feat: build step 4 — /mediation/, /arbitration/, /med-arb/; source ADRIC's rules

Three pages, five in the build, zero JavaScript. /arbitration/ carries §4's
paired-disclosure condition on four surfaces and Q39's struck universal appears
in no form. /med-arb/ meets the procedural-fairness objection at the level of
process design and ships deliberately without Pouya's own protocol commitments,
which are Q54.

docs/01 directed the mediation page to name the "ADRIC Model Mediation Rules".
No such document exists — 0 occurrences across all four of ADRIC's rules pages
against 10 for "National Mediation Rules"; "Model" belongs to the Model Dispute
Resolution Clause inside the rules. Caught only because R14 requires the source
before the claim. docs/reference/adric-rules.md + adric-extract/ carry it, with
the digest drift measured rather than assumed: the HTML changes per request, the
text extracts are byte-stable, so the extracts are the artefact.

Four review passes, 21 defects, and the pattern was mine: I wrote the Q54 gate
into the page and then breached it four times, then round 2 found two survivors
of round 1's own fixes and one defect round 1's fix created. Also removed a
<title> naming a practised role §4 does not grant, a habitual presupposing awards
issued, and a claim about what ADRIC's rules permit that my own reference doc
says is unsupported.

Two instrument failures caught before they became conclusions: touch targets
measured over file:// with no CSS loaded (uniform 18px, including on a .btn with
a 44px floor), and a schema.org validator call that parsed nothing and returned
0 warnings for everything. Both re-run with the instrument validated first.

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

src/components/DefinitionGrid.astro

@@ -0,0 +1,63 @@
+---
+/**
+ * A `<dl>` of term/description pairs on an auto-fitting grid.
+ *
+ * EXTRACTED AT STEP 4 ON `adversarial-reviewer`'S FINDING, and the finding was
+ * that `/mediation/`'s `.formats` and `/arbitration/`'s `.cols` were the same
+ * component under two names — identical markup, near-identical CSS, and
+ * `.cols` was already serving two different content types on one page.
+ * `/practice/*` at step 5 wants it a fourth time. Same argument that extracted
+ * `ContactBand`: two call sites, one already divergent, pages to come.
+ *
+ * `<dl>` RATHER THAN A DIV GRID, for `CredentialRow`'s reason: each pair is a
+ * term and its description, so a screen reader gets them as an associated pair
+ * rather than as a visual arrangement. Wrapping each `<dt>`/`<dd>` in a `<div>`
+ * inside the `<dl>` is valid HTML and is what makes the grid tractable.
+ *
+ * A `<dt>` IS NOT A HEADING and must not become one. These sit under the
+ * section's `<h2>`; promoting them to `<h3>` would be a heading level that adds
+ * nothing to the outline, and `docs/02` forbids skipped levels either way.
+ */
+interface Props {
+  items: ReadonlyArray<{ name: string; body: string }>;
+  /** The grid's per-column floor, passed to `.grid-autofit` as `--grid-min`.
+   *  The `min(N, 100%)` guard lives there, in one place. */
+  minColumn?: string;
+}
+const { items, minColumn = '17rem' } = Astro.props;
+---
+
+<dl class="grid-autofit defs" style={`--grid-min: ${minColumn}`}>
+  {
+    items.map((item) => (
+      <div class="def">
+        <dt class="def-name">{item.name}</dt>
+        <dd class="def-body">{item.body}</dd>
+      </div>
+    ))
+  }
+</dl>
+
+<style>
+  /* Columns and the `min()` guard come from `.grid-autofit` (global.css); this
+     sets only the gap. It re-implemented them for one pass — a second copy of
+     the guard, inside the extraction made to remove copies of the guard.
+     `--grid-min` is passed inline by the caller because a parent cannot style
+     this component's root, and a custom property is the one mechanism that
+     crosses that boundary. */
+  .defs {
+    gap: var(--space-7);
+  }
+  .def-name {
+    font-family: var(--font-mono);
+    font-size: var(--text-xs);
+    font-weight: var(--weight-medium);
+    letter-spacing: var(--tracking-wide);
+    text-transform: uppercase;
+    color: var(--text-meta);
+  }
+  .def-body {
+    margin-block-start: var(--space-2);
+    line-height: var(--leading-body);
+  }
+</style>

