fix(v3): split canonical lens bridge ledger row

[briansrls]

Summary

• split the canonical-lens bridge ledger row into the narrow retired PR docs(pb): pin canonical lens bridge blocker #1183 slice and the broader open patching residual
• keep the live residual structurally visible for test_runner.rs exact-string/name-keyed paths until PB-Runtime interpreter-as-data or a typed lens-registry carrier retires it
• refresh generated bootstrap snapshots, parse manifest, and the bridge-ledger runner ratchet wording for the added row

Verification

• cargo run -p v3-compiler --features bootstrap-regen-fresh --bin regen_bootstrap
• cargo run -p v3-compiler --features bootstrap-regen-fresh --bin regen_bootstrap -- --verify
• cargo test -p v3-compiler refresh_handwritten_parse_snapshot_manifest -- --ignored
• cargo test -p v3-compiler --test integration bridge_ledger_carrier_test:: -- --nocapture
• cargo test -p v3-compiler --test integration m1_5_verification_test::r3_bridge_retirement_ledger_zero_fixture_reports_open_rows_at_head -- --nocapture
• cargo fmt --all --check
• git diff --check

[briansrls]

Director — tracked. Bridge-row sub-class split per Option 2 ratification at #1276 c#4358798673. Touches bridge_ledger.dag (1 row → 2 rows: pr1183_slice_retired Retired + patching_residual Open) + carrier test ratchet update. +26/-14 reasonable scope. Mgr will retitle from session-dashboard placeholder when marked ready. — sent from zesty-bear-812

[briansrls]

Review metadata

• Provider / model: claude / claude-opus-4-7
• Commit: 42675330 · Trigger: schedule
• Comparison: origin/main @ 7addae94 ... review/pr-1396-42675330 @ 42675330
• Thinking: 16s wall

---

Verdict: APPROVE — Diff is narrowly scoped: one ledger row split into a retired PR #1183 slice and a named open residual, with regenerated bootstrap fixture, manifest hash bump, and test-comment refresh. Both rows carry name/owner/status/authority per the BridgeLedgerRow shape, the new Open row points at a concrete disposition doc, and the runner test (which fails while any Open row exists) is correctly relaxed from "three Open rows" to a count-agnostic comment. No invariant or modeling-discipline issues observed in the diff.

[briansrls]

Review metadata

• Provider / model: openai-pro / gpt-5-5-pro
• Commit: 42675330 · Trigger: manual
• Comparison: main @ 89382fc9 ... bridge-ledger-canonical-lens-split @ cc9ada2c
• Conversation: View conversation

---

1. Story of the diff

This PR narrows an over-broad bridge-ledger claim. The previous single bridge_canonical_lens_name_dispatch_retired row is split into a precise retired PR #1183 slice, bridge_canonical_lens_name_dispatch_pr1183_slice_retired, and a still-open residual, bridge_canonical_lens_name_patching_residual, so the ledger no longer implies that all canonical-lens-name string patching is gone when only the typed-dispatch slice was retired (src/v3/std/bridge_ledger.dag:43-51, src/v3/std/bridge_ledger.dag:137, src/v3/std/bridge_ledger.dag:143-146). The bootstrap snapshots are regenerated to carry the same source-level ledger data (src/v3/compiler/src/bootstrap_generated.rs:801, src/v3/compiler/src/bootstrap_generated_without_parse_surface.rs:801), and the parse corpus manifest hash is updated for the changed .dag file (src/v3/compiler/tests/integration/parse_corpus_manifest.txt:30). The existing bridge-ledger-zero test comment is also generalized so it describes the invariant, not a stale count of open rows (src/v3/compiler/tests/integration/bridge_ledger_carrier_test.rs:262-267).

2. Invariant categories

1. LAYER MODEL (substrate vs implementation).

Compliant — this touches substrate-authored .dag data, but it does not introduce a new substrate carrier or variant; it refines the authoritative bridge_ledger rows in source and mirrors them in both generated bootstrap fixtures (src/v3/std/bridge_ledger.dag:143-146, src/v3/compiler/src/bootstrap_generated.rs:801, src/v3/compiler/src/bootstrap_generated_without_parse_surface.rs:801).

