[codex] Add R3 evaluator debt sweep receipt

docs/audit/r3-evaluator-pr-1275-1500-debt-sweep.md

@@ -0,0 +1,110 @@
+# R3 Evaluator PR #1275-#1500 Debt Sweep
+
+**Status:** audit receipt, 2026-05-06. This is a source/PR-history sweep, not
+an implementation brief.
+
+**Scope:** merged PRs #1275-#1500 that touched
+`src/v3/compiler/src/lib.rs` evaluator surfaces:
+`evaluator::Value`, `eval_value`, `eval_port`, `eval_node`, `eval_branch`,
+`eval_loop`, `eval_bind`, and `eval_transform_node`.
+
+**Verification inputs:** local `git log --all -- src/v3/compiler/src/lib.rs`,
+`gh pr view` for the PR rows below, live `src/v3/compiler/src/lib.rs`, and
+the current in-repo evaluator dispatch/readiness docs:
+
+- `docs/briefs/r3-evaluator-dispatch.md`.
+- `docs/briefs/r3-pr-e5-loop-readiness-audit.md`.
+- `docs/briefs/r3-pr-e6-lens-fold-readiness-audit.md`.
+- `docs/briefs/r3-pr-e6-g1a-static-lens-fold-dispatch-packet.md`.
+- `docs/briefs/x1b-evaluator-impact-audit.md`.
+- `docs/briefs/r2-evaluator-test-runner-authority-ratchet.md`.
+- `ROADMAP.md` Pattern B / `test_runner.rs` predicate-authority debt.
+
+**Authority caveat:** this receipt does not settle the current E6-G0/G1
+implementation state. The #1275-#1500 sweep predates those later PRs, and the
+in-repo dispatch/readiness docs still carry E6 callable/fold work as an active
+coordination surface. Treat E6/Callable/G0 as open for Phase 3 reconciliation
+until the live evaluator and dispatch/readiness docs agree that the callable
+surface is executable and retired.
+
+## Summary
+
+The PRs that materially built the R2 evaluator body surface are legitimate
+hand-Rust capacity slices, not accidental parallel substrate authority rows:
+#1374, #1387, #1407, #1426, #1476, and #1496.
+
+The material residual classifications are:
+
+- #1426 introduced an explicit Bool literal-vs-variant branch identity gap.
+  That gap is retired by #1467/#1473 through shared
+  `Dag::bool_runtime_variant_id` authority.
+- #1476 keeps `LoopBound::Descent` fail-closed. This remains active and routes
+  through the Substrate termination-proof carrier before Evaluator consume-side
+  wiring.
+- #1407/#1496 left callable pressure for later E6 work. Keep E6/Callable/G0
+  open for Phase 3 authority reconciliation; this receipt does not assert that
+  callable execution is retired. The same handoff should preserve the split
+  between G1.a static lens fold/report production and G1.b generic dispatch
+  held on X1.b S1/S3.
+- #1484 is not a body-evaluator regression. `SymbolicCostExprEquals` remains
+  a runner/test-predicate gap tracked outside the `lib.rs` body evaluator.
+- #1459 and #1466 touched `lib.rs` via broader generated/substrate fallout,
+  but this sweep found no evidence promoting either into evaluator-lane causal
+  history.
+
+## Chronological Rows
+
+| PR | Merged-at | Evaluator-path touch | Hand-Rust / debt classification | Current dissolution state |
+|---|---:|---|---|---|
+| #1374 `feat(evaluator): PR-E E2 EvalFrame lookup and Bind environment` | 2026-05-01T07:04:37Z | `EvalFrame` / evaluator stack lookup helpers in `lib.rs`. | Hand-Rust capacity slice inside the existing evaluator module. PR body explicitly excludes runtime carrier changes, list fallback, and body evaluation. | Consumed by later E1/E4/E5/Bind slices. No new debt row found. |
+| #1387 `feat(evaluator): PR-E E1 value behavior execution` | 2026-05-01T08:51:37Z | `Behavior::Value`, `eval_value`, `eval_port`, `eval_node`, and `evaluate_body`. | Hand-Rust evaluator capacity. Uses the Rust runtime `Value` mirror and constructs only `LiteralValue` in this slice. | Active foundation under the broader PB-zero hand-Rust retirement trajectory. |
+| #1407 `feat(evaluator): PR-E E3 eager Transform application` | 2026-05-01T20:04:31Z | `eval_transform_node` for supported operator targets. | Hand-Rust capacity plus explicit unsupported transform fences. These are fail-closed residuals, not hidden bridge carriers. | FieldProject and Arrow/UserDefined callable execution are later E6 surfaces. Carry E6/Callable/G0 open into Phase 3 authority reconciliation, alongside G1.a static lens fold/report production and G1.b held on X1.b S1/S3. |
+| #1425 `fix(v3): P0 DB-8 receipt.json key pin + SG-0 census` | 2026-05-01T20:04:41Z | Touched `lib.rs`, but not evaluator-lane behavior. | Receipt-key / SG-0 census work outside the evaluator lane. | Not an evaluator debt row. |
+| #1426 `feat(evaluator): PR-E E4 Branch arm coverage` | 2026-05-01T21:29:41Z | `Behavior::Branch`, `eval_branch`, `Value::VariantValue`, branch frame/payload path execution. | Hand-Rust capacity plus explicit Bool literal-vs-variant identity gap. The PR deferred Bool `if` evaluation to shared substrate identity instead of adding a local map. | Retired by #1467/#1473 for evaluator and `lens_apply` Bool matching. |