src/data/schema.ts

@@ -179,7 +179,21 @@ export function personNode(
  *
  * No `priceRange`, no `aggregateRating`, no `review` — the last two have no
  * underlying data and §4 Forbidden bars the fabricated testimonial that the
- * previous site carried.
+ * previous site carried. And no `availableLanguage`: it is not in this type's
+ * domain either (see `serviceGraph`), and the `Person` beside it in the graph
+ * carries `knowsLanguage`.
+ *
+ * ⚠️ **`serviceType` AND `provider` ARE STILL OUT OF DOMAIN ON THIS TYPE, AND
+ * THAT IS UNRESOLVED.** Measured 2026-08-28 against `validator.schema.org`:
+ * `/` returns `UNKNOWN_FIELD` for `serviceType`, `availableLanguage` and
+ * `provider` on `ProfessionalService`, which is a `LocalBusiness` and takes
+ * none of the three. `availableLanguage` is removed above because nothing is
+ * lost. The other two carry real information — the commercial scoping and the
+ * link to the Person — so removing them costs more than the warning does, and
+ * the modelled fix is a `Service` node with the business as `provider`, which
+ * is a change to the home page's structured data rather than a tidy-up. **The
+ * other four pages now validate with 0 warnings**; this node is the last one.
+ * Raised for the step-7 SEO pass, not left as folklore.
  */
 export function professionalServiceNode(imageUrl?: string) {
   return {
@@ -202,7 +216,6 @@ export function professionalServiceNode(imageUrl?: string) {
       'Commercial arbitration',
       'Mediation-arbitration (med-arb)',
     ],
-    availableLanguage: ['en', 'fa'],
     email: `mailto:${CONTACT.email}`,
     ...(imageUrl ? { image: imageUrl } : {}),
   };
@@ -250,3 +263,112 @@ export function aboutGraph(imageUrl?: string) {
     '@graph': [personNode(imageUrl, { memberships: true })],
   };
 }
+
+/**
+ * `Service` for the three process pages — `/mediation/`, `/arbitration/`,
+ * `/med-arb/` (build step 4). docs/04's structured-data table.
+ *
+ * THE PERSON NODE TRAVELS WITH IT, for the reason `homeGraph` gives: a `@graph`
+ * makes `provider: {'@id': …}` resolve inside this document rather than relying
+ * on a crawler fetching `/about/` and joining two. Same `@id` either way, so a
+ * consumer that does fetch both merges rather than duplicates.
+ *
+ * NO `BreadcrumbList`. These are one hop from the root and show no visible
+ * breadcrumb; docs/04 requires the markup to MATCH visible breadcrumbs, so
+ * emitting one would assert navigation the page does not show. Breadcrumbs
+ * begin at `/practice/<area>/` and `/insights/<slug>/`.
+ *
+ * NO `offers` AND NO `priceRange` until `/fees/` exists (build step 9) — same
+ * gate docs/04 puts on `ProfessionalService`.
+ *
+ * NO `availableLanguage` EITHER, AND THAT IS NOT AN OVERSIGHT. schema.org's
+ * `domainIncludes` for it is `ContactPoint`, `Course`, `LodgingBusiness`,
+ * `ServiceChannel`, `TouristAttraction` — **not `Service`**. It shipped here for
+ * one pass and `validator.schema.org` returned `UNKNOWN_FIELD` twice per page on
+ * all three, against **0 warnings** on `/about/`. The `Person` in the same
+ * `@graph` already carries `knowsLanguage: ['en','fa']`, so nothing is lost. The
+ * modelled alternative is `availableChannel → ServiceChannel → availableLanguage`,
+ * which is more machinery than the fact is worth.
+ *
+ * ⚠️ `serviceType` IS A UNION, NOT A STRING, AND THAT IS THE POINT. An
+ * unscoped `"Arbitration"` is Q39's struck universal — false as a universal,
+ * swept four times, reached a public page once — so it is not a value this
+ * function will accept. A comment asking a caller not to pass it is a warning; a
+ * union is a build error. Widening the union requires a §4 Offerings row, and
+ * the row is the only thing that should ever widen it.
+ */
+/** `docs/04`'s ratified `serviceType` strings — **not** §4's row labels, which
+ *  read "Arbitration — sole arbitrator (commercial)" and "Med-Arb —
+ *  mediation-arbitration". Adding a member needs a §4 Offerings row, but do not
+ *  copy the row's wording: doing that "verbatim" yields the unscoped
+ *  `Arbitration` this union exists to bar. */
+export type PublishableServiceType =
+  'Mediation' | 'Commercial arbitration' | 'Mediation-arbitration (med-arb)';
+
+export function serviceGraph(opts: {
+  path: string;
+  name: string;
+  serviceType: PublishableServiceType;
+  description: string;
+  imageUrl?: string;
+}) {
+  return {
+    '@context': 'https://schema.org',
+    '@graph': [
+      {
+        '@type': 'Service',
+        '@id': `${SITE.url}${opts.path}#service`,
+        name: opts.name,
+        serviceType: opts.serviceType,
+        description: opts.description,
+        url: `${SITE.url}${opts.path}`,
+        provider: { '@id': PERSON_ID },
+        areaServed: [
+          { '@type': 'City', name: 'Toronto' },
+          { '@type': 'AdministrativeArea', name: 'Ontario' },
+        ],
+      },
+      personNode(opts.imageUrl),
+    ],
+  };
+}
+
+/**
+ * `/med-arb/`'s graph — the Service, the Person, and a `FAQPage`.
+ *
+ * docs/04 lists `FAQPage` for this page and for `/for-parties/`, with the
+ * condition that matters: *"Only where the visible page genuinely is Q&A. Never
+ * fabricate questions to farm a rich result."* So the node is BUILT FROM THE
+ * SAME ARRAY THE PAGE RENDERS — a question cannot enter the structured data
+ * without appearing on the page, and the two cannot drift.
+ */
+export function medArbGraph(opts: {
+  faq: ReadonlyArray<{ q: string; a: string }>;
+  imageUrl?: string;
+}) {
+  const base = serviceGraph({
+    path: '/med-arb/',
+    name: 'Med-arb (mediation-arbitration)',
+    serviceType: 'Mediation-arbitration (med-arb)',
+    description:
+      'Mediation that converts to binding arbitration if the mediation does ' +
+      'not resolve the dispute. One neutral, both phases, agreed in writing ' +
+      'in advance. Commercial matters.',
+    imageUrl: opts.imageUrl,
+  });
+  return {
+    ...base,
+    '@graph': [
+      ...base['@graph'],
+      {
+        '@type': 'FAQPage',
+        '@id': `${SITE.url}/med-arb/#faq`,
+        mainEntity: opts.faq.map((item) => ({
+          '@type': 'Question',
+          name: item.q,
+          acceptedAnswer: { '@type': 'Answer', text: item.a },
+        })),
+      },
+    ],
+  };
+}

