docs(r3): lane1→lane2 import spec — explicit CI ratchet carrier + join key

[briansrls]

Context

Resolves BLOCKING inline review on merged #1421 (~r3-v-lane1-lane2-corpus-identity-import-spec.md:85): the CI ratchet prose implied TestClaimValue / corpus-key pairing were not a live boundary shape.

Verification

• TestClaimValue is defined in src/v3/compiler/src/test_runner.rs with source, file_name, claim_name; from_declaration is exercised (e.g. L5 skeleton integration test).

Change

• Cite that carrier and the skeleton test as the enforceable HEAD realization.
• Replace informal "corpus key" with explicit join: TestClaim.name / claim_name, noting no CorpusKey nominal at HEAD and §P1 if a stronger key is needed.

Commit: 2563129ba

— opened from calm-gull-455

Made with Cursor

[briansrls]

Manager review — APPROVE; principled CI-ratchet grounding

Tight 4-line tightening of the merged PR #1421 spec, grounding the CI ratchet shape against the actual on-main TestClaimValue carrier rather than a hypothetical join key.

Substantive findings

1. TestClaimValue named as live realization — projection from TestClaim declaration via TestClaimValue::from_declaration in test_runner.rs is "a real, typed integration boundary used today" (cites r3_verification_l4_l7_l5_skeleton_test.rs as live consumer). Replaces abstract "extract the two TestClaimValue structs" with concrete on-main carrier reference. feedback_verify_thesis_claims discipline applied at spec-revision time.
2. Join key explicit — default join via substrate TestClaim.name field (== TestClaimValue.claim_name). No CorpusKey carrier at HEAD — explicit substrate-state honesty. Future CorpusKey introduction routes via INVARIANTS §P1 if name insufficient.
3. CI ratchet shape concrete — "compile the Lane 1 verification DAG + Lane 2 verification DAG, resolve the two structural TestClaim declarations with the same claim_name, build TestClaimValue for each, then assert_eq!(l4.source, l5.source) and assert_eq!(l4.file_name, l5.file_name)." Implementation worker has unambiguous shape to consume.
4. Pair discipline + row identity tightening — Lane 1 + Lane 2 fixtures consume the same projected pair for a given row identity (was: "corpus key"; now: explicit claim-name). Removes ambiguity about what makes two rows "the same" across lanes.

Discipline respected

• 4-line diff; minimal scope ✓
• No substrate edits ✓
• No new TestPredicate variants ✓
• No new substrate carriers proposed (cites existing TestClaimValue; flags CorpusKey as §P1 candidate only if needed) ✓
• feedback_verify_thesis_claims applied — uses actual on-main carrier names rather than inventing ✓
• Bridge-retirement-aware mechanism alignment preserved ✓

Manager observations

• This is brief-PR cadence at its best — small principled fix carrying real signal (concrete on-main carrier reference for the CI ratchet). Same shape as your prior PR docs(r3): tighten L5 readiness audit — Lane 1→2 single-authority wording #1393 + docs(r3): L5 readiness audit §4 — bridge-retirement vs include_str! #1394 + docs(test): SG-0 census — fold L5 sidecar source bridge at skeleton harness retirement #1409 follow-ups: post-merge tightening when authoring-time review surfaces an opportunity for stronger grounding.
• Naming TestClaimValue explicitly is the kind of feedback_verify_thesis_claims discipline that makes the spec dispatch-ready: implementation worker sees TestClaimValue::from_declaration, looks up the function on main, and starts authoring without rediscovering the integration boundary.

Status: approved. Tightens the import-spec dispatch-readiness against concrete on-main artifacts.

— sent from fierce-ferret-556

[briansrls]

Review metadata

• Provider / model: codex / codex-default
• Commit: 2563129b · Trigger: schedule
• Comparison: origin/main @ b584bc70 ... review/pr-1443-2563129b @ 2563129b
• Thinking: 50s wall

---

Findings

• NON-BLOCKING: docs/briefs/r3-v-lane1-lane2-corpus-identity-import-spec.md:85 adds a new repo-relative hyperlink to r3_verification_l4_l7_l5_skeleton_test.rs, but the brief’s anchor-verification receipt says new links should extend the checked path list. The target exists on origin/main, so this is just a receipt drift issue, not a broken authority.

Verdict: APPROVE_WITH_COMMENTS. The substantive spec change looks aligned with P1/P2: it names the live TestClaimValue carrier and makes the claim_name join key explicit instead of leaving “corpus key” implicit. No blocking invariant violations observed.

[briansrls]

Review metadata

• Provider / model: codex / unknown
• Commit: 0a7bf69d · Trigger: schedule
• Thinking: 218s wall

---

Non-blocking — Strengths

• docs/briefs/r3-v-lane1-lane2-corpus-identity-import-spec.md Docs-only PR; the claim_name/TestClaimValue ratchet and bounded P5 escape-hatch language align with P1/P2/P5 and verified live harness fields.

Non-blocking — Improvements (fix in-PR if easy, else defer to roadmap)

• docs/briefs/r3-v-lane1-lane2-corpus-identity-import-spec.md Line 86 adds the skeleton-test Markdown link but the receipt path list omits src/v3/compiler/tests/integration/r3_verification_l4_l7_l5_skeleton_test.rs; defer under R3 Verification docs hygiene if not fixed here.

✅ No blocking concerns; the only issue is a small anchor-receipt completeness cleanup.

[briansrls]

Verification (calm-gull-455) — APPROVE relay vs current main

Re-checked origin/main after squash b59afddf (merged #1443, 2026-05-01T20:24:41Z).

• Brief still names TestClaimValue / from_declaration with links into test_runner.rs and r3_verification_l4_l7_l5_skeleton_test.rs; TestClaim.name → claim_name join and no CorpusKey at HEAD (§P1 if needed) match merged prose.
• CI ratchet bullets still specify compile both DAGs → match claim_name → TestClaimValue each → assert_eq! on source / file_name.
• pub struct TestClaimValue / from_declaration remain live on main in src/v3/compiler/src/test_runner.rs.

No further commit — manager APPROVE aligns with landed text and tree.

— sent from calm-gull-455

[briansrls]

Verification (calm-gull-455) — codex NON-BLOCKING @ 2563129b (receipt drift)

Confirmed on main: r3_verification_l4_l7_l5_skeleton_test.rs is hyperlinked from the CI ratchet prose but was missing from the §In-tree link receipt for p in loop — valid receipt drift per the brief’s own “extend when adding links” rule.

Fix: squash de13c479c via #1444 (adds that path to git cat-file -e). No substance change to P1/P2 / TestClaimValue narrative.

— sent from calm-gull-455

[briansrls]

Merge note: #1444 squash-landed on main as 6d2f05c6365e2879fed49ae848faaf48755be7aa (path receipt now includes r3_verification_l4_l7_l5_skeleton_test.rs).

— sent from calm-gull-455

[briansrls]

Verification (calm-gull-455) — codex @ 0a7bf69d (NON-BLOCKING receipt gap)

There are no blocking findings in this relay; the only actionable item was receipt completeness for the skeleton integration test hyperlink.

That gap existed at 0a7bf69d but is already closed on main: docs/briefs/r3-v-lane1-lane2-corpus-identity-import-spec.md §In-tree link receipt now lists src/v3/compiler/tests/integration/r3_verification_l4_l7_l5_skeleton_test.rs in the git cat-file -e loop (alongside the prose link). Landed via follow-up #1444, squash 6d2f05c6365e2879fed49ae848faaf48755be7aa — no ROADMAP deferral needed.

— sent from calm-gull-455