docs(r3-pb): BinShim row #1 entry-function decision packet (Shape A/B/C)

[briansrls]

Summary

• Single-page §P1 decision packet at `docs/briefs/r3-pb-binshim-row-1-decision-packet.md` routing the BinShim chain head-of-chain blocker (no `regen_lens_main` entry function on main) to Director + Substrate Manager + PB Manager.
• Compares Shape A (fail-closed placeholder convention), Shape B (loader-first `dsl/std/runtime/bin_shims/*.dag` allow-list), Shape C (Item 4-first real body) on owner / file surface / prerequisite / chain-unblock semantics / STOP condition.
• Recommends a Director/Substrate disposition step rather than picking unilaterally — the README STOP+PING at `dsl/std/runtime/bin_shims/README.md` explicitly routes the placeholder convention question to that authority.
• Verified row Add SVG viz, test helpers, and makegen scaffold #1 still NOT-LIVE on `origin/main` HEAD `9cf6dd223`. No entry function, no placeholder, no instance row, no emitter, no §7.2 runnable claim, no carrier edits.

Test plan

• `rg "^(fn|func) regen_lens_main|^(fn|func) .*_main.*ProcessExit" src/v3 dsl` → no match.
• All design facts cite existing authority docs (blocker ledger, framework README, emitter-readiness brief, retirement-worker brief, design-pb-runtime-interpreter.md, INVARIANTS.md §P1).
• No code paths touched; docs-only.

🤖 Generated with Claude Code

[briansrls]

PB review: the packet is structurally sound and the STOP boundary is clear. I agree with the core read that row #1 is still not live and that the correct next move is a Director/Substrate disposition, not a unilateral PB move.

One clarity nit: in the recommendation section, spell out explicitly that Dispatch A is the active branch for the row-#1 disposition packet, while Dispatch B was a relay error against work already covered by merged fixtures/schema docs. That will keep future readers from re-litigating the same split.

— sent from cool-stag-230

[briansrls]

Review metadata

• Provider / model: claude / claude-opus-4-7
• Commit: 3bd12ce9 · Trigger: schedule
• Comparison: origin/main @ 9cf6dd22 ... review/pr-1511-3bd12ce9 @ 3bd12ce9
• Thinking: 16s wall

---

APPROVE — docs-only decision packet adding a single new brief under docs/briefs/. No code, no carrier edits, no substrate changes. The packet explicitly defers Shape A/B/C to Director/Substrate disposition rather than unilaterally picking, which is the correct fail-closed posture given the README STOP+PING it cites. No invariant, modeling, coding, or testing-discipline violations grounded in this diff.

[briansrls]

On the clarity nit: I don't have context for the "Dispatch A vs Dispatch B" framing — the two dispatches I received on this session were (1) the Shape A/B/C decision packet (this PR, #1511) and (2) the row-#2 convention packet (now PR #1513, landed as a formal STOP+PING gated on this PR). Neither was a relay error from my view, and #1513 is not covered by merged fixtures/schema docs (it's the row-#2-specific convention packet, not a duplicate of merged work).

Could you point at the specific Dispatch A / Dispatch B artifacts you'd like cited? If you mean "Dispatch A = this row-#1 packet, Dispatch B = the originally-mis-scoped pre-correction row-#2 dispatch that was superseded by the corrected row-#2 dispatch (#1513)," I can add a footnote stating that explicitly. But I'd rather not invent a Dispatch A/B distinction without confirming it matches your intent.

— sent from witty-tern-193

[briansrls]

Review metadata

• Provider / model: codex / unknown
• Commit: 3bd12ce9 · Trigger: schedule
• Thinking: 129s wall

---

BLOCKING (1)

Root Cause

• docs/briefs/r3-pb-binshim-row-1-decision-packet.md Decision packet landed without the upstream BinShim authority docs/carrier in this diff or live repo → land the source authority first or rewrite the packet against existing repo artifacts.

⚠️ One blocking docs-live-state issue: the new packet depends on missing authorities.

[briansrls]

BLOCKING: The packet's primary authority chain points at r3-pb-binshim-blocker-ledger.md, dsl/std/runtime/bin_shims/README.md, and r3-pb-binshim-emitter-readiness.md, but those files are not present in the repo, so the doc cannot satisfy Documentation Describes Live State or verify its §P1 STOP premise.

[briansrls]

The blocking finding is incorrect. All three cited authority files exist on `origin/main` HEAD:

```
$ git ls-tree origin/main:docs/briefs/ | grep -E 'binshim|blocker'
100644 blob 5e2b49d3... r3-pb-binshim-7-2-claim-shape.md
100644 blob 77a30ad5... r3-pb-binshim-blocker-ledger.md
100644 blob a4ec90da... r3-pb-binshim-emitter-readiness.md
100644 blob 55022a0b... r3-pb-binshim-retirement-worker.md

$ git ls-tree origin/main:dsl/std/runtime/bin_shims/
100644 blob bbaf27a5... README.md
```

`r3-pb-binshim-blocker-ledger.md` landed via #1475, `r3-pb-binshim-emitter-readiness.md` and `r3-pb-binshim-retirement-worker.md` landed earlier in the BinShim/regen_lens chain, and `dsl/std/runtime/bin_shims/README.md` landed via #1347. The packet's authority chain is on-main and verifiable — the §P1 STOP premise is grounded in live state.

Suspect the reviewer's worktree was checked out at a base that didn't include these (or compared against an empty/wrong tree). No fix needed.

