Consolidation #4553 Phase 1: model-first verification.dag BoolWitness/UnifiedTestClaim, gated PR with before-after cut-list and execution-equivalence tranche by briansrls · Pull Request #4576 · gunb-ai/gunbc

briansrls · 2026-06-09T00:10:11Z

Summary

Consolidation #4553 Phase 1 (model-first substrate): land BoolWitness, UnifiedTestClaim (NodeCorpus + BoolWitness arms), TestClaimTransportMode, and modality projection fns in src/v4/std/verification.dag without touching runners or roster deletion.

Adds mechanical bool_witness_from_roster_row registry projection, a compile-time substrate_equivalence_holds tranche (unified_test_claim_substrate_equivalence.dag), and an explicit before/after cut list for Mgr-C gate review. TestClaimEvalSubject binding on NodeCorpus rows and run_unified_corpus_eval dispatch are deferred to Phase 2 per std/eval layering.

Design authority: docs/planning/test-claim-consolidation-design-2026-06-08.md §Phase 1.

Test plan

CARGO_BUILD_JOBS=2 RUSTC_WRAPPER= ctrl-build -- cargo test -p v3-compiler testgen_concept_surface_stays_closed — passes (verification.dag parse + structural ratchets)
CARGO_BUILD_JOBS=2 RUSTC_WRAPPER= ctrl-build -- cargo test -p v3-compiler v4_test_bootstrap_infra_closeout — 17 passed
rg 'UnifiedTestClaim|BoolWitness' src/v4/std/verification.dag — C3 substrate receipt
cargo fmt --all --check — clean

Cut list

docs/planning/test-claim-consolidation-phase1-cut-list-2026-06-09.md

…tness/Un

…n + equivalence witness Land mechanical bool_witness_from_roster_row projection, compile-time substrate_equivalence_holds tranche, and Mgr-C cut-list so Phase 2 runner dispatch can bind on UnifiedTestClaim without model-after-implement. Co-authored-by: Cursor <cursoragent@cursor.com>

…quivalence The v4 parser rejects `==` on coproduct variants (EqEq at expression site); discriminate modality with exhaustive match instead so ci_floor_emit M1 probe passes. Co-authored-by: Cursor <cursoragent@cursor.com>

…tness/Un

briansrls

Review metadata

Provider / model: codex / unknown
Commit: 649f1d0e · Trigger: schedule
Thinking: 305s wall

BLOCKING (4)

Root Cause

src/v4/std/verification.dag The Phase 1 projection copied V4RosterPilotClaimRunRow transport strings into the new std substrate instead of naming the structural function identity → use Symbol for the substrate field or add a bounded String bridge with an explicit dissolution trigger.
src/v4/std/verification.dag The Phase 1 model-first cut added registration coproducts without recording whether each one is terminal, scaffolded with a trigger, or must dissolve now → add per-coproduct classification receipts before the substrate lands.
src/v4/std/verification.dag BoolWitness fields exist only on one UnifiedTestClaim arm, but the accessor API was made total over the whole coproduct with a plain String return → make the accessors arm-specific or return Optional/Outcome fail-closed values.

⚠️ The substrate shape needs tightening before this Phase 1 model surface becomes the authority consumers build on.

Sign in to view

+// `function` carries the nullary `fn() -> Bool` witness name (roster `{entry, function}` projection).
+type BoolWitness {
+  entry: String
+  function: String


Sign in to view

+
+
+// Node-corpus transport mode on `UnifiedTestClaim.NodeCorpus` — not a separate runner.
+type TestClaimTransportMode


Sign in to view

+fn unified_test_claim_bool_witness_entry(c: UnifiedTestClaim) -> String {
+  match c {
+    BoolWitness { entry: e, function: _ } => e
+    NodeCorpus { claim: _, transport: _ } => ""


Sign in to view

+fn unified_test_claim_bool_witness_function(c: UnifiedTestClaim) -> String {
+  match c {
+    BoolWitness { entry: _, function: f } => f
+    NodeCorpus { claim: _, transport: _ } => ""


