docs(D-SAG-6-1): reference ledger and stale-status corrections for the D-SAG-6 audit

.claude/board/INTEGRATION_PLANS.md

@@ -1,3 +1,7 @@
+## 2026-10-28 — D-SAG-6-1 reference ledger and stale-status corrections → `.claude/plans/d-sag-6-1-audit-reference-ledger-v1.md`
+
+**Status:** PROPOSED, docs only; prerequisite reading for D-SAG-6-1 (#1760), not the audit. **Board:** [entry](entries/2026-10-28-d-sag-6-1-audit-reference-ledger.md). Lists the references #1760 §7 lacks (ruff `ruff_spo_triplet` files and plans, `ogar-from-ruff`, CFA #1813/#1814/#1818, #1792, #1799, contract files, D-DTM-1, ndarray primitives, crai-rs CR-H0/H1) and corrects stale status (crai-rs lists #1813–#1830 as open; ruff has no Rust frontend; `ProposedMethod` is not defined in code).
+**Confidence:** reference ledger only; D-SAG-6-1 unstarted.
 ## 2026-10-28 — D-DT64-6: DR64 routing, DT64 bind store, per-cell WHEN×ON×DO behaviour, OGIT reference predicates → `.claude/plans/d-dt64-6-routing-bindstore-behavior-v1.md`
 
 **Status:** PROPOSED, docs only; nothing authorized. **Board:** [entry](entries/2026-10-28-d-dt64-6-routing-bindstore-behavior.md), [STATUS_BOARD](STATUS_BOARD.md) D-DT64-6. Records eight user decisions (three DT64 levels; primary DR#3 Identity × Properties; WHEN×ON×DO per DR#3 cell; DR64 = direct routing, DT64 = DTO bind store; immutable classid → behaviour via hotplug; global needs no classid; OGIT read unpinned; DR64 references existing identity and the global→local prefetch holds a frame-keyed register copy) and stages D-DT64-6-0..5 with 6-1a (classify the verb files both readers miss: lance-graph accepts 12, OGAR 511, of 523), OGIT reader reconciliation, one `(src, verb, dst)` table from the entity `ogit:allowed` blocks with no second parser, DT64 predicate registration, consumer HotPlug access. Requests an amendment to D-DT64-Q0's "One-sentence contract" (persistent binding store); not applied.

.claude/board/entries/2026-10-28-d-sag-6-1-audit-reference-ledger.md

@@ -0,0 +1,13 @@
+# D-SAG-6-1: reference ledger and stale-status corrections for the D-SAG-6 audit
+
+**Status:** PROPOSED / DOCS ONLY. Plan: [d-sag-6-1-audit-reference-ledger-v1.md](../../plans/d-sag-6-1-audit-reference-ledger-v1.md). Parent: D-SAG-6 (#1760). No code, test or authorization; the D-SAG-6-1 audit itself is still unstarted.
+
+## What this records
+
+- References the D-SAG-6-1 audit needs that #1760 §7 does not list: ruff `ruff_spo_triplet` (`ir.rs`, `expand.rs`, `reassemble.rs`, `role.rs`), ruff's behavioural-ore and R2IL oracle plans, OGAR `ogar-from-ruff`, the merged CFA docs (#1813, #1814, #1818), #1792, #1799, `spog_tenants.rs`, `epistemic_state5.rs`, `lance-graph-cypher-quack`, D-DTM-1, the `mcp.exo.red` workbench entry, the ndarray mask/popcount primitives for the first fixture, and crai-rs CR-H0/CR-H1 (private).
+- MEASURED: ruff `f2cb523` has SPO frontends for C++, C#, Python, Ruby, MeTTa and SQLAlchemy, none for Rust; `ProposedMethod` has zero hits under lance-graph `crates/`.
+- Stale status corrected: crai-rs CR-H0 lists #1813–#1830 as open (all merged); #1831 calls crai-rs the successor to #1760 (crai-rs is a sidecar that waits on D-SAG-6-1).
+
+## Open
+
+Whether D-SAG-6-1 proposes a Rust frontend for ruff; where `ProposedMethod` is defined before D-SAG-6-2 needs it.

.claude/board/entries/README.md

@@ -25,10 +25,11 @@ index row, (3) no duplicate entry id. Checks 1 and 2 are deliberately
 opposite directions; the stranding this convention prevents shows up in
 exactly one of them, never both.
 
