scriptoriumadmin/hypertwist
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Backfill retained web content provenance board

Browse source
This commit is contained in:
axiomlogicnexus 2026-05-27 07:46:46 +02:00
parent 9d28442b59
commit ce95754057
3 changed files with 152 additions and 3 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

16
docs/HT_REPO_INCORPORATION_AUDIT_2026-05-11.md
View file

@ -306,8 +306,13 @@ These are not part of the current 71-repo shallow-eval set, but they are live in
| Source | Current surface | Count signal | Notes |
| --- | --- | ---: | --- |
| `SpeedCubeDB` | retained web content | 185 runtime bundles / 185 direct donor deck specs | Live retained-web-content incorporation surface |
| `CubingApp` | retained web content | 25 direct donor deck specs | Live retained-web-content incorporation surface |
| `SpeedCubeDB` | retained web content | 185 runtime bundles / 185 direct donor deck specs | Live retained-web-content incorporation surface; broad advanced and multi-puzzle algorithm corpus beneath the first-party catalog layer |
| `CubingApp` | retained web content | 25 direct donor deck specs | Live retained-web-content incorporation surface; narrower one-handed and shaped-subset adjunct beneath the first-party catalog layer |
Use
[HYPERTWIST_TRAINING_CATALOG_RETAINED_WEB_CONTENT_PROVENANCE_BOARD_2026-05-27.md](C:/HyperTwist/docs/HYPERTWIST_TRAINING_CATALOG_RETAINED_WEB_CONTENT_PROVENANCE_BOARD_2026-05-27.md:1)
for the current first-party owner split and the exact retained-web-content
grounding role of these two live catalog sources.
## Current saveability view
@ -370,7 +375,12 @@ That does not make them legally cleared or deeply audited. It only means this pa
## Recommended next implementation slice
The next clean bounded move is to preserve the twenty landed lanes as the current truth, keep the `onionhoney/roux-trainers` clean-room status explicit everywhere, keep the landed `cubing/cubing.js` `MPL` boundary explicit everywhere, keep the landed `cutelyaware/magiccube4d` attribution/provenance boundary explicit everywhere, keep the landed `google/model-viewer/packages/shared-assets` allowlist/provenance boundary explicit everywhere, keep the landed `PostHog/posthog` root-`MIT`-core-versus-`ee/` subtree boundary explicit everywhere, keep the landed `screenpipe/screenpipe` permissive-core-versus-`ee/` subtree boundary explicit everywhere, keep the landed `remotion-dev/remotion` commercial-license and package-split boundary explicit everywhere, and continue from the closed retained-set packet sequence with `Phase 5R-A` for `cubing/alg.js`, while leaving the optional `Phase 3R-G` browser comparison lane deferred unless the landed primary browser spatial stack exposes a real gap. After that, do a second pass focused on the additional live non-71 surfaces (`SpeedCubeDB` and `CubingApp`) so the active training catalog has a complete provenance board before any further widening work resumes.
The training-catalog provenance-board pass for the additional live non-71
surfaces (`SpeedCubeDB` and `CubingApp`) is now complete. The next clean
bounded move is not another trainer-shell widening by default; it is a
source-backed pass on any remaining live non-repo or reference-incorporation
surface that still feeds first-party materialized catalog outputs without an
equally explicit provenance board.
For the readable all-rows follow-up board that pairs this audit with the reset schedule, use:

137
docs/HYPERTWIST_TRAINING_CATALOG_RETAINED_WEB_CONTENT_PROVENANCE_BOARD_2026-05-27.md Normal file
View file