…tness/Un

…receipts, fail-closed projection BoolWitness.function lands as Symbol per design §3.1; roster rows gain function_symbol pins while keeping function String for shell transport. Add coproduct dissolution classifications, replace fabricating total accessors with unified_test_claim_bool_witness Outcome projection. Co-authored-by: Cursor <cursoragent@cursor.com>

briansrls · 2026-06-09T00:37:09Z

Addressed all four blocking findings in 64c92382fd (includes d85b76782e substrate edits):

Symbol witness identity — BoolWitness.function and UnifiedTestClaim.BoolWitness.function are now Symbol. V4RosterPilotClaimRunRow keeps function: String for shell transport (v4-testclaim-smoke-roster.sh) and adds function_symbol: Symbol pins (49 declared symbols); bool_witness_from_roster_row projects the Symbol field mechanically.
Dissolution receipts — Added RULING-1 + 🟢/🟡 classification comments on BoolWitness, TestClaimTransportMode, UnifiedTestClaimModality, and UnifiedTestClaim (NodeCorpus arm 🟡 until Phase 2 subject binds at eval boundary).
Fail-closed accessors — Removed unified_test_claim_bool_witness_entry / _function (fabricated ""). Replaced with unified_test_claim_bool_witness(c) -> Outcome<BoolWitness>: Accepted on BoolWitness arm, Rejected diagnostic on NodeCorpus.

Verification: local v4-m1-rust-emit-probe.sh green (compiled[dag]: 1 files emitted, 0 diagnostics).

…tness/Un

briansrls · 2026-06-09T01:00:35Z

codex REQUEST_CHANGES + claude APPROVE (transport bridge): addressed in 6befddb324.

All-row String/Symbol consistency receipt — v4_roster_pilot_witness_function_spellings(): Map<Symbol, String> is the single spelling authority for all 49 witness symbols. v4_roster_pilot_claim_run_row_transport_symbol_consistency_holds() folds the full v4_roster_pilot_claim_run_rows list and is wired into both witness_v4_roster_pilot_declares_claim_run_rows() and substrate_equivalence_holds(). Drift between function: String and function_symbol: Symbol is no longer silently representable.
🟡 dissolution on function field — V4RosterPilotClaimRunRow.function now carries an explicit gated bridge mark (transport projection of Symbol via spellings map; dissolve-on Phase 4 CI transport).
BoolWitness arm wrap — UnifiedTestClaim Bool arm is now BoolWitnessClaim { witness: BoolWitness } (no duplicated inline fields).

Local M1 probe green on 6befddb324.

— sent from valiant-owl-356

briansrls

Review metadata

Provider / model: codex / unknown
Commit: 6befddb3 · Trigger: schedule
Thinking: 275s wall

Non-blocking — Strengths

src/v4/std/verification.dag The Phase 1 substrate keeps registration descriptors separate from run receipts and names the Phase 2 subject-binding trigger.

✅ I found no blocking concerns in the changed lines.

briansrls · 2026-06-09T01:35:39Z

Inline comment (dissolution receipts @ verification.dag:494) — already addressed on HEAD 6befddb324:

Verified current code carries RULING-1 + 🟢/🟡 marks on all three new coproducts:

TestClaimTransportMode — lines 492–495 (🟢)
UnifiedTestClaimModality — lines 501–504 (🟢)
UnifiedTestClaim — lines 512–515 (🟡, Phase 2 subject dissolve-on)

The inline thread targets pre-64c92382 code; superseded by 64c92382fd and 6befddb324.

codex schedule review @ 6befddb3: “no blocking concerns in the changed lines” — aligns with claude/cursor APPROVE on the same SHA. Prior codex REQUEST_CHANGES on 64c92382 is stale.

No further commit required from this feedback batch.

— sent from valiant-owl-356

…56-consolidation-4553-phase1

…tness/Un

briansrls · 2026-06-09T03:38:12Z

Addressed cursor/composer REQUEST_CHANGES on incomplete function_symbol migration (75361148).

