feat: build step 2 — the home page; close Q35, Q37, Q39; Q39's answer corrected the register

Pouya's rulings and the page are one commit, not two, because they are not
separable: Q37 changed the credential label the page renders, Q39 scoped the
arbitration copy, and Q35 gave Med-Arb the row that lets the footer link stand.
Splitting them would produce a commit where the register says one thing and the
page says another.

RULINGS

Q39 — ANSWERED, and my assertion was FALSE as a universal. Pouya checked
rather than defended. Family arbitration in Ontario is gated: prescribed
training, 14h on screening for domestic violence and power imbalances, 30h of
Ontario family law for arbitrators outside the bar, 10h ongoing per two years.
claims-auditor produced that counter-example from inside the repo on 2026-08-26
and it was right. The source is now IN the repo per R14 —
docs/reference/ontario-family-arbitration-training.md, retrieved with its
provenance — and §4 states which half it establishes: the family gate directly,
the commercial half only by absence. So "commercial arbitration in Ontario
requires no licence and no designation" is recorded as Pouya's stated position,
unstamped. What disposes of the question is the scope exclusion: he does not do
family arbitration, so it is NOT OFFERED and the gate never bears on the
practice.

The false universal survived in three more live places, found by grep, not
recall: §4's NOT-NEGOTIABLE boundary bullet (the boundary paragraph breaching
its own boundary), §9's Q33 closure row, and a comment in SiteHeader.astro.

Q35 — ANSWERED, all four items, and the answer supplied a gate that was
missing. Med-Arb gets an Offerings row and stays in the footer sitemap. The six
subject-matter labels pass test 1. §4 gains "Subject-matter areas — the
publication gate": competence to accept an appointment, AND a page that frames
it as positioning without claiming history. Nothing in the nav or footer is
unrowed any more.

Q37 — ANSWERED. "Law and engineering" becomes "Legal training and engineering
practice". A degree is not a practice; the parallel was doing the implying.

Q38/R13 — the committed SVG does not close it; the walk-back went too far. It
renders faithfully BECAUSE it is the raster. R13 stays open.

R1 — surfaced and acknowledged; the interim licensure framing is now carried by
a shipped page rather than a spec, which raises the stakes.

BUILD STEP 2

Seven of docs/01's eight home sections, six new components, zero JavaScript.
Section 7 (Latest insights) ships at step 7 with the collection it lists —
recorded in docs/01 and in the page, not just here. Four credential slots, not
three: §4's paired-disclosure condition requires the Q.Arb stage on any page
that offers arbitration. No booking link (R6). The masthead tagline is
suppressed on / (it duplicated the hero eyebrow). The step-1 proof sheet is
deleted, and five live references to it were found by grep.

WHAT THE REVIEWS FOUND — 24 findings across the two passes

claims-auditor FAILED it on ten, every one implication or scope rather than
fabrication, which is where D13 says the risk lives. The four that mattered:
"I mediate and arbitrate" asserted a track record §4 does not hold for
arbitration; the JSON-LD asserted arbitration twice and stated the stage
nowhere (a crawler-only claim is still a claim); "at one published rate with
preparation time included" misdescribed money against docs/07's two day rates
and capped prep allowance; and "Law and engineering are not blended here" was
Q37's struck parallel relocated into prose one day after Pouya struck it.
It also found the Canadian Tax Foundation missing from §9 Q8 — a stale second
copy of a fact that would have bitten at step 3.

adversarial-reviewer found fourteen, three blocking. The first was
class="section-head" on <SectionHeading> never matching — the parent-cannot-
style-a-child defect for the FOURTH time, written into a diff where I had just
added fresh warnings about it to two other components. Measured: 0px margin,
0px gap, headings over the card edges, with astro check and eslint both clean.
I had looked at a screenshot of that section and passed over it. Fixed with a
page-owned wrapper (48px, measured) and the prop is deleted from six components
so it is now a build error. Also: the credential row was never "two up on a
phone" and its comment said it was; PROCESS was hardcoded in the page against
the reason written in site.ts; 83px of residual overflow at a 200% default font
size, now 3px.

