
+    request = validate_regional_surface_plan_draft(
+        illustrative_regional_surface_plan_draft()
+    )
+
+    assert write_regional_surface_plan_request_atomic(request, None) is False
+    assert list(tmp_path.iterdir()) == []
+
+
+def test_request_replace_failure_preserves_last_known_good(
+    tmp_path: Path, monkeypatch: pytest.MonkeyPatch
+) -> None:
+    from rate_of_closure.application import atomic_text_files
+
+    request = validate_regional_surface_plan_draft(
+        illustrative_regional_surface_plan_draft()
+    )
+    target = tmp_path / "regional-plan.json"
+    target.write_text("last-known-good", encoding="utf-8")
+    monkeypatch.setattr(
+        atomic_text_files.os,
+        "replace",
+        lambda _source, _target: (_ for _ in ()).throw(OSError("replace failed")),
+    )
+
+    with pytest.raises(OSError, match="replace failed"):
+        write_regional_surface_plan_request_atomic(request, target)
+
+    assert target.read_text(encoding="utf-8") == "last-known-good"
+    assert not list(tmp_path.glob(".*.tmp"))
+
+
+@pytest.mark.parametrize(
+    "text, message",
+    [
+        ("{not-json", "JSON"),
+        ('{"request_id":"one","request_id":"two"}', "[Dd]uplicate"),
+        (" " * (MAX_REGIONAL_PLAN_WIRE_BYTES + 1), "maximum wire size"),
+    ],
+    ids=("corrupt", "duplicate", "oversize"),
+)
+def test_request_read_rejects_corruption_duplicates_and_oversize(
+    tmp_path: Path, text: str, message: str
+) -> None:
+    target = tmp_path / "invalid.json"
+    target.write_text(text, encoding="utf-8")
+
+    with pytest.raises((TypeError, ValueError), match=message):
+        read_regional_surface_plan_request(target)
+
+
+def test_request_read_bounds_a_file_that_grows_after_metadata_check(
+    tmp_path: Path, monkeypatch: pytest.MonkeyPatch
+) -> None:
+    target = tmp_path / "growing.json"
+    target.write_bytes(b"{}")
+    requested_sizes: list[int] = []
+
+    class GrowingBinaryFile:
+        def __enter__(self) -> GrowingBinaryFile:
+            return self
+
+        def __exit__(self, *_args: object) -> None:
+            return None
+
+        def read(self, size: int) -> bytes:
+            requested_sizes.append(size)
+            return b" " * size
+
+    monkeypatch.setattr(
+        Path,
+        "read_text",
+        lambda *_args, **_kwargs: pytest.fail("unbounded read_text used"),
+    )
+    monkeypatch.setattr(Path, "open", lambda *_args, **_kwargs: GrowingBinaryFile())
+
+    with pytest.raises(ValueError, match="maximum wire size"):
+        read_regional_surface_plan_request(target)
+
+    assert requested_sizes == [MAX_REGIONAL_PLAN_WIRE_BYTES + 1]
+
+
+def test_request_read_rejects_invalid_utf8(tmp_path: Path) -> None:
+    target = tmp_path / "invalid-utf8.json"
+    target.write_bytes(b"\xff")
+
+    with pytest.raises(ValueError, match="UTF-8"):
+        read_regional_surface_plan_request(target)
+
+
+def test_import_rejects_non_editor_provenance_without_coercion() -> None:
+    request = validate_regional_surface_plan_draft(
+        illustrative_regional_surface_plan_draft()
+    )
+    payload = request.to_dict()
+    payload["provenance"]["producer"] = "external.course.authority"
+    external = regional_material_plan_request_from_dict(payload)
+
+    with pytest.raises(ValueError, match="editor producer"):
+        editor_draft_from_regional_surface_plan_request(external)
+
+    payload = request.to_dict()
+    payload["provenance"]["input_sha256"] = "0" * 64
+    mismatched = regional_material_plan_request_from_dict(payload)
+    with pytest.raises(ValueError, match="digest does not match"):
+        editor_draft_from_regional_surface_plan_request(mismatched)
+
+
+def test_import_rejects_editor_provider_with_unsupported_axis() -> None:
+    request = validate_regional_surface_plan_draft(
+        illustrative_regional_surface_plan_draft()
+    )
+    payload = request.to_dict()
+    payload["axis_unit"] = [0.0, 0.0, 1.0]
+    rotated = regional_material_plan_request_from_dict(payload)
+
+    with pytest.raises(ValueError, match="axis qualification"):
+        editor_draft_from_regional_surface_plan_request(rotated)
+
+
+def test_edited_import_rebinds_provenance_instead_of_reusing_stale_digest() -> None:
+    request = validate_regional_surface_plan_draft(
+        illustrative_regional_surface_plan_draft()
+    )
+    imported = editor_draft_from_regional_surface_plan_request(request)
+    edited = replace(imported, request_id="edited-plan")
+
+    rebound = regional_surface_plan_request_for_draft(edited, request)
+
+    assert rebound.request_id == "edited-plan"
+    assert rebound.provenance.producer == EDITOR_PROVIDER_ID
+    assert rebound.provenance.input_sha256 != request.provenance.input_sha256
+
+
+def test_editor_rejects_unsafe_integer_valued_number() -> None:
+    draft = illustrative_regional_surface_plan_draft()
+    unsafe = replace(
+        draft,
+        base_surface=replace(draft.base_surface, firmness_pa=10_000_000_000_000_000.0),
+    )
+
+    with pytest.raises(ValueError, match="cross-runtime safe range"):
+        validate_regional_surface_plan_draft(unsafe)
diff --git a/tests/rate_of_closure/test_workspace_files.py b/tests/rate_of_closure/test_workspace_files.py
index a7dcd85d3..5e260af4a 100644
--- a/tests/rate_of_closure/test_workspace_files.py
+++ b/tests/rate_of_closure/test_workspace_files.py
@@ -70,13 +70,15 @@ def fail_serialization(_document: WorkspaceDocument) -> str:
 def test_replace_failure_preserves_existing_and_removes_temporary_file(
     tmp_path: Path, monkeypatch: pytest.MonkeyPatch
 ) -> None:
+    from rate_of_closure.application import atomic_text_files
+
     target = tmp_path / "showcase.json"
     target.write_text("last-known-good", encoding="utf-8")
 
     def fail_replace(_source: str | Path, _target: str | Path) -> None:
         raise OSError("replace failed")
 
-    monkeypatch.setattr(workspace_files.os, "replace", fail_replace)
+    monkeypatch.setattr(atomic_text_files.os, "replace", fail_replace)
 
     with pytest.raises(OSError, match="replace failed"):
         workspace_files.write_workspace_atomic(_document("New"), target)