src/data/site.ts

@@ -50,18 +50,20 @@ export const CREDENTIALS = {
   designations: ['Q.Med (ADRIC / ADRIO)'],
   inProgress: ['Q.Arb — commenced August 2026'], // [verified 2026-08-26]
   /**
-   * "Chartered Med-Arbitrator" — ADRIO's own term
-   * (docs/reference/adrio-designations.md). This read "Chartered
-   * Mediator-Arbitrator" until 2026-08-28, which was wrong; Pouya caught it and
-   * it was his own error, carried from the strategy brief and never sourced.
+   * "Chartered Med-Arbitrator" — ADRIO's own term, and ADRIC's navigation label
+   * too (docs/reference/adrio-designations.md; docs/reference/adric-rules.md
+   * Finding 3, which corroborates it from a second body). Never
+   * "Mediator-Arbitrator".
    *
-   * UNCONSUMED AS OF 2026-08-28 — nothing imports `CREDENTIALS.goal`. `/about/`'s
-   * credential arc hand-types all three designations instead, which is the drift
-   * shape this repo keeps paying for (see ContactBand's 52ch/46ch divergence).
-   * It is corrected rather than deleted because `/med-arb/` at build step 4 is
-   * its natural consumer: either that page uses it, or this line comes out.
+   * `GOAL_NAME` IS THE STRING TO RENDER. This designation's name was wrong on a
+   * public page for the life of the file, and the reason it survived four review
+   * passes is that it existed in several hand-typed copies rather than one
+   * constant. `/arbitration/` and `/med-arb/` (step 4) render `GOAL_NAME`;
+   * `/about/`'s arc still hand-types it, which is the remaining copy.
    */
   goal: 'C.Med-Arb (Chartered Med-Arbitrator)',
+  /** Just the expansion, for prose. §11 Glossary is the authority for it. */
+  goalName: 'Chartered Med-Arbitrator',
   education: ['JD, Bond University'],
   certifications: [
     'Kompass Arbitration Certificate Program',