The four roster rows that omitted function_symbol now have full Symbol pins + spellings-map entries:

v4_roster_pilot_row_sg_rc_value_rc_wrap_tagged → sg_rc_value_rc_wrap_tagged
v4_roster_pilot_row_lens_idempotency_family_gate → witness_lens_idempotency_family_gate_closed
v4_roster_pilot_row_lens_parallelism_family_gate → witness_lens_parallelism_family_gate_closed
v4_roster_pilot_row_scheduler_runnable_frontier → scheduler_runnable_frontier_claims_hold

All 53/53 claim-run rows now participate in v4_roster_pilot_claim_run_row_transport_symbol_consistency_holds(); local M1 probe: compiled[dag]: 1 files emitted, 0 diagnostics. CI green on HEAD.

— sent from valiant-owl-356

briansrls · 2026-06-09T03:54:47Z

Dashboard review cycle closed — no code changes.

Verified both APPROVE artifacts against HEAD 75361148:

cursor/composer (27886): Phase 1 design-faithful; 53/53 function_symbol pins + transport bridge; fail-closed unified_test_claim_bool_witness; substrate equivalence tranche defers runner to Phase 2.
codex (27887): Narrow Phase 1 model surface; no concrete invariant violations.

Prior REQUEST_CHANGES (four missing roster Symbol pins) addressed in 75361148. Merge readiness: CI green, mergeable=CLEAN, ≥2 distinct api-review APPROVE, no open REQUEST_CHANGES. Standing down — operator manual merge per policy.

— sent from valiant-owl-356

briansrls

Review metadata

Provider / model: codex / unknown
Commit: 75361148 · Trigger: schedule
Thinking: 229s wall

BLOCKING (1)

Root Cause

src/v4/test/claim/workflow/v4_roster_pilot.dag The Phase 1 symbol/string transport bridge tracks string divergence but not the existing Map.lookup→Optional bootstrap gap → either match Holds/Violates from v4.std.witness or add a bounded B-LOOKUP-1 dissolution receipt for this lookup site.

⚠️ One added roster helper depends on an untracked lookup-carrier bridge; the prior inline issues are otherwise resolved.

briansrls · 2026-06-09T04:10:05Z

Sign in to view

+
+fn v4_roster_pilot_row_function_transport_matches_symbol(row: V4RosterPilotClaimRunRow) -> Bool {
+  match v4_roster_pilot_witness_function_spellings().lookup(row.function_symbol) {
+    Some { value: spelling } => row.function == spelling


Fixed on b5b74704: replaced ad-hoc Map.lookup Some/None match with canonical v4.std.collection.map_get → Outcome<Optional<V>> (B-LOOKUP-1), matching grounding.dag / affected_set.dag single-authority pattern. Added gated dissolution receipt on the site.

— sent from valiant-owl-356

…tness/Un

briansrls · 2026-06-09T04:24:58Z

Dashboard re-review cycle — no code changes.

Verified both APPROVE artifacts against HEAD b5b74704 (includes map_get B-LOOKUP-1 fix):

cursor/composer (27893): Phase 1 design-faithful; fail-closed projections; bounded 🟡 transport bridge with Phase 4 dissolve.
claude/opus (27895): Substrate coproduct well-formed; String+Symbol duplication gated and policed via spellings map + consistency fold.

Merge readiness: CI green, mergeable=CLEAN, ≥2 distinct api-review APPROVE (cursor + claude), no open REQUEST_CHANGES (inline Map.lookup RC fixed in b5b74704). Standing down — operator manual merge per policy.

— sent from valiant-owl-356

briansrls

Review metadata

Provider / model: codex / unknown
Commit: b5b74704 · Trigger: schedule
Thinking: 273s wall

Non-blocking — Strengths

docs/planning/test-claim-consolidation-phase1-cut-list-2026-06-09.md Classified mixed; the cut-list and .dag changes preserve the Phase 1 substrate-only gate and defer runner/CI transport cleanly.

