## [Claude #4] Stage 6 — Verification Round #4

### Re-read issue
Scope = `build_layout_css` 분기 확장 to **top-1-bottom-2 / top-2-bottom-1 / left-1-right-2 / left-2-right-1 / grid-2x2** (issue text: "현재 horizontal-2 / vertical-2 만 dynamic — 다른 preset 도 dynamic").
Guardrail = Kei `build_containers_type_b` (page_structure) 회귀 X + zone-level ratio 변경 시 Step 8 trace 설명 가능.

### Full pytest
```
python -m pytest -q tests
89 passed in 6.95s
```
Phase Z2 subset: 47/47 passed.

### Before (Round #3) vs After (Round #4)
- `git rev-parse --short HEAD` = **201099e** (unchanged from Rounds #1–#3).
- `git diff HEAD -- src/phase_z2_pipeline.py src/space_allocator.py tests/phase_z2/` → **empty** (no implementation movement since Round #3 closed NOT VERIFIED).
- Dirty/untracked worktree files are all outside the verifier scope (Z2 frame templates, samples, unrelated scripts).

### AS-IS (HEAD = 201099e — PR 1 foundation)

**Static evidence — dispatch is still rows / cols only** (src/phase_z2_pipeline.py:1214–1221):
```python
# ── Dynamic branch — topology dispatch (PR 1: rows / cols only) ──
if topology == "rows":
    return _build_rows_dynamic(preset, zones_data, gap)
if topology == "cols":
    return _build_cols_dynamic(preset, zones_data, gap)
# PR 2 will dispatch T / inverted-T / side-T-{left,right} / 2x2 here.
# PR 3 will dispatch single here.
return _build_fr_default(preset)
```

Override path (src/phase_z2_pipeline.py:1205–1212) still warn-and-fallthrough for non rows/cols presets (PR 1 lock — PR 2 promotes to strict ValueError).

**Runtime probe** — `build_layout_css(preset, {z1..z4: {content_weight:1.0, min_height_px:100}})`:

| Preset | dynamic_rows | dynamic_cols | computation |
|---|---|---|---|
| `top-1-bottom-2` | False | False | `fr_default_from_preset` |
| `top-2-bottom-1` | False | False | `fr_default_from_preset` |
| `left-1-right-2` | False | False | `fr_default_from_preset` |
| `left-2-right-1` | False | False | `fr_default_from_preset` |
| `grid-2x2` | False | False | `fr_default_from_preset` |
| `single` | False | False | `fr_default_from_preset` |

**Test evidence — PR 1 lock-in tests still in place** (tests/phase_z2/test_build_layout_css_pr1.py:145–154):
```python
"""T-shape (top-1-bottom-2) falls through to fr_default in PR 1 …"""
result = build_layout_css("top-1-bottom-2", zones)
assert result["computation"] == "fr_default_from_preset"
```
Closing B-4 requires updating this assertion when PR 2 lands.

### TO-BE (issue scope target)
All 5 listed presets (top-1-bottom-2 / top-2-bottom-1 / left-1-right-2 / left-2-right-1 / grid-2x2) must be dynamic. Currently **0/5 dynamic**. PR 1 (HEAD) ships the foundation only — col-axis solver, unified per-zone geometry mapper, length-locked widths/widths_ratios/dynamic_cols on every return path, retry gate (with `retry_skipped_reason`), and `zone_widths_px_planned` / `zone_col_ratios_planned` Step 8 trace fields. PR 2 (T / inverted-T / side-T-{left,right} / 2x2) + PR 3 (single) remain outstanding per the code comments at lines 1219–1220.

### Guardrail check
`build_containers_type_b` / `page_structure` paths untouched — `git diff HEAD -- src/space_allocator.py` is empty since PR 1 merge. No regression detected.

### Gaps vs Close

| Axis | Issue requirement | Current state | Gap |
|---|---|---|---|
| top-1-bottom-2 dynamic | required | fr_default fallthrough | NOT MET |
| top-2-bottom-1 dynamic | required | fr_default fallthrough | NOT MET |
| left-1-right-2 dynamic | required | fr_default fallthrough | NOT MET |
| left-2-right-1 dynamic | required | fr_default fallthrough | NOT MET |
| grid-2x2 dynamic | required | fr_default fallthrough | NOT MET |
| build_containers_type_b guardrail | preserved | preserved (no diff) | OK |
| Step 8 trace explainability | required | partial — `zone_widths_px_planned` ships but only populated on rows/cols dynamic returns | partial |
| Override path strict ValueError | PR 2 contract | still warn-and-fallthrough | NOT MET |

### Conclusion
B-4 issue scope is **NOT closed by HEAD 201099e**. PR 1 is a deliberately-scoped foundation lock; the topology dispatch + override strict-mode promotion for the 5 listed presets is the binding deliverable of PR 2 (with `single` deferred to PR 3). Four consecutive verification rounds (Claude #1–#3 + Codex #1–#3 + this round) agree on this AS-IS reading; no implementation movement has occurred between rounds.

### Evidence block
Commands run:
- `git rev-parse --short HEAD` → `201099e`
- `git status --short` → dirty/untracked items are all out-of-scope (Z2 templates, samples, ad-hoc scripts)
- `git diff HEAD -- src/phase_z2_pipeline.py src/space_allocator.py tests/phase_z2/` → empty
- `grep -n "falls through\|fr_default_from_preset\|top-1-bottom-2\|PR 1" tests/phase_z2/test_build_layout_css_pr1.py` → PR 1 lock-in test confirmed at lines 145–154
- Runtime probe via `python -c "from src.phase_z2_pipeline import build_layout_css …"` on all 6 presets (results above)
- `python -m pytest -q tests` → 89 passed in 6.95s

Files checked:
- D:\ad-hoc\kei\design_agent\src\phase_z2_pipeline.py (lines 1195–1221 dispatch block)
- D:\ad-hoc\kei\design_agent\tests\phase_z2\test_build_layout_css_pr1.py (lock-in assertions)
- D:\ad-hoc\kei\design_agent\src\space_allocator.py (guardrail surface — unchanged)

Verified facts:
- HEAD unchanged at 201099e across Rounds #1–#4.
- 0/5 target presets are dynamic in HEAD.
- PR 1 lock-in test still asserts top-1-bottom-2 → fr_default_from_preset.
- Kei build_containers_type_b / page_structure paths show no diff vs PR 1 merge.
- pytest 89/89 green; phase_z2 subset 47/47 green.

FINAL_CONSENSUS: NO