@ -0,0 +1,137 @@
# HyperTwist Training Catalog Retained Web Content Provenance Board - 2026-05-27
## Status
This document is the source-backed provenance and hierarchy note for the two
additional live non-71 training-catalog surfaces already materialized in
HyperTwist:
- `SpeedCubeDB`
- `CubingApp`
These are not repo-row lanes in the current retained-set manifest.
They are retained web-content sources already transformed into first-party
training catalog surfaces.
It answers one narrow question:
- do these retained web-content sources require reopening already landed
first-party training/coaching implementation, or do they instead need a
provenance-board backfill that makes their exact catalog role explicit
Result:
- no product-code reopening is required
- yes provenance-board and feature-canon backfill is required
## Source basis
- `C:\HyperTwist\docs\HT_REPO_INCORPORATION_AUDIT_2026-05-11.md`
- `C:\HyperTwist\docs\v6_5_deep_manual_pack\HyperTwist\FEATURE_REGISTRY.md`
- `C:\HyperTwist\UnrealHyperTwist\Content\HyperTwistTraining\MaterializedCatalog\hypertwist-deck-import-specs.json`
- `C:\HyperTwist\UnrealHyperTwist\Content\HyperTwistTraining\MaterializedCatalog\hypertwist-runtime-content-bundles.json`
- `C:\HyperTwist\UnrealHyperTwist\Content\HyperTwistTraining\MaterializedCatalog\hypertwist-content-pack-manifests.json`
- `C:\HyperTwist\UnrealHyperTwist\Content\HyperTwistTraining\MaterializedCatalog\raw-extracts\httrack-retained-exercises.jsonl`
- `C:\HyperTwist\UnrealHyperTwist\Content\HyperTwistTraining\MaterializedCatalog\hypertwist-reference-incorporation-targets.json`
## Counts confirmed from current first-party materialization
- `SpeedCubeDB`: `185` direct donor deck specs and `185` runtime bundles
- `CubingApp`: `25` direct donor deck specs and `25` runtime bundles
Track distribution currently materialized:
- `SpeedCubeDB`
- `advanced-3x3-systems`: `70`
- `big-cube-systems`: `6`
- `megaminx-systems`: `66`
- `pyraminx-systems`: `8`
- `skewb-systems`: `20`
- `square-1-systems`: `15`
- `CubingApp`
- `one-handed-systems`: `8`
- `pyraminx-systems`: `1`
- `skewb-systems`: `3`
- `square-1-systems`: `13`
## Slice-local verdict
| Slice | Primary authority | Secondary value kept | Result |
|---|---|---|---|
| First-party rewritten course layer, materialized catalog packs, and live training-shell/runtime ownership | first-party current code and course layer | retained web-content sources stay subordinate provenance-grounding surfaces | unchanged live implementation owner |
| Broad advanced and multi-puzzle retained algorithm corpus across advanced `3x3`, big cube, megaminx, broad pyraminx, broad skewb, and broad square-1 catalog families | `SpeedCubeDB` retained web content | `CubingApp` remains the narrower one-handed and shaped-subset adjunct | unchanged retained broad catalog corpus ground beneath first-party packs |
| Narrow one-handed and shaped-subset adjunct corpus across one-handed `CMLL`, Sarah-intermediate trigger ladders, pyraminx last-layer, and square-1 `CSP` / `OBL` family subsets | `CubingApp` retained web content | `SpeedCubeDB` remains the broader adjacent advanced corpus | unchanged retained narrow adjunct corpus ground beneath first-party packs |
## Why no product-code reopening is required
HyperTwist already owns the live runtime and course-shell layer above both web
sources:
- first-party content-pack manifests
- first-party deck import specs
- first-party rewritten track goals and course wording
- materialized runtime bundle structure
The two retained web-content sources ground catalog breadth, not trainer-shell
ownership.
That means:
- no first-party training-session runtime reopen is needed
- no coaching/review-plan/follow-up reopen is needed
- no retained repo-row hierarchy needs reopening from this pass
## Current exact grounded role
### `SpeedCubeDB`
Retained as the broad advanced and multi-puzzle algorithm corpus beneath the
first-party catalog layer.
Its currently materialized footprint includes:
- dense `1LLL` and adjacent advanced `3x3` families
- big-cube parity and reduction-adjacent families
- megaminx `OLL` / `PLL` breadth
- broad skewb `Sarah Advanced` families
- broad square-1 and pyraminx deck grounding
It is not the owner of HyperTwist's trainer-shell runtime, review-plan logic,
or coaching orchestration.
### `CubingApp`
Retained as the narrower shaped-subset and one-handed adjunct beneath the
first-party catalog layer.
Its currently materialized footprint includes:
- one-handed `CMLL` family splits
- pyraminx last-layer adjunct grounding
- skewb `Sarah Intermediate` trigger ladders
- square-1 `CSP` / `OBL` shaped subset ladders
It is not the owner of the broader imported algorithm-corpus layer, and it is
not the owner of HyperTwist's trainer-shell runtime.
## Backfill result
### Earlier incorporation audit reopened
- `C:\HyperTwist\docs\HT_REPO_INCORPORATION_AUDIT_2026-05-11.md`
### Feature canon reopened
- `C:\HyperTwist\docs\v6_5_deep_manual_pack\HyperTwist\FEATURE_REGISTRY.md`
## Final call
`SpeedCubeDB` and `CubingApp` stay as live retained web-content provenance
grounds inside the current first-party training catalog.
`SpeedCubeDB` is the broader advanced and multi-puzzle corpus ground.
`CubingApp` is the narrower one-handed and shaped-subset adjunct ground.
Neither one reopens the already landed first-party training/coaching runtime.

2
docs/v6_5_deep_manual_pack/HyperTwist/FEATURE_REGISTRY.md
View file

@ -199,6 +199,8 @@ repo.
| Feature | Status | Primary authority | Notes |
|---|---|---|---|
| Imported training catalog packs | Implemented now | first-party current code | Already in product. |
| Retained advanced and multi-puzzle algorithm corpus grounding | Implemented now | `SpeedCubeDB` retained web content + first-party current code | Current live imported catalog breadth across advanced `3x3`, big cube, megaminx, broad pyraminx, broad skewb, and broad square-1 families remains grounded in retained `SpeedCubeDB` web content beneath the first-party rewritten course layer. This does not displace the first-party training runtime or the narrower `CubingApp` one-handed/shaped-subset adjunct. |
| Retained one-handed and shaped-subset adjunct corpus grounding | Implemented now | `CubingApp` retained web content + first-party current code | Current live imported adjunct depth across one-handed `CMLL`, Sarah-intermediate skewb triggers, pyraminx last-layer, and square-1 `CSP` / `OBL` families remains grounded in retained `CubingApp` web content beneath the first-party rewritten course layer. This does not displace the broader `SpeedCubeDB` corpus ground or the first-party training runtime. |
| Hypercubing knowledge/reference bundle | Implemented now | landed knowledge packet + first-party consumption | Live knowledge family. |
| Curated training atlas / wiki-like memory | Deep-source grounded retained | first-party target + memory doctrine | Not yet source-grounded enough for stronger external-lane ownership. |