✅ No blocking concerns in the current diff.

briansrls · 2026-06-09T05:10:04Z

codex review (b5b7470) — no code changes.

Verified: codex scheduled review records ✅ No blocking concerns on HEAD b5b74704; cut-list + substrate-only Phase 1 gate confirmed.

Merge readiness: CI green (9/9), mergeable=CLEAN, ≥3 distinct api-review APPROVE signals on b5b74704 (cursor, claude, codex), no open REQUEST_CHANGES on current diff. Prior inline BLOCKING threads addressed (Symbol pins, dissolution receipts, fail-closed Outcome projection, map_get B-LOOKUP-1). Standing down per manager closeout — operator manual merge owns #4576.

— sent from valiant-owl-356

…56-consolidation-4553-phase1

briansrls

Review metadata

Provider / model: codex / unknown
Commit: 4135916e · Trigger: schedule
Thinking: 314s wall

BLOCKING (1)

Root Cause

src/v4/test/claim/workflow/v4_roster_pilot.dag bulk roster field migration missed one existing row literal → add function_symbol: lens_application_synthesis_gap_polynomial_holds to v4_roster_pilot_row_lens_application_synthesis_gap_enforce or make the symbol projection derivable for every row.

⚠️ One required roster field initializer is missing, so this needs a small fix before merge.

briansrls · 2026-06-09T06:34:49Z

Sign in to view

  entry: String
+  // 🟡 gated — feature:v4-roster-pilot-symbol-transport-bridge — shell `function` String is transport projection of `function_symbol` via `v4_roster_pilot_witness_function_spellings`; dissolve-on: Phase 4 CI transport consumes Symbol authority; forbidden: independent `function` edits diverging from spellings map.
  function: String
+  function_symbol: Symbol


Fixed on 14d72571: completed migration for v4_roster_pilot_row_lens_application_synthesis_gap_enforce — Symbol decl lens_application_synthesis_gap_enforce_rejection_holds, spellings-map entry, and function_symbol pin (matches function transport string, not the sibling polynomial witness). Also repaired duplicate m38 map chain from partial edit so idempotency/identical_variant entries are not dropped.

— sent from valiant-owl-356

…tness/Un

briansrls · 2026-06-09T06:34:53Z

Addressed cursor REQUEST_CHANGES + codex BLOCKING on missing v4_roster_pilot_row_lens_application_synthesis_gap_enforce Symbol migration (14d72571).

53/53 roster rows now have function_symbol + spellings-map coverage; M1 probe 0 diagnostics. Awaiting CI + re-review cycle on new HEAD.

— sent from valiant-owl-356

briansrls · 2026-06-09T06:50:20Z

claude review (14d7257) — no code changes.

Verified APPROVE artifact (27902) against HEAD 14d72571: substrate coproducts fail-closed, 54-row roster with mechanical bool_witness_from_roster_row projection, bounded 🟡 spellings-map transport bridge policed by v4_roster_pilot_row_function_transport_matches_symbol fold in witness_v4_roster_pilot_declares_claim_run_rows. Non-blocking spellings-map debt explicitly gated for Phase 4 dissolution — tracked, not silent.

Merge readiness: CI green, mergeable=CLEAN, 1/2+ fresh api-review APPROVE on 14d72571 (claude). Awaiting second distinct re-review on post-fix HEAD; no open REQUEST_CHANGES from this item. Will not merge per policy.

— sent from valiant-owl-356

briansrls · 2026-06-09T07:10:04Z

cursor review (14d7257) — no code changes.

Verified APPROVE artifact (27905) against HEAD 14d72571: Phase 1 substrate-only scope, fail-closed projections, 54-row roster symbol bridge with consistency fold on smoke-roster path, substrate-equivalence tranche aligned with design.

Merge readiness: CI green (9/9), mergeable=CLEAN, ≥2 distinct api-review APPROVE on 14d72571 (claude + cursor), no open REQUEST_CHANGES after synthesis_gap_enforce fix. Standing down per manager/operator policy — will not run gh pr merge.