1. INVARIANTS.md + modeling-discipline.md.

Compliant — Boundary Discipline / single-authority is improved: the comment now says split rows enumerate sub-classes when one prose bridge has both a retired slice and remaining residual, avoiding one row carrying two incompatible statuses (src/v3/std/bridge_ledger.dag:115-118). Fail-closed is also preserved in the test contract: ledger-zero must still Fail and name every open row rather than pretending the ledger is already clean (src/v3/compiler/tests/integration/bridge_ledger_carrier_test.rs:262-267).

1. CODING.md.

Compliant — the change is data-first and keeps behavior out of hidden Rust logic: the source ledger rows carry the new facts directly (src/v3/std/bridge_ledger.dag:137, src/v3/std/bridge_ledger.dag:143-146), while the generated Rust only materializes those rows as bootstrap data (src/v3/compiler/src/bootstrap_generated.rs:801). No new methods, object surfaces, hidden state, or primitive sentinel logic are introduced.

1. TESTING.md.

Compliant — no new test was necessary for a row split because the existing behavior-driven bridge-ledger-zero test remains aimed at the runner contract: fail while open rows exist, and name every open row (src/v3/compiler/tests/integration/bridge_ledger_carrier_test.rs:262-267). The parse corpus manifest was updated hermetically for the changed ledger file (src/v3/compiler/tests/integration/parse_corpus_manifest.txt:30).

1. LOCKED DESIGN DECISIONS.

N/A — the diff references design/brief authority strings, but it does not alter a locked design document or diverge from a locked thesis claim in the changed lines.

1. TRACKED vs UNTRACKED DEBT.

Compliant — the new residual bridge is tracked: the comment bounds the live class to exact-string canonical-lens-name remnants in test_runner.rs, including include_str!, name arms, and generic name-keyed lookups (src/v3/std/bridge_ledger.dag:47-50); it names the dissolution path as PB-Runtime interpreter-as-data or a typed lens-registry carrier (src/v3/std/bridge_ledger.dag:50-51); and the row records owner/status/authority (src/v3/std/bridge_ledger.dag:143-146).

3. Verdict

APPROVE

The PR makes the ledger more truthful by separating a retired typed-dispatch slice from a still-open string-patching residual, and the source, generated bootstrap data, manifest, and test commentary stay aligned. I did not find a blocking or non-blocking issue tied to the changed lines.

[briansrls]

Review metadata

• Provider / model: codex / unknown
• Commit: cc9ada2c · Trigger: schedule
• Thinking: 143s wall

---

Non-blocking — Strengths

• src/v3/std/bridge_ledger.dag The split keeps the retired PR #1183 slice separate from the still-open name-patching residual and documents the residual scope plus dissolution trigger.

✅ No blocking concerns.

[briansrls]

Review metadata

• Provider / model: cursor / composer-2
• Commit: cc9ada2c · Trigger: schedule
• Comparison: origin/main @ 89382fc9 ... review/pr-1396-cc9ada2c @ cc9ada2c
• Thinking: 24s wall

---

Findings

None. The diff only adjusts the canonical bridge_ledger substrate (split one retired row into a retired PR #1183 slice row plus an Open residual row with authority pointing at docs/briefs/r2-pb-canonical-lens-bridge-disposition.md and prose naming PB-Runtime / typed-registry as the dissolution path), syncs generated bootstrap fixtures and the parse corpus manifest, and softens an integration-test comment so it does not hard-code row counts. That matches P2 (single authority for the ledger) and tracked bridge expectations from INVARIANTS.md / docs/modeling-discipline.md rather than contradicting them. Nothing here conflicts with CODING.md or TESTING.md for new code style or test discipline.

Verdict

APPROVE — Change is narrow, ledger semantics are clearer, and residual work is explicitly owned and pointed at a named authority; no rubric violations visible in the diff.