src/pages/arbitration.astro

@@ -0,0 +1,327 @@
+---
+/**
+ * `/arbitration/` — build step 4. Spec: docs/01 §`/arbitration/`, docs/03.
+ *
+ * ⚠️ THIS PAGE CARRIES §4's PAIRED-DISCLOSURE CONDITION, AND IT IS A CONDITION
+ * ON THE OFFERING ITSELF, NOT A DISCLAIMER BOLTED TO IT. Both halves ship or
+ * neither does:
+ *
+ *   1. All three forms are offered NOW — sole, party-appointed, co-arbitration
+ *      — in commercial matters. §4 Offerings rows all three.
+ *   2. The Q.Arb PATHWAY commenced August 2026, with C.Med-Arb as the endpoint.
+ *
+ * "Pathway", never "designation": a designation that "commenced in August 2026"
+ * reads as held since then, which §4 Forbidden bars. Pouya's instruction is that
+ * being open about the stage is the differentiator — so it is stated plainly and
+ * high on the page, not hedged and not buried.
+ *
+ * ⚠️ AND NOTHING HERE MAY SAY OR IMPLY THAT ARBITRATION IS UNGATED IN ONTARIO.
+ * Q39's struck universal — "anyone may be appointed an arbitrator in Ontario;
+ * nothing in law gates the role behind a designation" — was FALSE as a
+ * universal, has been swept four times, and reached a public page once. The
+ * page states what is offered and where the credentialing stands. It makes no
+ * claim about what the law requires of anyone.
+ */
+import BaseLayout from '../layouts/BaseLayout.astro';
+import Button from '../components/Button.astro';
+import ContactBand from '../components/ContactBand.astro';
+import DefinitionGrid from '../components/DefinitionGrid.astro';
+import CredentialRow from '../components/CredentialRow.astro';
+import Eyebrow from '../components/Eyebrow.astro';
+import Pill from '../components/Pill.astro';
+import SectionHeading from '../components/SectionHeading.astro';
+import { getImage } from 'astro:assets';
+import ogDefault from '../assets/og-portrait.jpg';
+import { serviceGraph } from '../data/schema';
+import { CREDENTIALS, CREDENTIAL_ROW, CREDENTIAL_ROW_ARB } from '../data/site';
+
+const ldImage = await getImage({
+  src: ogDefault,
+  format: 'jpeg',
+  width: 1200,
+  height: 630,
+});
+const graph = serviceGraph({
+  path: '/arbitration/',
+  name: 'Commercial arbitration',
+  serviceType: 'Commercial arbitration',
+  description:
+    'Sole-arbitrator, party-appointed and co-arbitration appointments in ' +
+    'commercial matters. Documents-only, expedited and full hearing tracks.',
+  imageUrl: new URL(ldImage.src, Astro.site).href,
+});
+
+/* §4 Offerings rows all three, each `[verified 2026-08-26 — Pouya, Q33/Q36]`,
+   each scoped commercial. Do not add a fourth without a row. */
+const APPOINTMENTS = [
+  {
+    name: 'Sole arbitrator',
+    body: 'One arbitrator, appointed by agreement or by the mechanism the contract names.',
+  },
+  {
+    name: 'Party-appointed',
+    body: 'Appointed by one side to a three-member tribunal, deciding with the other two.',
+  },
+  {
+    name: 'Co-arbitration',
+    body: 'Sitting with co-arbitrators, usually where the matter spans more than one discipline.',
+  },
+];
+
+/* Tracks are docs/01 §`/arbitration/` item 2. The flat fees behind the first
+   two are D14's card and live on `/fees/`; no figure appears here. */
+const TRACKS = [
+  {
+    name: 'Documents only',
+    body: 'No hearing. Written submissions, the documentary record, and an award. The right track where the dispute is about what the contract says rather than about what happened.',
+  },
+  {
+    name: 'Expedited',
+    body: 'A compressed timetable fixed at the outset, with page limits and a short hearing. Chosen when the commercial cost of the dispute staying open exceeds the value of a full process.',
+  },
+  {
+    name: 'Full hearing',
+    body: 'Pleadings, disclosure, witnesses, experts, oral argument. Where the facts are genuinely contested and someone has to hear them tested.',
+  },
+];
+---
+
+<BaseLayout
+  title="Commercial Arbitration · Pouya Lajevardi · Toronto"
+  description="Sole, party-appointed and co-arbitration appointments in commercial matters, Toronto. Tracks, rules and awards — and where the Q.Arb pathway stands."
+  jsonLd={graph}
+>
+  {/* ---- 1. Hero — and the paired disclosure starts here ---------------- */}
+  <section class="section hero">
+    <div class="wrap">
