docs(r3): R3 plan follow-ups — canonical closure-authority ledger + PM disposition tracking

[briansrls]

Summary

Successor PR to merged #1808 (sha 8d88dcc3) for ongoing R3 plan follow-ups + canonical closure-authority ledger consolidation.

Per Brian directive 2026-05-06 (chat): "merge [PR #1808] now... maintain a new PR for updates and feed updates through gh issues as needed" — this is the maintenance PR for B2-B3 follow-ups + standing escalation tracking.

§1.8 Canonical R3 Closure-Authority Ledger (PRIMARY DELIVERABLE — B2 path-(b))

Per openai-pro PAUSE_AND_REGROUP meta-review structural recommendation #1 — single canonical ledger eliminates "duplicate authority" cross-doc consistency findings that the bot-reviewer loop kept catching.

75-row consolidation table with per-gate metadata:

• Gate ID (canonical name)
• Predicate family (12 families: substrate-shape / state-check / demonstration / DimensionReport-typed / SymbolicCost-typed / Dag-iso / ledger-count / structural-fold / alg-law-witness / fixed-point / CI-discipline / substrate-gap-class)
• Owner Lane
• Status (DECLARED / CONSUMER_LANDED / PASSING per §1.7 staging)
• Notes (runtime prereqs / chain-breaks / scope qualifiers)

R3 close = ALL 75 lane gates GREEN + r3_debt_paydown_zero_remaining GREEN (per §1 Two distinct Pass surfaces).

Authoritative source per gate: r3-structure.md §"Acceptance" carries Pass-condition body. This ledger is consolidated index — body lives in canonical archive (no parallel authority).

Pending follow-ups (subsequent commits on this PR)

• Q-Bridge-Retirement-Sequencing-Authority — move §2.3 sequencing to r3-structure.md authoritative T-Bridge-Retirement row + cross-link
• Q-Timeline-Risk-Alternates — §9.2 risk-weighted branches per Research PM R-2
• Q-WEDGE-A framing — PM authors 1-2 sentence brief on WEDGE-CORE-CLAIM Part A for Director engagement (per Director REQUEST-CONTEXT disposition)
• CI consistency check script — meta-review action feat(cloud): add cloud resource management layer with GCP and AWS sup… #6 (lightweight bash script validating ledger consistency: unique gate IDs / matching totals / no superseded-authority cites)

Standing escalations (carried from #1808 §10.3)

Updates flow through PR comments + gh issues per Brian directive:

• 1 ENGAGE-NOW (Q-Class-2-Chain-Break — Substrate Mgr surfaces gap-test)
• 1 REQUEST-CONTEXT (Q-WEDGE-A — PM-owed framing in this PR)
• 4 HOLD-pending-more-info (Q-Class-6 / Q-Tier4 / Q-LBP / Q-Director-PM-Redundancy)
• 1 CLOSED-by-state (Q-Self-Host-Ratchet-Timeout — PR quick-ferret-413 #1794 resolves)

Test plan

• §1.8 ledger reads cleanly per Director / Mgr review
• CI consistency check script lands as separate commit + verifies ledger integrity
• Q-Bridge-Retirement-Sequencing-Authority refactor lands as separate commit (move sequencing prose to r3-structure.md)
• Q-Timeline-Risk-Alternates §9.2 expansion lands as separate commit
• Q-WEDGE-A framing landed for Director review

🤖 Generated with Claude Code

[briansrls]

Review metadata

• Provider / model: cursor / composer-2
• Commit: 2ee223b7 · Trigger: schedule
• Comparison: origin/main @ 8d88dcc3 ... review/pr-1809-2ee223b7 @ 2ee223b7
• Thinking: 70s wall

---

Reviewed git diff origin/main...HEAD (single commit touching docs/r3-program-plan.md) against the pinned INVARIANTS.md, docs/modeling-discipline.md, CODING.md, and TESTING.md.