[briansrls]

Review metadata

• Provider / model: openai-pro / gpt-5-5-pro
• Commit: cc9ada2c · Trigger: manual
• Conversation: View conversation

---

1. Story of the diff

This PR makes the bridge ledger more precise instead of treating the whole canonical-lens-name bridge as already retired. The old single retired row is narrowed to bridge_canonical_lens_name_dispatch_pr1183_slice_retired, while a new explicit bridge_canonical_lens_name_patching_residual row remains Open with PB-Runtime ownership and a disposition authority (src/v3/std/bridge_ledger.dag:137, src/v3/std/bridge_ledger.dag:143-146). The surrounding prose now explains the split: the PR #1183 dispatch slice is retired via typed DeclarationRef identity, but the broader exact-string canonical-lens-name residual still lives in test_runner.rs through include_str!, name arms, and generic name-keyed lookups until PB-Runtime interpreter-as-data or a typed lens-registry carrier lands (src/v3/std/bridge_ledger.dag:43-51). Generated bootstrap fixtures and the parse corpus manifest are updated to reflect the .dag source change (src/v3/compiler/src/bootstrap_generated.rs:801, src/v3/compiler/src/bootstrap_generated_without_parse_surface.rs:801, src/v3/compiler/tests/integration/parse_corpus_manifest.txt:30).

2. Invariant categories

1. LAYER MODEL (substrate vs implementation).

Compliant — this touches a Dag-resident .dag data authority, and the change keeps the canonical fact in bridge_ledger rather than adding an implementation-side parallel ledger; the generated Rust mirrors the same row split (src/v3/std/bridge_ledger.dag:142-147, src/v3/compiler/src/bootstrap_generated.rs:801).

1. INVARIANTS.md + modeling-discipline.md.

Compliant — Boundary Discipline / single-authority is improved by splitting an overbroad retired row into a retired slice plus an open residual, with the comment explicitly saying split rows enumerate sub-classes when one prose bridge has both states (src/v3/std/bridge_ledger.dag:115-117). Fail-closed/progress discipline is also preserved because the residual is not masked as retired; it is authored as status: Open (src/v3/std/bridge_ledger.dag:143-145).

1. CODING.md.

N/A — no new hand-written Rust behavior, methods, helpers, or error/result APIs are introduced; the Rust edits are regenerated fixture data plus a test comment update (src/v3/compiler/src/bootstrap_generated.rs:801, src/v3/compiler/tests/integration/bridge_ledger_carrier_test.rs:262-267).

1. TESTING.md.

Compliant — the existing bridge-ledger-zero test commentary was adjusted away from hard-coded row counts and toward the behavior contract: fail while any rows are open, name every open row, and re-arm as pass once all rows retire (src/v3/compiler/tests/integration/bridge_ledger_carrier_test.rs:262-267). The parse corpus manifest was also updated for the changed .dag source (src/v3/compiler/tests/integration/parse_corpus_manifest.txt:30).

1. LOCKED DESIGN DECISIONS.

N/A — the diff references authority docs and future retirement mechanisms, but it does not modify a locked design doc or introduce a divergence from a locked decision in the reviewed diff (src/v3/std/bridge_ledger.dag:50-51, src/v3/std/bridge_ledger.dag:146).

1. TRACKED vs UNTRACKED DEBT.

Compliant — the new open residual is tracked rather than free-floating: documentation names the residual, bounds it to the exact-string canonical-lens-name class and current test_runner.rs surfaces, and gives dissolution triggers via PB-Runtime interpreter-as-data or a typed lens-registry carrier (src/v3/std/bridge_ledger.dag:47-51). The data row also names owner, open status, and authority (src/v3/std/bridge_ledger.dag:143-146).

3. Verdict

APPROVE

No findings. The PR makes the bridge ledger more faithful by refusing to overclaim retirement, keeps the canonical authority in the .dag ledger, and updates generated/bootstrap and corpus artifacts consistently.