+      <Eyebrow dot>Arbitration</Eyebrow>
+      <h1 class="display hero-h">Available now, and open about the stage.</h1>
+      <p class="hero-lede">
+        I accept sole, party-appointed and co-arbitration appointments in
+        commercial matters. The Q.Arb pathway commenced in August 2026;
+        C.Med-Arb is the endpoint. Both are true at once, and they belong
+        together rather than one of them surfacing later.
+      </p>
+      <div class="hero-creds">
+        <CredentialRow slots={[...CREDENTIAL_ROW, CREDENTIAL_ROW_ARB]} />
+      </div>
+    </div>
+  </section>
+
+  {/* ---- 2. The credentialing stage, on its own ------------------------- */}
+  {
+    /* §4 Offerings: "neither half may be dropped". The offering is in the hero;
+      this is the other half, given its own section rather than a footnote,
+      because Pouya's instruction is that the openness IS the differentiator. */
+  }
+  <section class="section section-inverse reveal">
+    <div class="wrap">
+      <div class="section-head">
+        <SectionHeading eyebrow="Where I am in the arc" level={2}>
+          <span slot="heading"
+            >Stated plainly rather than discovered later.</span
+          >
+        </SectionHeading>
+      </div>
+      <ol class="grid-autofit stage" role="list">
+        <li>
+          <Pill>Held</Pill>
+          <h3 class="stage-name">Q.Med</h3>
+          <p class="stage-body">
+            Qualified Mediator, through the ADR Institute of Canada and the ADR
+            Institute of Ontario.
+          </p>
+        </li>
+        <li>
+          <Pill>Commenced August 2026</Pill>
+          <h3 class="stage-name">Q.Arb</h3>
+          <p class="stage-body">
+            Qualified Arbitrator. Newly commenced — not held, and not nearing
+            completion.
+          </p>
+        </li>
+        <li>
+          <Pill>The endpoint</Pill>
+          <h3 class="stage-name">C.Med-Arb</h3>
+          <p class="stage-body">
+            {CREDENTIALS.goalName}. The designation this practice is built
+            toward.
+          </p>
+        </li>
+      </ol>
+      <p class="stage-note">
+        If the stage of the arc bears on an appointment decision, it should bear
+        on it before the appointment and not after.
+      </p>
+    </div>
+  </section>
+
+  {/* ---- 3. Appointments ------------------------------------------------ */}
+  <section class="section reveal">
+    <div class="wrap">
+      <div class="section-head">
+        <SectionHeading
+          eyebrow="Appointments"
+          level={2}
+          lede="Commercial matters. I do not accept family arbitration."
+        >
+          <span slot="heading">Three forms.</span>
+        </SectionHeading>
+      </div>
+      <DefinitionGrid items={APPOINTMENTS} />
+    </div>
+  </section>
+
+  {/* ---- 4. Tracks ------------------------------------------------------ */}
+  <section class="section section-alt reveal">
+    <div class="wrap">
+      <div class="section-head">
+        <SectionHeading
+          eyebrow="Tracks"
+          level={2}
+          lede="Settled with the parties before the first procedural order, not defaulted to."
+        >
+          <span slot="heading">How much process the dispute needs.</span>
+        </SectionHeading>
+      </div>
+      <DefinitionGrid items={TRACKS} />
+    </div>
+  </section>
+
+  {/* ---- 5. Rules ------------------------------------------------------- */}
+  <section class="section reveal">
+    <div class="wrap">
+      <div class="section-head">
+        <SectionHeading eyebrow="Rules" level={2}>
+          <span slot="heading">Under whose rules.</span>
+        </SectionHeading>
+      </div>
+      <div class="prose">
+        {
+          /* Sourced: docs/reference/adric-rules.md. ADR CHAMBERS IS DELIBERATELY
+            NOT NAMED — docs/01 item 3 lists it, and nothing in this repository
+            sources what rules it publishes. docs/07 cites it for published FEE
+            ranges, which is a different claim. Name it here once it is sourced;
+            do not name it from recall (R14). */
+        }
+        <p>
+          <strong>The ADRIC Arbitration Rules.</strong> The ADR Institute of Canada
