@@ -1,3 +1,11 @@
+## 2026-10-12 (4) — software-awareness-graph-v1 — the software's own methods, programs and evidence as graph rows → `.claude/plans/software-awareness-graph-v1.md`
+
+**Status:** MEASURED for one concept's methods (D-SAG-1-0..3 and D-SAG-1-5 in the PR); D-SAG-1-4 and D-SAG-1-6 queued.
+
+- A `MethodTable` projects to typed rows (class, method, field, condition, fold, law, program, primitive, lowering, refusal, claim, capability, test). Each row has a tier (declared / derived / compiled / verified / hypothesis) and a provenance.
+- The edge table is a Quack table: dependency questions are Drafts lowered and run on mask-risc. Validation is the compiler's verdicts recorded as edges.
+- H1 (fuse a predicate into `MaskedSumI32`) comes from the rows. Measured gap: 1.6–1.8× against a single-pass scalar reference on two tiers. Missing: `Pred` fusion and a sum `FusedFold`.
+- First blocker: multi-hop questions need a relation carrier, not a mask fed between programs.
 ## 2026-10-12 (3) — gremlin-composable-algebra-v1 — Gremlin steps as typed building blocks of ClassId methods → `.claude/plans/gremlin-composable-algebra-v1.md`
 
 **Status:** MEASURED for one population (D-CMH-G1-0..3 shipped in the PR); D-CMH-G1-4..5 queued.
 
@@ -1,3 +1,17 @@
+## D-SAG-1 — Software-awareness graph (2026-10-12)
+
+Plan: `.claude/plans/software-awareness-graph-v1.md`. No new opcode, Quack variant, engine or registry. Nothing applies a transformation.
+
+| D-id | scope | status | gate / falsifier |
+|---|---|---|---|
+| D-SAG-1-0 | inventory: Bible work (6-lane fixture, English KJV; no 7-language graph in code), OGAR identities, dir-sim provenance, Quack/mask-risc | In PR | every row cites code or says documented-only |
+| D-SAG-1-1 | typed schema + projection of prepared methods with tiers and provenance | In PR | `tests/awareness.rs` projection test; one identity, one kind |
+| D-SAG-1-2 | queries A–D through Quack; compiler verdicts as edges | In PR | answers == the programs; falsifiers 1–10, each disable-verified |
+| D-SAG-1-3 | H1 candidate + tier-labelled measurement; MDL measure with counterexample | In PR | `examples/awareness_loop.rs`; MDL test |
+| D-SAG-1-4 | multi-hop dependency via relation carrier / factored `EqU32Via` | Queued (needs D-CMH-G1-4) | two-hop answer == closure computed from the programs |
+| D-SAG-1-5 | G1 `CALLS` edges + method-reference refusals | In PR | `method_references_are_calls_edges_and_cycles_are_the_compilers`; callers via Quack; `MethodCycle` recorded, never re-derived |
+| D-SAG-1-6 | persist as a Lance dataset; source revision from git; canonical body digest | Queued | reload equals projection; toolchain change keeps digests |
+
 ## D-CMH-G1 — Gremlin as a composable ClassId algebra (2026-10-12)
 
 Plan: `.claude/plans/gremlin-composable-algebra-v1.md`.
 
@@ -0,0 +1,17 @@
+# D-SAG-1 — the software's own methods as graph rows (2026-10-12)
+
+**Status:** MEASURED (one concept, one hop). Plan: `.claude/plans/software-awareness-graph-v1.md`.
+
+- **Projection.** `crates/lance-graph-glove-parity/src/awareness.rs` projects a `MethodTable` into typed node and edge rows. Every edge carries a tier and a provenance. The edge table is a Quack table, queried by `Draft` → `lower` → mask-risc, landing on the owning method in one program.
+- **Evidence chain.** Under SAP's real claims the graph explains `billable_hours`' refusal down to `claim:hours_logged:hypothesized`. `FirstOfRun` counts as demonstrated only once the compiler's `EvidenceNotIndependent` refusal is projected. IAM and SAP needs stay hypotheses.
+- **H1.** The tiled `MaskedSumI32` program measures 1.6–1.8× a single-pass scalar reference:
+  - native: 603.5/379.1 µs and 619.6/362.8 µs;
+  - config-v3: 617.4/347.6 µs;
+  - 1,048,576 rows.
+
+  A fused mask-risc terminal is absent; `fused_ternlog` declines `Pred` and `FusedFold` is `Count`/`Any` only.
+- **MDL.** Factoring one shared condition never shortens the description, so G1's `is_billable` is a semantic reuse, not a syntactic one. The first metric dropped a composite's atoms from the vocabulary, and the label-vs-identity counterexample caught it.
+- **Disable runs.** 14 runs, all red. One disable first stayed green; the missing test case then exposed a kind-confusion defect in `claim` (fixed).
+- **Blocker.** Multi-hop questions need a relation carrier or a factored `EqU32Via`, not a mask passed between programs.
+
+OPEN: method identity is caller-issued (OGAR mints none); the body digest is build-local.
@@ -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.
 
-315 entries, 2026-08-06 .. 2026-10-12.
+316 entries, 2026-08-06 .. 2026-10-12.
 
 | date | entry id | finding | file |
 |---|---|---|---|
+| 2026-10-12 | `D-SAG-1` |  | [2026-10-12-software-awareness-graph.md](2026-10-12-software-awareness-graph.md) |
 | 2026-10-12 | `opencog-pln-entropy-research-map` | Source inventory and falsifiable PLN/ECAN/Shannon/Moore attention comparison: entropy target, evidence overlap, graph projection, resource fairness | [2026-10-12-opencog-pln-entropy-research-map.md](2026-10-12-opencog-pln-entropy-research-map.md) |
 | 2026-10-12 | `nars-self-evidence-literature` | NARS internal trace versus evidence, Jøsang dependent/independent fusion and fission, and the snapshot-stable fold assumption; source-grounded red tests for merged #1496 | [2026-10-12-nars-self-evidence-literature.md](2026-10-12-nars-self-evidence-literature.md) |
 | 2026-10-12 | `hyperon-metta-lg-translation` | Hyperon/MeTTa to LG sibling map: atom/hypergraph roundtrip, binding and rewrite parity, typed folds, scoped effects and 15 open questions | [2026-10-12-hyperon-metta-lg-translation.md](2026-10-12-hyperon-metta-lg-translation.md) |