[AdaWorldAPI]

D-QR-CAP: five-PR architecture / experimental roadmap (docs only)

Capstone mapping choice: existing lance_graph_contract::class_view::{ClassView,FieldMask,WideFieldMask} becomes the candidate structural field-mapping basis adjacent to Quack's bind-once numeric boundary. This does not make a field-presence bit an edge, permission, or identity witness. The capstone seeks fewer handrolled edges by expressing a typed relationship once and lowering it onto borrowed ordinal lanes and optional Boolean QR tiles.

Documents

• Relational ISA / ClassView capstone
• Cross-consumer independent-oracle and falsifier matrix

Four independently reviewable companion draft PRs

Consumer	PR	Board issue	Initial oracle
SAP/CATS	lance-graph #1661	EXO-122	existing CATS/BAPI/hashing
IAM/dir-sim	lance-graph #1662	EXO-123	existing dir-sim View/actors
HubSPO-rs	HubSPO #47	EXO-124	live/list masks plus synthetic association fixtures until D-STO-3 exists
stockfish-rs	stockfish #23	EXO-125	Shakmaty movegen + exact NNUE full-refresh oracle

Capstone board: EXO-121. Coordination: EXO-50.

Geometry and gates

• FieldMask: existing 64-bit structural presence. WideFieldMask: zero-alloc low 64 bits, wider 64-bit word chunks for up to 256 u8-indexed field positions. Existing ClassView::project still takes FieldMask; no silent trait-widening claim.
• QR64 = 64×64 Boolean = 512 B; QR256 = 256×256 Boolean = 8 KiB. These exclude borrowed SoA, ordered off/pos mapping, sidecar and numeric values.
• Method8×Slot8 unary palette geometry is NOT Source8×Target8 adjacency. GF(2) XOR, Boolean OR/AND and NNUE wrapping/saturating int math have separate laws.
• Reuse #1659 physical borrowed lanes / order+prefetch and #1660 algebra/rotation/proof research; neither is modified here. EXO-119 owns the proposed PowerShell ordinal-template lens.
• ADMISSION B #1640 remains binding, including preparation. No cross-table value-steered position reads. ACL/authority and source identity/version witnesses stay distinct from the ClassView field mask.
• No production code, tests or benchmarks in this PR; this is a preregistered plan, not evidence of speedup. No main branch modifications and no merge authorization.

Acceptance

Each topic must supply independently checked correctness, negative/refusal/mutation tests, evidence that zero-copy survives, no invented schema/permission, separate cold/hot/materialization costs, and ADOPT / KEEP_DIRECT / REJECT / DEFER. A negative ROI on QR staging is an acceptable result. Graph edge generation is unnecessary where it can be derived from an admitted law and witnessed inputs.

Horizontal ISA/PART_OF ↔ proposed vertical HHTL16 zipper extension (operator hypothesis)

Additional capstone appendix: D-QR-ABI-0 detailed rails/ABI mapping. It cross-checks existing FacetCascade 16-byte facet_classid(4) + 6×(hi8:lo8) against two distinct proposed 16×8-bit ISA / PART_OF logical vertical zipper views. The present HHTL NiblePath 16-nibble router uses 4-bit steps, and must not be mislabeled as an already-persisted HHTL16 16×8 representation. Local u8 rail positions require witnessed codebook/identity mapping; a 6×2×8-bit address does not replace the QR256 8-KiB Boolean truth plane. This is a testable mapping hypothesis, not an adopted format.

The four consumer plans now contain their own explicit typed-rail negative oracles. Board decisions and unresolved implementation handoff remain EXO-121, EXO-119, EXO-120, and EXO-50.

Additional capstone rail mapping note

The capstone branch also contains D-QR-ABI-0: ClassView/WideFieldMask and ISA/PART_OF horizontal rails, a separate additional documentation commit present when the branch was verified (d66cb0b1). It maps existing 6×(hi:u8, lo:u8) FacetCascade rails and proposes future ISA/PART_OF HHTL16 zipper views without claiming they are already ABI-compatible or physically stored. The mapping and all rail laws remain experimental; no new persistent encoding or hardware ISA is approved. There are three docs in this capstone branch, no source-code changes.

[coderabbitai]

Important

• 🔍 Trigger review

This repository does not receive automatic reviews because it has fewer than 10 stars.

⚙️ Run configuration

• Configuration used: Organization UI
• Review profile: CHILL
• Plan: Advanced
• Run ID: 37c71fb2-ce57-4331-ad41-631e02f5f848

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Warning

Billing warning: we have not been able to collect payment for this subscription for more than 72 hours. Please update the payment method or pay any pending invoices in Billing to avoid service interruption.

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-10-21T13:16:21.060724Z	fb4941a	Draft marked ready

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.