Seven more I found myself first, including <Picture widths> declaring the
untouched 1600px master as the <img src> fallback (254,626 B for a 476px slot,
and the build log said "before: 349kB" either way), and a prop named `as`
silently turning off type-checking for a whole component.

VERIFICATION — run, not asserted. Full figures in AGENTS.md entry (w).

npx tsc 6.0.3 · check 0/0/0 · lint clean · build clean · audit 0
1 <script> and it is JSON-LD; non-JSON-LD scripts 0; no JS bundle; identical
page with script execution disabled (444 nodes, 6,578 chars)
Phase 5 minifier check: no `animation` shorthand beside animation-timeline
overflow 0 at 14 widths, AND every mark measured at 1.5654-1.5657 vs 1.5657
one h1, no heading skips, focus order == DOM order across 44 focusables
31 painted contrast pairs at 3 widths, 0 failures
72/72 hit-test points across 8 cards resolve to the card's link
print 0 hidden, reduced-motion 0 hidden
Lighthouse NOT RUN — tool unavailable until step 7 (R11)
HTML validator NOT RUN, and 4 of 12 srcsets carry a duplicate 1x descriptor

Opened for Pouya: Q40 (one OG image for nineteen pages), Q41 (does Q37 reach
prose; may the comparative be restored; what LAT pre-hearing mediation means),
Q42 (the four "also offered" processes have no row), Q43 (the process timings
are published commitments with no row).

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_0148NztQskLKKApP5SzAA78e

docs/02-design-system.md

@@ -183,7 +183,7 @@ except the reveal of the hero.
 