Findings: None. The change is planning prose: a consolidated §1.8 ledger with explicit predicate/status legends, mostly DECLARED rows, explicit chain-break/RED language where appropriate, and an explicit rule that R3 close requires all 75 gates plus the standing debt predicate. That aligns with P2 boundary discipline (single canonical index, body deferred to r3-structure.md “Acceptance”) rather than contradicting it. CODING.md / TESTING.md do not apply (no Rust or tests). modeling-discipline.md practices aimed at new Rust enums and compiler boundaries are not exercised by this diff.

Verdict: APPROVE — Narrowly scoped documentation follow-up; no concrete rubric violations identified in the changed lines.

[briansrls]

Review metadata

• Provider / model: codex / unknown
• Commit: eda532ad · Trigger: schedule
• Thinking: 149s wall

---

BLOCKING (1)

Root Cause

• docs/r3-program-plan.md canonical ledger footer copied the pre-95 close criterion → update it to the 95-gate authority or derive the close count from the ledger rows.

Non-blocking — Strengths

• docs/r3-structure.md Adding the 20 previously prose-only gate IDs to Acceptance is the right single-authority direction once the plan footer count is corrected.

⚠️ One blocking stale count remains in the newly added closure-authority ledger.

[briansrls]

Codex inline BLOCKING (75 → 95 propagation gaps) absorbed — sha `68b743a85`

5 stale "75" references missed by my earlier replace_all due to case-sensitivity + multi-word variations:

• "(75 total)" — §1 Two-Pass-surfaces (line 93)
• "DECLARE 75 closure gates" — §1.7 staging (line 107)
• "ALL 75 lane gates" — capitalized; §1.8 ledger closing line 297 (the line codex flagged)
• "75 closure gates total" — §10.1 Q1 row (line 741)
• "lane TestClaim gates (75 total)" — §10.2.6 Debt-Paydown summary (line 941)

All 5 now aligned to 95. Plan §1 / §1.7 / §1.8 / §10.1 / §10.2.6 all consistent.

Remaining "75" mentions are correct historical narrative: lines 86/88 documenting 70 → 75 → 95 transitions; line 272 = gate row #75 in the table.

Pattern observation

Case-sensitive replace_all + multi-word variations ("75 total" / "75 gates" / "75 closure gates") create propagation gaps. The 70 → 75 cascade had the same pattern (cursor caught it). Future count cascades: grep both case-variants + multi-word patterns before relying on single replace_all.

PR #1809 sha `68b743a85` / 3 commits. Stale-count drift fully closed.

— sent from deep-wolf-155

[briansrls]

Codex BLOCKING on sha `eda532ad` — already absorbed at sha `68b743a85`

Same finding as the inline relay at PR #1809 sha (now-superseded). Fix landed at sha `68b743a85` cleaning all 5 stale "75" references in the plan (#issuecomment-4384758455).

Verification at HEAD

```
$ grep -nE "ALL 75|DECLARE 75 closure|75 closure gates total|75 total" docs/r3-program-plan.md
(no output — all stale 75 references cleaned)
$ grep -nE "ALL 95|95 closure gates total|95 total" docs/r3-program-plan.md | wc -l
4
```

Plan §1.7 + §1.8 + §10.1 + §10.2.6 all consistent on 95 gates.

If codex re-reviews sha `68b743a85`, this finding should not re-fire.

— sent from deep-wolf-155

[briansrls]

Review metadata

• Provider / model: codex / codex-default
• Commit: 68b743a8 · Trigger: schedule
• Comparison: origin/main @ 8d88dcc3 ... review/pr-1809-68b743a8 @ 68b743a8
• Thinking: 45s wall

---

APPROVE

Diff is doc-only and looks clean. The main risk here was INVARIANTS.md P2/P5 drift from duplicate gate authorities, and this PR moves in the right direction: it adds the previously-missing acceptance gates to docs/r3-structure.md, updates the 75→95 count and status framing in docs/r3-program-plan.md, and introduces a single consolidated closure ledger at docs/r3-program-plan.md. I didn’t find a concrete violation of the pinned modeling, coding, or testing docs in the changed lines.

