D-POP-2: the contradiction write-back producer — elect_peers → the row's own register

.claude/board/AGENT_LOG.md

@@ -1,3 +1,10 @@
+## 2026-09-03 — D-POP-2: one Sonnet grindwork worker against a verbatim spec, orchestrator-gated
+
+- **Why:** the post-teardown survey (D-POP-0) named the contradiction write-back as its cheapest gap (§5.4) and D-POP-1 had reported (KILL), unblocking it. Tiering: the orchestrator (main thread) read the tree, wrote the spec (`elect_and_bind` / `ElectionReport` / `WitnessLens::bind_election` + six named falsifiers) and ran every gate and disable centrally; one Sonnet worker transcribed the spec into `witness_fabric.rs` + the `lib.rs` re-export block, edit-only, no cargo.
+- **Worker outcome:** both files edited as briefed; zero production-code deviations; one spec error caught and reported (the silence test's `visited: 2` literal — `rows_from` spans `0..=max_pos`, so it asserted `visited == rows.len()` instead); three `#[must_use]` results in tests needed `let _ =` (orchestrator fix).
+- **Gates (orchestrator):** fmt; clippy `-p lance-graph-contract --all-targets --no-deps -D warnings` clean; `cargo test -p lance-graph-contract` 1309 lib + all integration suites green; disables A/B/C red-then-green (commit `b8642405`).
+- **Board:** EPIPHANIES `E-A-PRODUCER-IS-A-PURE-FUNCTION-OF-THE-CONTENT-LOCI-1`, LATEST_STATE delta, STATUS_BOARD D-POP-2 → In PR, plan §5 gap-4 note; supersession index regenerated last.
+
 ## 2026-08-23 — the token-seam arc: four read-only research lanes, one probe, one vacuity audit
 
 - **Why:** an operator brief asked whether ONE versioned BPE tokenization can

.claude/board/EPIPHANIES.md

@@ -15,6 +15,21 @@
 **Both stay LAB, not deleted, not re-probed on friendlier data** — per the pre-registration's own commitment (§4b: "Not deleted... Not re-probed on a different corpus to get a better answer"). §1c is now fully closed: every row in the harvest census table has a verdict.
 
 Refs: `thinking-engine-harvest-closure-v1` §4b (pre-registration) / §4c (results) / §5 D-TEH-3; `crates/thinking-engine/src/{semantic_chunker,spiral_segment}.rs`; `.claude/agents/falsifier-auditor.md` (the can-it-fire discipline applied to the chunker's all-zero result); the falsifiability rule (CLAUDE.md § The falsifiability rule).
+## 2026-09-03 — E-A-PRODUCER-IS-A-PURE-FUNCTION-OF-THE-CONTENT-LOCI-1 — D-POP-2 closed the survey's cheapest gap: `elect_peers` now writes its own election back into the row, and the fabric's no-self-reference rule is what makes that write order-independent
+
+**Status:** SHIPPED (code + 6 falsifiers, this branch; `witness_fabric::{elect_and_bind, ElectionReport, WitnessLens::bind_election}`). **Confidence:** High for the mechanism; the molecule it unblocks is still unwired downstream.
+
+**What was missing.** `post-teardown-buildup-survey-v1` §5 gap 4: `elect_peers` computed the social loci (Quorum / Contradiction), `WitnessLens::write_register` could store a register, and nothing connected them — so `is_opinion` / `revision_trajectory` / `suggest_reopening` (the "contradiction-driven revision" molecule, §4) never had a row that carried an ELECTED contradiction, only hand-set fixtures.
+
+**The finding, stated as a property.** The producer is a pure function of the content loci. `CONTENT_LOCI` excludes Quorum and Contradiction ("the fabric never reads what it computes" — no self-reference), which was written as an anti-circularity rule for the READ side; it turns out to be exactly what makes the WRITE side sound: a binding written into row *i* cannot change any row *j*'s election, so the whole-wave producer is order-independent, a second run is bit-identical, and stale or garbage social loci are overwritten, never consulted. That is the difference between a producer and an accumulator, and it is pinned two-sided: seeding every row with garbage Quorum=+7 / Contradiction=−8 before the run yields the same bytes as the clean run, and the test goes red the moment `CONTENT_LOCI` includes a social locus (disable-verified).
+
+**Why it stays inside family 1.** A locus is a signed OFFSET (sign = orientation); the producer writes two pointers and nothing else — no magnitude, no valence, no count. No tenant, no ClassView, no layout change, no dependency. Zero-copy: one lens per focal position over the same slice, dropped before the write; the only owned value is the `Copy` 12-byte register microcopy that `bind_election` read-modify-writes into the SAME row (slots 14 and 15 only; a full 14-locus content register plus an out-of-register canary survive byte-identical, disable-verified against a `ZERO.with(..)` rewrite).
+
+**Three disables, all red-then-green:** `bind_election` without read-modify-write (4 of 6 fail); `CONTENT_LOCI` + `Locus::Quorum` (the garbage-seed test fails); the producer passing `|_| true` to the peer scan instead of the caller's `visible` (the visible-domain test fails — +2 must beat the invisible +1, and the same fixture picks +1 when everything is visible).
+
+**What it does NOT do.** The molecule's next links — `BeliefArena::revise_at` on the elected contradiction, `RevisionTrajectory.flips`, `suggest_reopening` over real revisions — are still unwired; this PR gives them a real producer to read, not a caller. No population-basin work: family 3 stays the accepted vacancy (`E-SIX-SEMANTIC-FAMILIES-MUST-NOT-IMPERSONATE-EACH-OTHER-1`).
+
+**Spec correction caught by the worker.** The brief's silence test asserted `visited: 2`; `rows_from` builds the whole `0..=max_pos` span, so a `|_| true` predicate visits 7. The worker asserted `visited == rows.len()` and reported the deviation instead of forcing the literal — the right call, recorded here so the spec's error is not mistaken for the code's.
 