 | Component | Notes |
 |---|---|
-| `InfinityMark` | ⚠️ **Currently a raster — a documented, temporary exception to this rule** (`AGENTS.md` Q38, R13). The mark is gradient-mesh artwork rather than flat vector paths. An SVG *is* held and it renders faithfully — but at 257,278 bytes against 3,063 for the AVIF a Retina browser takes, carrying seven embedded base64 PNGs that would breach `CLAUDE.md`'s no-base64-images rule. The exception is about payload and composition, not about fidelity. It renders AVIF/WebP with a PNG fallback; a Retina device takes 3,063 B of AVIF, `alt=""` when decorative, aspect ratio pinned to `667 / 426`. **Restore this rule the moment the commissioned vector master lands.** Until then the rule stands unchanged for every other mark or icon: inline SVG, `currentColor`, `aria-hidden` when decorative, never a PNG |
+| `InfinityMark` | ⚠️ **Currently a raster — a documented, temporary exception to this rule** (`AGENTS.md` Q38, R13). The mark is gradient-mesh artwork rather than flat vector paths. An SVG *is* held and it renders faithfully — **and it renders faithfully because it IS the raster**: 257,278 bytes wrapping **seven embedded base64 PNGs**, against **3,063 B** for the AVIF a Retina browser takes in the header — **84×**. *Figures re-stated 2026-08-27 because a second, larger call site now exists and the single number had become misleading:* the mark ships at **two intrinsic widths**, 64 px (header, footer) and 232 px (the home page's approach section, which renders at 225.5 px). At 232 px a DPR-2 device takes **14,555 B** and DPR-3 **22,639 B**, so the ratio against the SVG there is ~11×, not 84×. The full ladders are in `docs/reference/brand-assets.md`; do not quote one number as if it covered both. Inlining it would breach `CLAUDE.md`'s no-base64-images rule. *Pouya settled the characterisation on 2026-08-27: a first draft called the file "a raster in a vector wrapper", a later draft withdrew that as unfair, and **the withdrawal went too far.** Both halves are true — the fidelity is real, and it is bought with embedded raster, which is exactly why fidelity was never the question.* The exception is about payload and composition. It renders AVIF/WebP with a PNG fallback; a Retina device takes 3,063 B of AVIF, `alt=""` when decorative, aspect ratio pinned to `667 / 426`. **Restore this rule the moment the commissioned vector master lands.** Until then the rule stands unchanged for every other mark or icon: inline SVG, `currentColor`, `aria-hidden` when decorative, never a PNG |
 | `SiteHeader` | **Sticky from 66 rem (1056 px) up; static below it.** Practice dropdown as CSS-only `<details>`. **(a)** The one-row header holds the brand, **seven** nav items and the CTA — Insights is the seventh, arriving on its own at build step 7. Binary search on the built page puts the true fit at **1047 px**; 66 rem is the clean token above it. Below 66 rem the nav takes its own row and the header measures **141 px** at tablet widths and **189 px** at 320–375 px. **(b)** "Condenses on scroll" is a hairline rule and a shadow, **not a size change** — a `position: sticky` header stays in normal flow, so shrinking its padding lifts every page below it, against the CLS < 0.05 budget. Longhands only; see the component on what the minifier does to the `animation` shorthand. *(This row has been wrong twice, instructively. It first said 60 rem / "~115 px", a height the header never took. It then said 64 rem "with 32 px of clearance" — that 32 px was `.header-inner`'s own `column-gap` mistaken for slack; the real figure at 1024 px with seven items was **−21.6 px**, and nothing overflowed only because flexbox crushed the logo inside the brand block. **Measure slack, not gaps.**)* |
 | `SiteFooter` | Three-column sitemap, contact block, designations, entity line |
 | `Eyebrow` | Mono label with optional maroon dot |
@@ -216,5 +216,44 @@ Not a polish pass. A build requirement.
 - Forms: real `<label>` elements, `aria-describedby` for hints, errors announced
   with `role="alert"` and tied to their field.
 - Touch targets ≥ 44 × 44 px.
-- Test at 200% zoom and at 320 px width.
+- Test at 200% zoom and at 320 px width. **Both measured 2026-08-27 on `/`:
+  document overflow 0 at 320, 360, 390, 414, 640, 768, 900, 1024, 1056, 1200,
+  1216, 1280, 1440 and 1920 CSS px, with zero elements extending past the
+  viewport.** Page zoom at 200% of 1280 is the 640 column and at 400% is the
+  320 column, so WCAG 1.4.4 and 1.4.10 are both covered by that sweep.
+
+  **A stricter case is not fully clean, and it is recorded rather than left to
+  be discovered.** With the reader's *default font size* at 200% — root at
+  32 px, a real accessibility setting and not page zoom — `/` measured **234 px**
+  of overflow at 390. Brought down in three measured steps:
+
+  | Fix | 390 px | 320 px |
+  |---|---|---|
+  | as first built | 234 px | 304 px |
+  | `minmax(min(Nrem, 100%), 1fr)` on three grids | 83 px | 153 px |
+  | `.credentials` made explicit `repeat(2, minmax(0, 1fr))`; `.feature` padding clamped and `overflow-wrap: anywhere` on its title; `.contact-action` `flex: 0 1 auto` + `min-inline-size: 0` | **3 px** | **65 px** |
+
+  Two things worth keeping. **`overflow-wrap: break-word` permits a break at
+  layout time but does not reduce min-content size** — `anywhere` does, and that
+  distinction was the whole of one of those fixes. And 1280 px stays **602 px**
+  over, from the header's deliberate `flex-wrap: nowrap` above 66 rem plus
+  `white-space: nowrap` on the brand name; the 320 px residual is the same header
+  plus the display headline's 104 px floor. Undoing either re-opens the measured
+  step-1 header decision, so they stand. All of this is beyond what this floor
+  requires — page zoom is clean — so it is a robustness margin rather than a
+  failure. Revisit if a real reader hits it.
+
+- **Measure the elements, not only the page.** A document-level overflow check
+  passes while a flex child absorbs the deficit by being crushed — that is how
+  step 1 shipped a logo at aspect 0.891 under a green "0 overflow at every
+  width". Assert the rendered geometry of the thing you care about. On `/`
+  every infinity-mark instance measures 1.5654–1.5657 against the master's
+  1.5657.
+- **A touch-target measurement of the wrong box is not a finding.** The eight
+  cards on `/` report 26–39 px-tall `<a>` elements and are fine: each card's
+  whole box is the link's hit area via `::after { inset: 0 }`, verified by
+  hit-testing nine points per card at three widths (24 cards, 9/9). Hit-test
+  before enlarging anything. WCAG 2.5.8's inline exception also applies to a
+  link sitting mid-sentence, and one such link on `/` is deliberately left at
+  164 x 21.
 - Every page must be readable and navigable with JavaScript disabled.