+          adopted a new edition effective 1 March 2025, alongside an arbitrator appointment
+          protocol and a set of forms — notice to arbitrate, request to administer,
+          request for the appointment of an arbitrator, urgent interim measures, challenge
+          to an arbitrator, notice of appeal.
+        </p>
+        <p>
+          <strong>Or ad hoc, or whatever the contract names.</strong> Where a contract
+          names a rule set, a seat and a language and leaves the rest to the tribunal,
+          that works. What does not is a process whose shape nobody agreed to in advance.
+        </p>
+      </div>
+    </div>
+  </section>
+
+  {/* ---- 6. Awards ------------------------------------------------------ */}
+  <section class="section section-alt reveal">
+    <div class="wrap">
+      <div class="section-head">
+        <SectionHeading eyebrow="Awards" level={2}>
+          <span slot="heading">In writing, with reasons.</span>
+        </SectionHeading>
+      </div>
+      <div class="prose">
+        <p>
+          An award should be in writing and give reasons — what was in dispute,
+          what the record showed, and why the conclusion follows. An award that
+          announces a result without the reasoning is not much use to the party
+          that lost, and it is no use at all to the relationship that has to
+          survive it.
+        </p>
+        {
+          /* THE DATE IS A COMMITMENT ABOUT PROCESS, NOT A PUBLISHED TURNAROUND.
+            §4 Forbidden bars a time-to-award statistic outright; no figure
+            appears here and none may be added. */
+        }
+        <p>
+          The date an award is due belongs in the first procedural order, along
+          with everything else, rather than being left open. No number is
+          published here: a turnaround time advertised in advance of a record is
+          a guess dressed as a commitment.
+        </p>
+      </div>
+    </div>
+  </section>
+
+  {/* ---- 7. Fees --------------------------------------------------------- */}
+  <section class="section reveal">
+    <div class="wrap">
+      <div class="section-head">
+        <SectionHeading eyebrow="Fees" level={2}>
+          <span slot="heading">Published in full.</span>
+        </SectionHeading>
+      </div>
+      <div class="prose">
+        <p>
+          Hourly, hearing day, and flat fees for documents-only and expedited
+          matters at two levels of complexity. The cancellation schedule is on
+          the same page.
+        </p>
+      </div>
+      <div class="cta">
+        <Button href="/fees/" variant="ghost">The rate card &rarr;</Button>
+      </div>
+    </div>
+  </section>
+
+  <ContactBand />
+</BaseLayout>
+
+<style>
+  .hero {
+    padding-block-start: var(--space-9);
+  }
+  .hero-h {
+    margin-block: var(--space-4) var(--space-5);
+    font-size: var(--text-5xl);
+    max-inline-size: 20ch;
+  }
+  .hero-lede {
+    max-inline-size: 58ch;
+    font-size: var(--text-lg);
+    line-height: var(--leading-body);
+    color: var(--text-secondary);
+  }
+  .hero-creds {
+    margin-block-start: var(--space-8);
+  }
+
+  /* `.stage`, NOT `.arc`. `/about/` has an `.arc` block of its own with the
+     same class names and different rendering — serif at --text-2xl there, mono
+     at --text-lg here — and Astro's scoping means neither reaches the other. Two
+     meanings for one name across two pages is a rename waiting to go wrong.
+     `role="list"` on the <ol> and no local `list-style: none`: global.css does
+     both for `ol[role='list']`, and WebKit drops list semantics when the marker
+     goes, which is why the role is there. Columns come from `.grid-autofit`. */
+  .stage {
+    --grid-min: 17rem;
+    gap: var(--space-7);
+  }
+  .stage-name {
+    margin-block: var(--space-4) var(--space-2);
+    font-family: var(--font-mono);
+    font-size: var(--text-lg);
+    font-weight: var(--weight-medium);
+  }
+  .stage-body {
+    max-inline-size: 40ch;
+    line-height: var(--leading-body);
+  }
+  .stage-note {
+    margin-block-start: var(--space-8);
+    max-inline-size: 54ch;
+    line-height: var(--leading-body);
+    color: var(--text-inverse-2);
+  }
+  .cta {
+    margin-block-start: var(--space-6);
+  }
+</style>