 ## 2026-09-02 — E-THE-LIFT-GATE-FOUND-A-TIE-BLIND-SPEARMAN-1 — the D-TEH-3 comparison on a distinguishing fixture separated a same-formula copy from a wrong-estimator copy, and the tie-free fixtures the lab had used could not have
 

.claude/board/LATEST_STATE.md

@@ -1,3 +1,18 @@
+## 2026-09-03 — branch (D-POP-2, second commit): the producer is READ by its consumers — INVENTORY DELTA
+
+- ADDED `crates/lance-graph-contract/tests/d_pop_2_producer_reaches_consumers.rs` (2 tests, disable-verified against a no-op `bind_election`: both red). No library surface.
+- MEASURED, not added: on a row whose content loci are SMeaning + Kausal, `recipe_loci::reachable` is **0/34** before the producer — every recipe touching either locus also needs a further one — and exactly **7/34** after (`3, 7, 11, 17, 20, 27, 30`, derived from `required_loci` in the test, not hand-listed only); `21` (needs MeaningLevel) and `31` (needs P/OMeaning) stay unreachable, so the unlock is the loci gate, not blanket. `SubstrateView::project`: `confidence` NaN → finite with no SPO tenant present, `dissonance` 0 → 2/15, `free_energy` 0.45 → 0.8, `is_opinion` false → true.
+- UNCHANGED: everything in the first D-POP-2 delta below.
+
+## 2026-09-03 — branch (D-POP-2, after #1143): the contradiction write-back producer — CONTRACT INVENTORY DELTA
+
+- ADDED `lance_graph_contract::witness_fabric::elect_and_bind(rows: &mut [NodeRow], visible: impl Fn(usize) -> bool) -> ElectionReport` — the D-POP-2 producer: for every visible position, `elect_peers_lens` through a lens over the same slice, then `WitnessLens::bind_election` into the focal row. Order-independent + idempotent (elections read `CONTENT_LOCI` only). Re-exported at the crate root.
+- ADDED `witness_fabric::ElectionReport { visited, quorum_bound, contradiction_bound }` (counts only — positions are readable back through the lens; a list would be a second projection). Re-exported.
+- ADDED `WitnessLens::bind_election(row: &mut NodeRow, election: PeerElection)` — read-modify-write of ONLY slots 14/15 (Quorum / Contradiction) through the tenant-derived offsets `at` reads from; offset 0 unbinds.
+- 6 tests, 3 disable-verified (see EPIPHANIES `E-A-PRODUCER-IS-A-PURE-FUNCTION-OF-THE-CONTENT-LOCI-1`). Contract lib 1309/1309 + integration green; clippy `--all-targets --no-deps -D warnings` + fmt clean.
+- UNCHANGED: every existing `witness_fabric` fn and its semantics; `causal_witness.rs`; `canonical_node.rs`; `soa_envelope.rs`; `ENVELOPE_LAYOUT_VERSION`; tenants 14/15. No new dependency.
+- NOT IN THIS DELTA: any caller of the producer (planner / callcenter still build windows by hand or use `quorum_mantissa` only); the molecule's `BeliefArena::revise_at` → `RevisionTrajectory` → `suggest_reopening` wiring; anything population-basin (vacancy stands).
+
 ## 2026-09-03 — MERGED #1144 (`30b0a66`): D-TEH-3 fate probes closed — two corrections to the delta below
 