-489 entries, 2026-08-06 .. 2026-10-28.
+490 entries, 2026-08-06 .. 2026-10-28.
 
 | date | entry id | finding | file |
 |---|---|---|---|
+| 2026-10-28 | `D-SAG-6-1` |  | [2026-10-28-d-sag-6-1-audit-reference-ledger.md](2026-10-28-d-sag-6-1-audit-reference-ledger.md) |
 | 2026-10-28 | `D-DT64-6` |  | [2026-10-28-d-dt64-6-routing-bindstore-behavior.md](2026-10-28-d-dt64-6-routing-bindstore-behavior.md) |
 | 2026-10-27 | `harvest-integration-composite-key-and-simd-scope` |  | [2026-10-27-harvest-integration-composite-key-and-simd-scope.md](2026-10-27-harvest-integration-composite-key-and-simd-scope.md) |
 | 2026-10-26 | `D-ONNX-INTAKE-0` |  | [2026-10-26-onnx-frontend-intake-bakeoff.md](2026-10-26-onnx-frontend-intake-bakeoff.md) |

.claude/plans/d-sag-6-1-audit-reference-ledger-v1.md

@@ -0,0 +1,51 @@
+# D-SAG-6-1: reference ledger and stale-status corrections for the D-SAG-6 follow-up audit
+
+**Date:** 2026-10-28. **Status:** PROPOSED / DOCS ONLY. **Parent:** [D-SAG-6](d-sag-6-rust-codegraph-math-logic-tracert-v1.md) (#1760), stage D-SAG-6-1 (the read-only capability audit). **Not:** the audit itself, and no code, test, ABI or implementation authorization.
+
+D-SAG-6 S0 (#1760, merged 2026-10-24) lists its entry points in §7. Since then, #1813–#1831 and the private crai-rs CR-H0/CR-H1 docs have added material the D-SAG-6-1 audit must read, and some of it now carries stale status. This plan records both, so the audit does not start from #1760's ledger alone. It adds nothing to D-SAG-6's design.
+
+Read on 2026-10-28 at lance-graph `main` `eeaaafb`, ruff `main` `f2cb523`, OGAR `main` `6b43a48`, ndarray `master` `38dd52e`, crai-rs `main` `378a24a`. Evidence states: VERIFIED-IN-CODE (read here), CURRENT-CONTRACT (merged doc), MEASURED (command given), OPEN.
+
+## 1. References D-SAG-6-1 needs that #1760 §7 does not list
+
+| Reference | Why the audit needs it | State |
+|---|---|---|
+| ruff `crates/ruff_spo_triplet/src/ir.rs` (`ModelGraph`) | the structural model Ruff hands over; data, not execution semantics | VERIFIED-IN-CODE (exists) |
+| ruff `crates/ruff_spo_triplet/src/expand.rs` | sorted, deduplicated SPO triples: event order and multiplicity are lost, so this view alone cannot separate two programs that differ only in effect order (D-SAG-6 F05) | VERIFIED-IN-CODE (exists); behaviour as documented in ruff and crai-rs CR-H0 |
+| ruff `crates/ruff_spo_triplet/src/reassemble.rs` | inverse projection for C++ only, not a general reverse oracle | VERIFIED-IN-CODE (exists) |
+| ruff `crates/ruff_spo_triplet/src/role.rs` (`PredicateRole`) | the 13 role bits (DESCRIBES … OPAQUE) that #1814 quotes as READS, DERIVES, EFFECTS, REQUIRES, AUTHORIZES, TRIGGERS | VERIFIED-IN-CODE |
+| ruff `.claude/plans/behavioral-ore-v1.md` | the separate ordered event stream (`events.tsv`, never deduplicated); a C++ research leg, not a Rust oracle | CURRENT-CONTRACT (ruff plan) |
+| ruff `.claude/plans/r2il-roundtrip-oracle-spec-v1.md` | states that SPO is not the behavioural parity oracle | CURRENT-CONTRACT (ruff plan) |
+| ruff `docs/OGAR-POLYGLOT-AST-INTEGRATION.md` | "structure transpiles; behavior does not": an RFC, not a finished round trip | CURRENT-CONTRACT (ruff doc) |
+| ruff has no Rust SPO frontend | frontends exist for C++, C#, Python, Ruby, MeTTa and SQLAlchemy (`crates/ruff_{cpp,csharp,python,ruby,metta,sqlalchemy}_spo`), none for Rust. D-SAG-6's "Ruff structural crosswalk" therefore has no Ruff input for Rust source today; the audit must record this as MISSING, not as a coverage gap of an existing arm | MEASURED: `ls crates \| grep spo` on ruff `f2cb523` |
+| OGAR `crates/ogar-from-ruff` (`lift_model_graph`, `lift_actions`) | the existing Ruff → OGAR bridge; the audit's "smallest already-owned location for a read-only projection adapter" must be compared against it | VERIFIED-IN-CODE (crate exists; functions named in its own docs) |
+| `docs/research/counterfactual-behavior/cfa-1-typed-incidence.md` (#1814) | ON×WHEN×DO over Ruff/OGAR/rust-analyzer records; "Ruff is structural crosswalk, not a Rust MIR oracle"; WHEN is a K3 guard | CURRENT-CONTRACT (merged research doc) |