+| #1459 `keen-swift-519` | 2026-05-02T03:49:09Z | Touched `lib.rs` amid broad generated/substrate changes. | No evaluator-specific hand-Rust/debt introduction identified from PR metadata available in this sweep. | Do not promote into evaluator causal history without narrower file/body evidence. |
+| #1466 `feat(std): T-Numeric-Construction Slice 3` | 2026-05-02T01:41:30Z | Touched `lib.rs` through regenerated/bootstrap fallout. | Numeric-construction substrate/std work, not evaluator semantics. | Not an evaluator debt row. |
+| #1467 `fix(v3): reify Bool branch scrutinees through shared variant identity` | 2026-05-02T01:15:31Z | Evaluator Bool branch reification; also `dag.rs` and `lens_apply.rs`. | Retires bridge risk by centralizing on `Dag::bool_runtime_variant_id`; does not introduce a local Bool declaration-id map. | Retires #1426 Bool branch identity gap. |
+| #1473 `test(evaluator): cover Bool branch reification` | 2026-05-02T03:48:42Z | Evaluator tests in `lib.rs`. | Test-only hand-Rust; adds fail-closed coverage when Bool authority is missing. | Reinforces #1467 retirement. |
+| #1476 `feat(evaluator): PR-E E5 loop cardinality execution` | 2026-05-02T03:49:06Z | `Behavior::Loop`, cardinality execution, accumulator frame threading. | Hand-Rust capacity. Keeps `LoopBound::Descent` as a typed fail-closed residual. | Active. `docs/briefs/r3-pr-e5-loop-readiness-audit.md` keeps Descent outside the cardinality slice; evaluator consume-side wiring waits on a termination-proof carrier. |
+| #1484 `feat(evaluator): PR-E E7 symbolic-cost-only analyze_complexity wrapper` | 2026-05-02T04:57:42Z | Touched `lib.rs`; primary wrapper is in `dimension.rs`. | Thin wrapper over the existing symbolic-cost analyzer. No parallel analyzer and no new variants per PR body. | `SymbolicCostExprEquals` remains a runner/test-predicate gap, tracked in ROADMAP Pattern B and `docs/briefs/r2-evaluator-test-runner-authority-ratchet.md`, not as a body-evaluator regression. |
+| #1496 `feat(evaluator): PR-E Bind callable entry` | 2026-05-02T08:33:41Z | `Behavior::Bind`, `eval_bind`, callable frame entry prerequisite. | Hand-Rust capacity. PR body excludes E6 lens fold, runner work, substrate changes, and new runtime carriers. | Later E6/G1 prerequisite. Carry E6/Callable/G0 open for Phase 3 authority reconciliation, plus G1.a static lens fold/report production and G1.b generic dispatch held on X1.b S1/S3. |
+
+## Live Follow-Up Map
+
+- `LoopBound::Descent`: active fail-closed evaluator residual. Owner path is
+  Substrate termination-proof carrier, then Evaluator consume-side wiring.
+- E6 callable/report path: keep E6/Callable/G0 open until the live evaluator
+  and dispatch/readiness docs agree that callable execution is executable and
+  retired. G1.a static lens fold/report production remains active evaluator
+  pressure, and G1.b generic dispatch stays deferred behind X1.b S1/S3.
+- Runner/test-predicate authority: `SymbolicCostExprEquals` and related
+  predicate execution gaps live under ROADMAP Pattern B and
+  `docs/briefs/r2-evaluator-test-runner-authority-ratchet.md`, not in the
+  #1275-#1500 `lib.rs` body-evaluator sweep.
+
+## Phase 3 Compile Handoff
+
+Feed these rows into the broader `docs/audit/r3-debt-sweep-2026-05-06.md`
+Phase 3 compile as evaluator-slice evidence only; this receipt does not claim
+global bridge counts.
+
+- **Pattern C / evaluator runtime mirror evidence:** #1387 establishes the
+  Rust evaluator `Value` execution surface that mirrors the substrate runtime
+  value family for body evaluation. Treat it as R2 evaluator capacity evidence,
+  with retirement pressure owned by the PB-zero / v3-authoritative runtime path.
+- **Pattern A / runtime-prereq evidence:** #1476 is the active Descent row.
+  It should feed the Phase 3 compile through the Descent residual tracked by
+  `docs/briefs/r3-pr-e5-loop-readiness-audit.md`, not as a closed evaluator
+  item.
+- **Pattern A / callable-report evidence:** #1407 and #1496 are historical
+  prerequisites for E6 callable execution. The compile should keep
+  E6/Callable/G0 open until the live evaluator and dispatch/readiness docs
+  agree that callable execution is executable and retired. The same row should
+  carry G1.a static lens fold/report production pressure and the G1.b
+  generic-dispatch hold on X1.b S1/S3.
+- **Pattern B / runner predicate evidence:** #1484 should feed the
+  `SymbolicCostExprEquals` runner/test-predicate row, with no claim that the
+  body evaluator regressed or introduced a parallel analyzer.
+- **Do-not-promote rows:** #1459 and #1466 should remain out of evaluator
+  causal history unless a later compile pass supplies narrower PR-body or
+  file-diff evidence.