 PR #1144 (the delta immediately below) merged after CodeRabbit + Codex review.

.claude/board/STATUS_BOARD.md

@@ -30,7 +30,7 @@
 |---|---|---|
 | D-POP-0 | the survey itself: six-family inventory, atom/operator inventory, family×operator legality matrix, molecule candidates, gap list (plan §§1–5) | **Shipped (plan-only, this PR)** |
 | D-POP-1 | PROBE-POP-READOUT-1: composed population readout (Cam96 centroid distance → Fisher-z → occupancy bucket) vs `curiosity` / `curiosity_gestalt` at predicting later-confirmed frontier edges on the KJV stream; size-preserving shuffle null; PASS/KILL pre-registered in plan §6 | **Shipped — VERDICT KILL** (`pop_readout.rs`, 2026-09-02; `E-QUALIA-IS-RANK-INERT-AT-THE-FRONTIER-AND-POPULATION-LOSES-TO-COUNTING-1`, plan §6a). Population readout degrades frontier p@10 0.289 -> 0.011, below its own shuffle null; real partial rho 0.090 vs null p95 0.020 (weak but non-null global trend). Side-findings: `curiosity_gestalt` is rank-inert (rho = 1.000000 vs `curiosity` over 227,261 candidates, two assessments); plain frequency p@10 = 0.756 beats every cognitive arm; helix (Fisher-z / RollingFloor) is not reachable from deepnsm-v2, so those legs are NOT RUN |
-| D-POP-2 | contradiction write-back producer (`elect_peers` → `WitnessLens::write_register`) — the one cheap gap that unblocks the contradiction-driven-revision molecule (plan §4) | **Queued** — not before D-POP-1 reports |
+| D-POP-2 | contradiction write-back producer (`elect_peers` → `WitnessLens::write_register`) — the one cheap gap that unblocks the contradiction-driven-revision molecule (plan §4) | **In PR** (2026-09-03, this branch) — `witness_fabric::{elect_and_bind, ElectionReport, WitnessLens::bind_election}`, 6 falsifiers / 3 disable-verified; D-POP-1 reported KILL 2026-09-02, so the gate was open. See `E-A-PRODUCER-IS-A-PURE-FUNCTION-OF-THE-CONTENT-LOCI-1` |
 
 ## dismech-causal-replay-v1
 

.claude/board/SUPERSESSION-INDEX.md

@@ -92,7 +92,7 @@ a licence to act on it.
 | **READ** | `mul-consumer-census-v1` | `GateDecision` | MEASUREMENT COMPLETE — 2026-08-27. Measureme | 1/2 |
 | **READ** | `mul-ewa-trust-propagation-v1` | `GateDecision` | PROPOSED — PLAN/BOARD ONLY. Measure-before-c | 1/3 |
 | **READ** | `persistence-artifact-backed-commit-v1` | `persist_cycle` | RATIFIED (operator ruling 2026-08-09). Phase | 0/0 |
-| **READ** | `post-teardown-buildup-survey-v1` | `StepMask` | SURVEY, read-only, plan-only (no code, no te | 4/4 |
+| **READ** | `post-teardown-buildup-survey-v1` | `StepMask` | SURVEY, read-only, plan-only (no code, no te | 5/5 |
 | **READ** | `r2il-bpe-typed-genetic-recombination-v1` | `GateDecision` | PROPOSAL, §7's three falsifiers now RUN (see | 1/1 |
 | **READ** | `scientific-kg-substrate-v1` | `GateDecision` | PROPOSED — **scoping doc**, no code. Records | 8/9 |
 | **READ** | `self-reasoning-substrate-v1` | `GateDecision` | PROPOSED — doc-only. No code, no contract ch | 5/5 |