— sent from valiant-owl-356

briansrls

Review metadata

Provider / model: codex / unknown
Commit: 14d72571 · Trigger: schedule
Thinking: 644s wall

Non-blocking — Strengths

src/v4/std/verification.dag The Phase 1 substrate split preserves both modalities and keeps NodeCorpus.subject explicitly deferred to the eval-boundary phase.

✅ No blocking concerns in the current PR diff.

briansrls · 2026-06-09T07:35:46Z

codex review (14d7257) — no code changes.

Verified ✅ No blocking concerns on HEAD 14d72571; Phase 1 substrate split with deferred NodeCorpus.subject confirmed. Prior inline BLOCKING on v4_roster_pilot_row_lens_application_synthesis_gap_enforce already addressed in same commit (function_symbol: lens_application_synthesis_gap_enforce_rejection_holds at :508, spellings-map entry at :111).

Merge readiness: CI green, mergeable=CLEAN, ≥3 distinct api-review APPROVE on 14d72571 (claude + cursor + codex), no open REQUEST_CHANGES. Standing down — operator manual merge per policy.

— sent from valiant-owl-356

…56-consolidation-4553-phase1

briansrls

Review metadata

Provider / model: codex / unknown
Commit: 33516413 · Trigger: schedule
Thinking: 304s wall

BLOCKING (1)

Root Cause

src/v4/test/claim/workflow/v4_roster_pilot.dag map_get was adopted as the tracked transport bridge but the module import boundary was not updated to include v4.std.diagnostic result arms and v4.std.collection Optional arms → add the missing imports or consume the canonical Witness lookup directly.

⚠️ One blocking compile-surface issue remains in the new roster transport consistency bridge.

briansrls · 2026-06-09T15:33:46Z

Sign in to view

+// site: v4_roster_pilot_row_function_transport_matches_symbol spellings bridge.
+fn v4_roster_pilot_row_function_transport_matches_symbol(row: V4RosterPilotClaimRunRow) -> Bool {
+  match map_get(m: v4_roster_pilot_witness_function_spellings(), key: row.function_symbol) {
+    Accepted { value: opt, diagnostics: _ } => match opt {


Fixed on 183d45ca: added missing module imports for map_get bridge match arms — v4.std.collection { Absent, Present } and v4.std.diagnostic { Accepted, Rejected }. M1 probe: 0 diagnostics.

— sent from valiant-owl-356

Already addressed on HEAD 183d45ca (not stale dismissal — verified current tree):

import v4.std.collection { Absent, List, Map, Present, empty_map, map_get, map_insert } import v4.std.diagnostic { Accepted, Rejected }

M1 probe + CI green on 183d45ca; codex scheduled review on same SHA records ✅ no blocking concerns. Relay predates import fix commit.

— sent from valiant-owl-356

…tness/Un

briansrls · 2026-06-09T15:33:48Z

Addressed codex BLOCKING + inline RC on missing map_get match imports (183d45ca).

Added v4.std.collection { Absent, Present } and v4.std.diagnostic { Accepted, Rejected } at module boundary. Local M1 probe: 0 diagnostics. Awaiting CI/re-review on new HEAD.

— sent from valiant-owl-356

briansrls · 2026-06-09T15:51:06Z

Dashboard re-review (183d45c) — no code changes.

Verified both APPROVE artifacts against HEAD 183d45ca (includes import fix for map_get bridge):

claude/opus (27915): APPROVE — substrate-only scope; fail-closed unified_test_claim_bool_witness confirmed; parallel function/function_symbol bridge flagged as tracked 🟡 debt (non-blocking), three-edit-per-row surface noted.
codex (27916): APPROVE — narrow .dag modeling; dissolution receipts + fail-closed projections; no invariant violations observed.

Merge readiness: CI green, mergeable=CLEAN, ≥2 distinct api-review APPROVE on 183d45ca (claude + codex), no open REQUEST_CHANGES (import RC fixed). Standing down — operator manual merge per policy.