+| `docs/research/counterfactual-behavior/cfa-5-compiler-analysis.md` (#1818) | asks where a behavioural projection binds `ProposedMethod` without a duplicate type; cites #1760, #1747, #1792 | CURRENT-CONTRACT |
+| `docs/research/counterfactual-behavior/README.md` and `thought-dependency-scratchpad.md` (#1813) | name D-SAG-6 as the source-harvesting and fixed-point owner | CURRENT-CONTRACT |
+| [#1792](https://github.com/AdaWorldAPI/lance-graph/pull/1792) | escaped-`Keep` refusal: a mask may not be reused outside its borrowed scope (D-SAG-6 F07) | CURRENT-CONTRACT |
+| [#1799](https://github.com/AdaWorldAPI/lance-graph/pull/1799) and the DT64 plans | DT64 versioned bindings, which crai-rs places between Ruff and the D-SAG-6 projection | CURRENT-CONTRACT |
+| `crates/lance-graph-contract/src/spog_tenants.rs` | the graph of an address comes from its classid, not a stored fourth column | VERIFIED-IN-CODE (exists) |
+| `crates/lance-graph-contract/src/epistemic_state5.rs` | CausalEdge64 readings need declared meaning and evidence; a syntax edge named `calls` is not causal evidence | VERIFIED-IN-CODE (exists) |
+| `crates/lance-graph-cypher-quack/src/lib.rs` | experimental Cypher → Quack with a few supported shapes; other patterns are typed refusals | VERIFIED-IN-CODE (exists) |
+| `.claude/plans/d-dtm-1-distributed-thought-delta-reasoning-interchange-v1.md` | "D-SAG-6 owns codegraph"; lists D-SAG-6 as the source-code reasoning Tracert owner | CURRENT-CONTRACT |
+| `.claude/board/entries/2026-10-27-harvest-integration-composite-key-and-simd-scope.md` | records the read-only workbench at `mcp.exo.red` as building on D-SAG-6 | CURRENT-CONTRACT (board entry) |
+| ndarray `src/simd.rs` re-exports: `*_to_mask` predicates, `popcount_raw`, `mask_ternlog_popcount`, `masked_group_count_u32*` | the concrete primitives for D-SAG-6 §4's first fixture, `count(filter(P, X))` versus `popcount(mask(P, X))` | VERIFIED-IN-CODE (re-exported names) |
+| crai-rs `docs/CR-H0-*.md`, `docs/CR-H1-*.md` (private) | negative-mold falsifiers NEG-01..10, CR-F01..18 and X-01..15; several restate F01, F04, F06, F07, F11 and must be reconciled, not duplicated | CURRENT-CONTRACT (other repo, read in full) |
+
+## 2. Corrections to stale or overstated status
+
+| Where | What it says | What is true on 2026-10-28 |
+|---|---|---|
+| crai-rs `CR-H0-CHASSIS-RETROFIT-MAP.md` §2, `CR-H0-SOURCE-LEDGER.md` | #1813–#1830 are "currently open docs proposals" | all eighteen are merged into lance-graph `main` as docs; still not implementations |
+| #1831 plan §2 and board entry | crai-rs CR-H0/CR-H1 is "the successor to #1760" | crai-rs keeps D-SAG-6 as the motor and forbids a new Codegraph owner in crai-rs (CR-H0 §7); its CR-H2 waits on the D-SAG-6-1 audit. crai-rs is a sidecar that consumes D-SAG-6, and D-SAG-6-1 remains unstarted in lance-graph |
+| #1760 §0 and §3 | Ruff "may independently corroborate supported source structure" | true for the six languages above; for Rust there is no Ruff frontend, so there is nothing for Ruff to corroborate yet |
+| #1760 §0, prompt "do not … invent new … ProposedMethod" | implies a `ProposedMethod` type exists to reuse | `ProposedMethod` appears in plans (SAG capstone, cfa-5) but has no definition in `crates/` (MEASURED: zero hits under `crates/`); the audit must treat it as a planned shape, not an existing type |
+
+## 3. What the audit does with this
+
+The D-SAG-6-1 deliverable stays as #1760 §6 defines it: a SHIPPED / LAB / PROPOSED / MISSING capability map and the smallest read-only adapter recommendation. This ledger adds two required rows to that map, "Ruff Rust frontend: MISSING" and "`ProposedMethod` type: PROPOSED only", and one required comparison, against `ogar-from-ruff`. The audit must also reconcile crai-rs's falsifier sets with F01–F15 rather than adopt a second numbering.
+
+## 4. Open
+
+1. Whether the audit should propose a `ruff_rust_spo` frontend or treat rust-analyzer / rustc MIR as the only Rust source of structure (#1760 already names them as the semantic sources).
+2. Where `ProposedMethod` will be defined, and by which owner, before D-SAG-6-2 needs it.