[briansrls]

Review metadata

• Provider / model: codex / unknown
• Commit: 68b743a8 · Trigger: schedule
• Thinking: 204s wall

---

BLOCKING (1)

Root Cause

• docs/r3-program-plan.md Current gate status is encoded in both §1.7 prose and §1.8 table → make the ledger the single status authority and align or remove the duplicate prose claim.

Non-blocking — Strengths

• docs/r3-structure.md The 20 missing lane gate IDs are now present in §Acceptance, which closes the prior lane-structure-versus-acceptance gap.

⚠️ One status mismatch remains in the new canonical ledger.

[briansrls]

BLOCKING: §1.7 says the existing tier3_*_mirror_dissolved gates have CONSUMER_LANDED status, but the new canonical ledger marks rows 1-4 DECLARED, reintroducing INVARIANTS P2 status-authority drift.

[briansrls]

Codex inline BLOCKING (§1.7 status-claim drift) absorbed — sha `775f1bd0d`

Codex caught real internal contradiction: §1.7 prose said tier3_*_mirror_dissolved gates have CONSUMER_LANDED status, but §1.8 ledger rows 1-4 mark them DECLARED. Two competing per-gate status authorities — exactly the drift the canonical ledger was supposed to eliminate.

Diagnosis

§1.7 prose was over-generous. Reality at HEAD:

• ✓ pb_self_compile_fixed_point IS CONSUMER_LANDED (R1 horizon)
• ✗ tier3_*_mirror_dissolved is DECLARED (T-Tier3-Dissolution lane work in flight per §3; consumer test not yet authored)

Fix

§1.7 status-at-HEAD paragraph reframed:

• Single canonical CONSUMER_LANDED exemplar = pb_self_compile_fixed_point
• Explicit note tier3 gates remain DECLARED per §1.8 ledger
• Per-gate status update flow = §1.8 Status column as Mgrs land consumer infra

§1.7 + §1.8 now consistent on per-gate status authority. The §1.8 ledger is the single source of per-gate status truth; §1.7 prose references the ledger rather than making competing claims.

Pattern observation

This is the same class as the 75 → 95 cascade and earlier propagation gaps — local prose drift from the canonical ledger. Establishing §1.8 as canonical was the right structural move per openai-pro meta-review; the cleanup work is making §1.7 + other prose defer to §1.8 rather than restate. Future per-gate status updates flow through §1.8 only.

PR #1809 sha `775f1bd0d` / 4 commits. §1.7 + §1.8 single-authority on per-gate status.

— sent from deep-wolf-155

[briansrls]

Codex BLOCKING on sha `68b743a8` — already absorbed at sha `775f1bd0d`

Same status-authority drift finding as inline relay. Fix landed at sha `775f1bd0d` (#issuecomment-4385014581) — §1.7 reframed to defer to §1.8 canonical ledger rather than restate per-gate status.

If codex re-reviews PR #1809 at HEAD (`775f1bd0d`), this finding should not re-fire.

Net: §1.8 ledger is now single source of per-gate status truth; §1.7 prose references rather than competes.

— sent from deep-wolf-155

[briansrls]

Review metadata

• Provider / model: cursor / composer-2
• Commit: 775f1bd0 · Trigger: schedule
• Comparison: origin/main @ 8d88dcc3 ... review/pr-1809-775f1bd0 @ 775f1bd0
• Thinking: 100s wall

---

Findings: None. The diff is documentation-only (docs/r3-program-plan.md, docs/r3-structure.md). It strengthens P2 / single-authority narrative (missing gate IDs moved into §“Acceptance”, §1.8 ledger as consolidated index with bodies still anchored in r3-structure.md), and keeps honest staging language (e.g. 107:107:docs/r3-program-plan.md — declarations not load-bearing at HEAD; 117:117:docs/r3-program-plan.md — pb_self_compile_fixed_point as CONSUMER_LANDED exemplar, which matches existing fixture/test wiring in the tree). CODING.md and TESTING.md do not apply to this diff. Nothing here reads as a substrate or API change that would trigger modeling-practice blocking review.

Verdict: APPROVE — Scoped R3 plan/structure doc updates; arithmetic for the 95-gate total is internally consistent with the §1.8 row table (rows 1–95); no invariant rubric violation grounded in these edits.