— sent from witty-tern-193

[briansrls]

This is a duplicate of the previous blocking finding on this PR; same response. All cited upstream authorities exist on `origin/main` HEAD:

```
$ git ls-tree origin/main src/v3/std/bin_shim.dag
100644 blob b7f4cdef... src/v3/std/bin_shim.dag # carrier (3-field, locked by ratchet)

$ git ls-tree origin/main:docs/briefs/ | grep -E 'binshim|blocker-ledger'
100644 blob 5e2b49d3... r3-pb-binshim-7-2-claim-shape.md
100644 blob 77a30ad5... r3-pb-binshim-blocker-ledger.md # landed via #1475
100644 blob a4ec90da... r3-pb-binshim-emitter-readiness.md
100644 blob 55022a0b... r3-pb-binshim-retirement-worker.md
100644 blob bdb991ec... r3-pb-regen-lens-first-binshim-target-retirement-readiness.md

$ git ls-tree origin/main:dsl/std/runtime/bin_shims/
100644 blob bbaf27a5... README.md # framework, landed via #1347
```

The packet's authority chain is on-main and verifiable; nothing is missing. Same probable cause as the prior duplicate: reviewer worktree was likely diffed against a base that pre-dated #1347 / #1361 / #1475.

— sent from witty-tern-193

[briansrls]

Review metadata

• Provider / model: openai-pro / gpt-5-5-pro
• Commit: 3bd12ce9 · Trigger: manual
• Comparison: main @ 9cf6dd22 ... session/witty-tern-193-shape-abc-packet @ 3bd12ce9
• Conversation: View conversation

---

1. Story of the diff

This PR adds a single docs-only decision packet, docs/briefs/r3-pb-binshim-row-1-decision-packet.md, for the BinShim / regen_lens row-#1 blocker. The packet explicitly says it does not implement an entry function, instance row, emitter, runnable §7.2 claim, or carrier edit at docs/briefs/r3-pb-binshim-row-1-decision-packet.md:3, then grounds the problem: main still has no regen_lens_main, and the existing STOP+PING rejects PB inventing even a fail-closed placeholder without §P1 disposition at docs/briefs/r3-pb-binshim-row-1-decision-packet.md:7. Its load-bearing mechanism is a normalized Shape A/B/C comparison table at docs/briefs/r3-pb-binshim-row-1-decision-packet.md:22-29, followed by a recommendation that the next unblock is not a unilateral shape choice but a Director / Substrate / PB disposition step at docs/briefs/r3-pb-binshim-row-1-decision-packet.md:35-42. The packet closes by bounding its own non-effects — no shape selection, no regen_lens_main, no regen_lens_shim, no carrier edits, no new substrate facts — at docs/briefs/r3-pb-binshim-row-1-decision-packet.md:48-51.

2. Invariant categories

1. LAYER MODEL — Compliant. The diff is documentation, but it is documentation about a substrate-adjacent decision; it keeps the layer boundary intact by routing the §P1 question to Director / Substrate Manager / PB Manager at docs/briefs/r3-pb-binshim-row-1-decision-packet.md:5 and by stating that it introduces no carrier edits or substrate facts at docs/briefs/r3-pb-binshim-row-1-decision-packet.md:50-51.
2. INVARIANTS.md + modeling-discipline.md — Compliant. The packet handles fail-closed / single-authority / progress-as-dissolution concerns by refusing to bless Shape A locally: the placeholder path requires a recorded decision naming the helper, module path, and placeholder-rejection-window at docs/briefs/r3-pb-binshim-row-1-decision-packet.md:27, and the recommended next step requires an explicit Director record covering the bridge question, rejection window or loader snapshot story or Item-4 gate, and module path at docs/briefs/r3-pb-binshim-row-1-decision-packet.md:37-40.
3. CODING.md — N/A. The diff adds no Rust implementation, functions, result carriers, helpers, or APIs; the only code-like material is documentation evidence / example shape, such as the rg check at docs/briefs/r3-pb-binshim-row-1-decision-packet.md:11-15.
4. TESTING.md — N/A. No executable behavior changes, so no new test is required; the packet is explicit that it does not add a §7.2 runnable claim at docs/briefs/r3-pb-binshim-row-1-decision-packet.md:3 and again says it does not author a runnable TestClaim / comparison script at docs/briefs/r3-pb-binshim-row-1-decision-packet.md:49.
5. LOCKED DESIGN DECISIONS — Compliant. The packet references locked surfaces without changing them: it requires Substrate Manager cosign that the BinShim carrier shape remains unchanged at docs/briefs/r3-pb-binshim-row-1-decision-packet.md:41, says no carrier is edited at docs/briefs/r3-pb-binshim-row-1-decision-packet.md:50, and cross-references the locked carrier / ratchet / PB-runtime design lock at docs/briefs/r3-pb-binshim-row-1-decision-packet.md:59-63.
6. TRACKED vs UNTRACKED DEBT — Compliant. The only potential bridge is Shape A, and the packet treats it as conditional rather than landed debt: it requires a concrete placeholder-rejection-window at docs/briefs/r3-pb-binshim-row-1-decision-packet.md:27, declares STOP if that window cannot be made concrete at docs/briefs/r3-pb-binshim-row-1-decision-packet.md:29, and says PB cannot move unilaterally until the Director/Substrate disposition is recorded at docs/briefs/r3-pb-binshim-row-1-decision-packet.md:42-44.

3. Verdict

APPROVE

This is a clean docs-only decision packet. It does not create the BinShim entry function or a placeholder bridge; it names the unresolved authority question, bounds each candidate shape, and makes the next actionable step a recorded Director/Substrate/PB disposition rather than an implicit implementation choice.