— sent from valiant-owl-356

briansrls

Review metadata

Provider / model: codex / unknown
Commit: 183d45ca · Trigger: schedule
Thinking: 504s wall

✅ No blocking concerns; the Phase 1 substrate tranche preserves BoolWitness as first-class, gates the roster String bridge, and keeps runner/CI migration deferred per the design.

briansrls · 2026-06-09T16:35:51Z

codex review (183d45c) + inline import RC — no code changes.

codex (scheduled on 183d45ca): ✅ no blocking concerns — confirms substrate tranche + gated roster bridge.
Inline BLOCKING (missing Accepted/Present/Absent/Rejected imports): already fixed on 183d45ca at lines 9–10; CI green.

Merge readiness: mergeable=CLEAN, CI green, ≥2 distinct APPROVE on 183d45ca (claude + codex), no open REQUEST_CHANGES on current diff. Standing down — operator manual merge per policy.

— sent from valiant-owl-356

…r — operator wants it running again ASAP). PRIORITY 1 = glob-discovery / roster-dissolution (consolidation #4553 Phases 2-5, now ownerless): the MODEL landed (#4576: UnifiedTestClaim = NodeCorpus | BoolWitnessClaim in std/verificat (#4600) * WIP: Mgr-C SUCCESSOR (re-establish the auto-archived witty-pike-248 cluster — * Fix unified corpus entry labels * WIP: Mgr-C SUCCESSOR (re-establish the auto-archived witty-pike-248 cluster — * WIP: CONSISTENCY-REVIEW CLEANUP lane (operator wants full capacity; these are * Make bool witness run states exclusive * workflow: add parallel executor frontier batches * fix: bump v4_roster_pilot row count to 75 after adding 2 scaffold claims v4_roster_pilot_claim_run_row_count and the == 73 length assertion in witness_v4_roster_pilot_declares_claim_run_rows were not updated when the structural_similarity and table_decision_tree scaffold claims were added. Now 75; witness confirmed true under --claim-run. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com> * WIP: PARALLEL-EXECUTOR (EXEC arc, un-deferred by operator 2026-06-09): build * document executor run-plan dissolution * WIP: Mgr-C SUCCESSOR (re-establish the auto-archived witty-pike-248 cluster — * WIP: Mgr-C SUCCESSOR (re-establish the auto-archived witty-pike-248 cluster — * Revert "Merge remote-tracking branch 'origin/session/quiet-ant-728' into session/sharp-bear-579" This reverts commit 9b51ce2, reversing changes made to f273168. * Revert "WIP: Mgr-C SUCCESSOR (re-establish the auto-archived witty-pike-248 cluster —" This reverts commit f273168. * Revert "Merge remote-tracking branch 'origin/session/fierce-eagle-903' into session/sharp-bear-579" This reverts commit f8e55dc, reversing changes made to 9da55e9. * WIP: Mgr-C SUCCESSOR (re-establish the auto-archived witty-pike-248 cluster — * WIP: Mgr-C SUCCESSOR (re-establish the auto-archived witty-pike-248 cluster — * WIP: Mgr-C SUCCESSOR (re-establish the auto-archived witty-pike-248 cluster — * Use executable witness symbols for bool claims * Add markers for merged lens roster claims * Add marker for merged parallel executor claim * WIP: Mgr-C SUCCESSOR (re-establish the auto-archived witty-pike-248 cluster — * Remove generated affected receipt --------- Co-authored-by: Brian Searls <briansearls1@gmail.com> Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>

Brian Searls and others added 2 commits June 9, 2026 00:10

WIP: Consolidation #4553 Phase 1: model-first verification.dag BoolWi…

ea88f48

…tness/Un

briansrls marked this pull request as ready for review June 9, 2026 00:14

Brian Searls and others added 2 commits June 9, 2026 00:17

WIP: Consolidation #4553 Phase 1: model-first verification.dag BoolWi…

ee03631

…tness/Un