src/pages/med-arb.astro

@@ -0,0 +1,274 @@
+---
+/**
+ * `/med-arb/` — build step 4. Spec: docs/01 §`/med-arb/`, docs/03 §Mediation /
+ * Arbitration / Med-Arb, docs/04 (FAQPage is specified for this page).
+ *
+ * docs/01 rates this the strongest candidate for the best-performing page on
+ * the site: real search demand, thin competition, and it maps to the practice's
+ * long-term arc. So it is written to be the page that actually explains the
+ * process, not a service blurb.
+ *
+ * THE FAIRNESS SECTION IS WRITTEN ABOUT THE PROCESS, NOT ABOUT POUYA, AND THAT
+ * IS DELIBERATE. docs/03 requires the procedural-fairness objection met head-on
+ * and the consent mechanics explained. What a med-arb AGREEMENT has to settle is
+ * process design and needs no §4 row. What POUYA will personally commit to —
+ * how he handles caucus information he cannot un-hear, and whether he would step
+ * out of the arbitral phase rather than decide on it — is a claim about his
+ * practice, has no §4 row, and is not invented here.
+ *   TODO(pouya): AGENTS.md Q54 — confirm the six conduct commitments drafted
+ *   there, or replace them. Three are this page's; two shipped here and on
+ *   `/mediation/` for one pass before `claims-auditor` removed them.
+ */
+import BaseLayout from '../layouts/BaseLayout.astro';
+import ContactBand from '../components/ContactBand.astro';
+import Eyebrow from '../components/Eyebrow.astro';
+import SectionHeading from '../components/SectionHeading.astro';
+import { getImage } from 'astro:assets';
+import ogDefault from '../assets/og-portrait.jpg';
+import { medArbGraph } from '../data/schema';
+import { CREDENTIALS } from '../data/site';
+
+const ldImage = await getImage({
+  src: ogDefault,
+  format: 'jpeg',
+  width: 1200,
+  height: 630,
+});
+
+/* ONE SOURCE FOR THE Q&A, rendered visibly AND fed to the FAQPage node. docs/04
+   allows FAQPage here "only where the visible page genuinely is Q&A" — so the
+   node is built from the same array the section renders, and a question cannot
+   be added to the structured data without appearing on the page. */
+const FAQ = [
+  {
+    q: 'The neutral who heard my client in caucus then decides the case. How is that fair?',
+    a: 'That is the objection, and it is the right one. A mediator learns things a decision-maker is not supposed to know — what a party would really take, what they are afraid of, what their own counsel thinks of their case. In med-arb the same person may go on to decide it. The answer is not that the concern is overstated. It is that med-arb is only defensible where the parties agree to it knowingly and in advance, in writing, with the switch and its consequences settled before the mediation phase begins — never improvised on the day because the room has run out of road.',
+  },
+  {
+    q: 'What happens to what we say in caucus?',
+    a: 'That is the question the agreement has to answer before anyone says anything. A med-arb agreement should settle three things in terms: what triggers the move from mediation to arbitration; what a party may say in caucus without it entering the arbitral record; and how a party flags, at the time, that something is being said for the mediation only. Vagueness on any of the three is what turns a procedural objection into a real one.',
+  },
+  {
+    q: 'When is med-arb the wrong process?',
+    a: 'When the parties are not equally informed or equally advised, because the caucus asymmetry compounds. When one side needs a finding on the record more than it needs a settlement. When the relationship is already so damaged that a joint session is unproductive and shuttle mediation would be the better tool on its own. And whenever a party agrees to it reluctantly: grudging consent at the outset is not the foundation a process like this one runs on.',
+  },
+];
+
+const graph = medArbGraph({
+  faq: FAQ,
+  imageUrl: new URL(ldImage.src, Astro.site).href,
+});
+---
+
+<BaseLayout
+  title="Med-Arb · Pouya Lajevardi · What It Is and When It Fits"
+  description="Med-arb is mediation that converts to arbitration if it does not resolve. What it is, how it differs from arb-med, the fairness objection, and when it fits."
+  jsonLd={graph}
+>
+  {/* ---- 1. Hero -------------------------------------------------------- */}
+  <section class="section hero">
+    <div class="wrap">
+      <Eyebrow dot>Med-Arb</Eyebrow>
+      <h1 class="display hero-h">
+        One neutral. Two processes. One agreement, written first.
+      </h1>
+      <p class="hero-lede">
+        Med-arb is mediation that converts to binding arbitration if the
+        mediation does not resolve the dispute. The parties get a real attempt
+        at settlement and a decision if one does not come, from a single
+        appointment — and they take on one real objection to get it. This page
+        is mostly about that objection.
+      </p>
+    </div>
+  </section>
+
+  {/* ---- 2. What it is, and how it is not arb-med ----------------------- */}
+  <section class="section section-alt reveal">
+    <div class="wrap">
+      <div class="section-head">
+        <SectionHeading eyebrow="The shape" level={2}>
+          <span slot="heading"
+            >Mediation first. Arbitration only if it is needed.</span
+          >
+        </SectionHeading>
+      </div>
+      <div class="prose">
+        <p>
+          The parties appoint one neutral for both phases. The matter is
+          mediated. Whatever settles, settles, and is recorded. Whatever does
+          not settle moves to arbitration in front of the same neutral, on the
+          terms the parties agreed before any of it began, and ends in an award.
+        </p>
+        <p>
+          The commercial case for it is time and cost. Without it, a mediation
+          that does not settle means starting again: a new neutral, a second
+          round of briefs, another set of dates months out. Med-arb removes that
+          gap, and whatever narrowing the mediation achieved carries into the
+          arbitration instead of being re-litigated.
+        </p>
+        {
+          /* ARB-MED IS NAMED AS A DIFFERENT PROCESS AND DELIBERATELY NOT DEFINED.
+            docs/01 asks for the contrast; nothing in `docs/reference/` sources a
+            definition of arb-med, and this repository does not publish a
+            definition of a third party's process from recall (R14). Flagging the
+            confusion is the useful half and it needs no source. Define it here