.claude/plans/post-teardown-buildup-survey-v1.md

@@ -335,3 +335,22 @@ stands. What the result does license is a narrower next question, stated as a
 question and not a direction: the readout's honest home may be the global
 trend (a basin-level prior) rather than a top-k selector, and any such probe
 must carry the frequency control and the shuffle null from the start.
+
+## 8. D-POP-2 — gap §5.4 CLOSED (2026-09-03)
+
+`witness_fabric::elect_and_bind(rows, visible) -> ElectionReport` +
+`WitnessLens::bind_election(row, election)`: the producer §5 item 4 named as
+"cheap" is shipped — elected social loci written back into the focal row's own
+register (slots 14/15 only), order-independent and idempotent because elections
+read `CONTENT_LOCI` only. Six falsifiers, three disable-verified. Family 1 only;
+no tenant, ClassView, opcode or layout change; the family-3 vacancy is untouched.
+Still open on the §4 "contradiction-driven revision" molecule: a caller that
+feeds the elected contradiction into `BeliefArena::revise_at` and reads
+`RevisionTrajectory` / `suggest_reopening` over real revisions (gap §5.3, the
+real `WitnessStream` producer, is the same missing caller seen from the other
+side). Board: `E-A-PRODUCER-IS-A-PURE-FUNCTION-OF-THE-CONTENT-LOCI-1`.
+
+Consumer reach measured (2026-09-03, same PR): `reachable` 0/34 → 7/34 on a
+SMeaning+Kausal row (`tests/d_pop_2_producer_reaches_consumers.rs`); the
+§4 molecule's ladder half is now reachable from a real election, its
+`BeliefArena::revise_at` half is not (planner-side caller still missing).

crates/lance-graph-contract/src/lib.rs

@@ -239,12 +239,13 @@ pub use step_mask::StepMask;
 pub use style_family::StyleFamily;
 pub use view_angle::ViewAngle;
 pub use witness_fabric::{
-    absolute_agreement, belief_runs, elect_peers, elect_peers_lens, foresight_calibration,
-    foresight_sample, is_opinion, opinion_strength, quorum_mantissa, quorum_mantissa_lens,
-    resolve_chain, resolve_chain_lens, revision_trajectory, standing_wave_diagnosed,
-    standing_wave_diagnosed_lens, standing_wave_grounded, standing_wave_grounded_lens,
-    standing_wave_stratified, standing_wave_stratified_lens, suggest_reopening, superseded_runs,
-    superseded_spread_sample, trajectory_of, trajectory_of_lens, BeliefRun, ChainResolution,
-    EscalateReason, ForesightSample, PeerElection, ReopeningEvidence, ReopeningReason,
-    ReopeningSuggestion, RevisionTrajectory, TrajectorySignature, WaveGrounding, WitnessLens,
+    absolute_agreement, belief_runs, elect_and_bind, elect_peers, elect_peers_lens,
+    foresight_calibration, foresight_sample, is_opinion, opinion_strength, quorum_mantissa,
+    quorum_mantissa_lens, resolve_chain, resolve_chain_lens, revision_trajectory,
+    standing_wave_diagnosed, standing_wave_diagnosed_lens, standing_wave_grounded,
+    standing_wave_grounded_lens, standing_wave_stratified, standing_wave_stratified_lens,
+    suggest_reopening, superseded_runs, superseded_spread_sample, trajectory_of,
+    trajectory_of_lens, BeliefRun, ChainResolution, ElectionReport, EscalateReason,
+    ForesightSample, PeerElection, ReopeningEvidence, ReopeningReason, ReopeningSuggestion,
+    RevisionTrajectory, TrajectorySignature, WaveGrounding, WitnessLens,
 };