fix(v3): remove stale bin_shim_carrier_test.rs ref from SG-0 expected

[briansrls]

Summary

Post-merge inconsistency from back-to-back merges:

• fix(v3): refresh BinShim ratchets after carrier merge #1372 at 05:52:48Z added bin_shim_carrier_test.rs to EXPECTED_HAND_AUTHORED_TEST in sg0_census_test.rs
• docs(audit): verify numeric construction algebra surfaces #1370 merged 4 seconds later at 05:52:52Z, including a "move BinShim carrier ratchet into existing module" commit that deleted src/v3/compiler/tests/integration/bin_shim_carrier_test.rs (-80 lines) and consolidated content into m2_substrate_inhabitance_test.rs (+63 lines)

The expected list now references a file that no longer exists, breaking SG-0 census on every PR after the back-to-back merges. Verified at HEAD:

$ git ls-tree origin/main -- src/v3/compiler/tests/integration/bin_shim_carrier_test.rs
(empty)
$ git grep -n "bin_shim_carrier_test" origin/main -- src/v3/compiler/tests/integration/sg0_census_test.rs
src/v3/compiler/tests/integration/sg0_census_test.rs:235:    "src/v3/compiler/tests/integration/bin_shim_carrier_test.rs",

This 1-line removal restores census consistency. The carrier-test ratchet content itself stays alive in m2_substrate_inhabitance_test.rs per the consolidation move.

Test plan

• CI green (sg0_census_test::sg0_v3_test_hand_authored_subratchet stops failing)
• CI green on parse_stage4_prep::handwritten_parse_snapshot_matches_manifest (unaffected — parse manifest doesn't reference the file)

🤖 Generated with Claude Code

[briansrls]

Review metadata

• Provider / model: cursor / composer-2
• Commit: fb060727 · Trigger: schedule
• Comparison: origin/main @ 1cbc02d2 ... review/pr-1384-fb060727 @ fb060727
• Thinking: 53s wall

---

Findings: None. The diff only drops bin_shim_carrier_test.rs from EXPECTED_HAND_AUTHORED_TEST in sg0_census_test.rs (hunk at sg0_census_test.rs:235 in the unified diff). There is no substrate or compiler logic change, so INVARIANTS.md and docs/modeling-discipline.md do not apply in a meaningful way. It does not conflict with CODING.md. Under TESTING.md, this aligns the SG-0 hand-authored census list with the tree: bin_shim_carrier_test.rs is not present in the worktree, so keeping it in the expected set would be the inconsistent behavior.

Verdict: APPROVE — Single-line maintenance fix; expectations match reality, no rubric violations in the diff.

[briansrls]

LGTM — clean post-merge-race fix.

Verified the back-to-back merge artifact:

• fix(v3): refresh BinShim ratchets after carrier merge #1372 (05:52:48Z) added bin_shim_carrier_test.rs to EXPECTED_HAND_AUTHORED_TEST
• docs(audit): verify numeric construction algebra surfaces #1370 (05:52:52Z, +4s) consolidated the carrier-test content into m2_substrate_inhabitance_test.rs and deleted the standalone file

Both PRs are individually correct; the race left EXPECTED_HAND_AUTHORED_TEST referencing a deleted file. The 1-line deletion in this PR resolves the drift cleanly:

-    "src/v3/compiler/tests/integration/bin_shim_carrier_test.rs",

Carrier-test ratchet content survives in m2_substrate_inhabitance_test.rs (per #1370's consolidation), so the substrate-acceptance signal is preserved.

This unblocks SG-0 census on main and unblocks downstream PRs (incl. my docs-only #1369) that have been waiting on the post-merge-race repair.

— sent from deep-wolf-155

[briansrls]

Review metadata

• Provider / model: codex / unknown
• Commit: fb060727 · Trigger: schedule
• Thinking: 99s wall

---

Non-blocking — Strengths

• src/v3/compiler/tests/integration/sg0_census_test.rs Classification: code/test infrastructure; removing the absent bin_shim_carrier_test.rs entry follows the SG-0 missing-expected-file shrinkage path.

ROADMAP — Verified

• PB-0 SG-0 ratchet: The change aligns with the pure-bootstrap ratchet direction by reducing the expected hand-authored Rust census instead of adding a new authority.

✅ No blocking concerns; targeted verification was blocked by an existing compile error in src/v3/compiler/src/infer.rs outside this PR.

[briansrls]

Review metadata

• Provider / model: openai-pro / gpt-5-5-pro
• Commit: fb060727 · Trigger: manual
• Conversation: View conversation

---

1. Story of the diff

This PR updates the SG-0 census expectation list so it no longer names src/v3/compiler/tests/integration/bin_shim_carrier_test.rs. The failure mode being fixed is not a runtime or substrate behavior bug; it is a stale ratchet entry inside EXPECTED_HAND_AUTHORED_TEST, where the expected hand-authored Rust test inventory still included a file that should no longer be counted. The chosen mechanism is minimal and appropriate: delete the obsolete path from src/v3/compiler/tests/integration/sg0_census_test.rs:235 while leaving the surrounding census entries and comments intact.

2. Invariant categories

1. LAYER MODEL — N/A. The diff only edits a Rust integration-test census expectation; it does not touch substrate declarations, Dag-resident types, cross-pass carriers, or dag.rs.
2. INVARIANTS.md + modeling-discipline.md — Compliant. Progress Is Dissolution / tracked residual discipline is served by removing the stale SG-0 residual entry at src/v3/compiler/tests/integration/sg0_census_test.rs:235, rather than preserving an obsolete bridge/test path as expected debt.
3. CODING.md — Compliant. The change is explicit static test data maintenance: deleting the stale string literal at src/v3/compiler/tests/integration/sg0_census_test.rs:235; it adds no hidden state, helper API, object method, fallback path, or new naming surface.
4. TESTING.md — Compliant. The PR directly updates the census test’s expected inventory by removing "src/v3/compiler/tests/integration/bin_shim_carrier_test.rs" at src/v3/compiler/tests/integration/sg0_census_test.rs:235; no additional regression test is needed because the edited expectation is the ratchet being repaired.
5. LOCKED DESIGN DECISIONS — N/A. The diff does not reference or alter thesis/design-doc locked decisions; it only reconciles a test inventory literal.
6. TRACKED vs UNTRACKED DEBT — Compliant. No new scaffold, TODO, bridge, shim, or temporary shape is introduced; the only debt-related movement is deleting one stale tracked residual path at src/v3/compiler/tests/integration/sg0_census_test.rs:235.

3. Verdict

APPROVE

The PR is a narrow SG-0 census correction that removes obsolete expected debt without introducing new substrate shape, implementation behavior, or test scaffolding. I found no blocking or non-blocking findings in the diff.