+            once a source is committed. */
+        }
+        <p>
+          <strong>It is not arb-med.</strong> The two names are one syllable apart
+          and the processes are not interchangeable. If a contract names one of them,
+          check which one before relying on it — against the rule set the contract
+          adopts, not against this page.
+        </p>
+      </div>
+    </div>
+  </section>
+
+  {/* ---- 3. The objection ----------------------------------------------- */}
+  <section class="section section-inverse reveal">
+    <div class="wrap">
+      <div class="section-head">
+        <SectionHeading
+          eyebrow="The objection"
+          level={2}
+          lede="Met here rather than further down, because it is the reason to read the page."
+        >
+          <span slot="heading">The fairness problem is real.</span>
+        </SectionHeading>
+      </div>
+      <div class="faq">
+        {
+          FAQ.map((item) => (
+            <div class="faq-item">
+              <h3 class="faq-q">{item.q}</h3>
+              <p class="faq-a">{item.a}</p>
+            </div>
+          ))
+        }
+      </div>
+    </div>
+  </section>
+
+  {/* ---- 4. Rules ------------------------------------------------------- */}
+  <section class="section reveal">
+    <div class="wrap">
+      <div class="section-head">
+        <SectionHeading eyebrow="Rules" level={2}>
+          <span slot="heading">There is a published rule set for this.</span>
+        </SectionHeading>
+      </div>
+      <div class="prose">
+        {
+          /* Sourced: docs/reference/adric-rules.md. Both quotations are ADRIC's
+            own words. "integrating seamlessly" stays inside the quotation marks —
+            §4 Forbidden bars superlatives in this site's own voice. */
+        }
+        <p>
+          The ADR Institute of Canada publishes <strong
+            >ADRIC Med-Arb Rules</strong
+          >, developed by a task force and put to the membership in 2019, and
+          designed, in ADRIC's words, to <q
+            >work in tandem with ADRIC's existing Mediation Rules and
+            Arbitration Rules</q
+          >. They were drafted for domestic commercial disputes, and ADRIC notes
+          that parties may also apply them to international or non-commercial
+          ones.
+        </p>
+        <p>
+          ADRIC's own framing is worth quoting: med-arb is
+          <q
+            >not merely the merging of separate mediation and arbitration
+            processes, but a unique process designed to meet the needs of
+            particular disputants</q
+          >, one that <q
+            >requires a high level of practitioner competence to do successfully</q
+          >.
+        </p>
+      </div>
+    </div>
+  </section>
+
+  {/* ---- 5. The endpoint ------------------------------------------------- */}
+  <section class="section section-alt reveal">
+    <div class="wrap">
+      <div class="section-head">
+        <SectionHeading eyebrow="Why this practice" level={2}>
+          <span slot="heading"
+            >C.Med-Arb is the designation this practice is built toward.</span
+          >
+        </SectionHeading>
+      </div>
+      <div class="prose">
+        <p>
+          {CREDENTIALS.goalName}, the ADRIC and ADRIO designation for exactly
+          this hybrid, is the endpoint of the arc. Q.Med is held. The Q.Arb
+          pathway commenced in August 2026. C.Med-Arb comes after both, and it
+          is the reason the two halves of this practice are being built together
+          rather than one after the other.
+        </p>
+        <p>
+          I accept med-arb appointments now, in commercial matters. The section
+          above is the part to read before proposing one: the agreement does the
+          work, and it does it before the mediation starts.
+        </p>
+        <p class="onward">
+          <a href="/mediation/">Mediation &rarr;</a>
+          <a href="/arbitration/">Arbitration &rarr;</a>
+          <a href="/fees/">Fees &rarr;</a>
+        </p>
+      </div>
+    </div>
+  </section>
+
+  <ContactBand />
+</BaseLayout>
+
+<style>
+  .hero {
+    padding-block-start: var(--space-9);
+  }
+  .hero-h {
+    margin-block: var(--space-4) var(--space-5);
+    font-size: var(--text-5xl);
+    max-inline-size: 18ch;
+  }
+  .hero-lede {
+    max-inline-size: 58ch;
+    font-size: var(--text-lg);
+    line-height: var(--leading-body);
+    color: var(--text-secondary);
+  }
+
+  .faq {
+    display: grid;
+    gap: var(--space-8);
+  }
+  .faq-q {
+    max-inline-size: 46ch;
+    font-size: var(--text-2xl);
+    line-height: var(--leading-tight);
+  }
+  .faq-a {
+    margin-block-start: var(--space-4);
+    max-inline-size: var(--width-prose);
+    line-height: var(--leading-body);
+    color: var(--text-inverse-2);
+  }
+
+  /* A ROW OF STANDALONE CTAs, NOT PROSE, so WCAG 2.5.8's inline-link exception
+     does not cover it. Measured at 390px before the fix: three targets at
+     21px tall, against the 44px floor docs/02 sets and the 24px minimum 2.5.8
+     sets. `inline-flex` + `min-block-size` is what `.btn`, the nav and the
+     footer all use. The middot separators went with the fix — spacing does the
+     same job without an aria-hidden glyph between two links. */
+  .onward {
+    display: flex;
+    flex-wrap: wrap;
+    gap: var(--space-2) var(--space-6);
+  }
+  .onward a {
+    display: inline-flex;
+    align-items: center;
+    min-block-size: 44px;
+  }
+</style>

src/pages/mediation.astro

@@ -0,0 +1,311 @@
+---
+/**
+ * `/mediation/` — build step 4. Spec: docs/01 §`/mediation/`, docs/03 §Mediation.
+ *
+ * Job (docs/01): convert counsel who have already decided on mediation and are
+ * choosing a neutral. So: procedural, specific, unembellished. No persuasion
+ * that the reader has already done for themselves.
+ */
+import BaseLayout from '../layouts/BaseLayout.astro';
+import Button from '../components/Button.astro';
+import ContactBand from '../components/ContactBand.astro';
+import DefinitionGrid from '../components/DefinitionGrid.astro';
+import CredentialRow from '../components/CredentialRow.astro';
+import Eyebrow from '../components/Eyebrow.astro';
+import PracticeCard from '../components/PracticeCard.astro';
+import SectionHeading from '../components/SectionHeading.astro';
+import { getImage } from 'astro:assets';
+import ogDefault from '../assets/og-portrait.jpg';
+import { serviceGraph } from '../data/schema';
+import { CREDENTIAL_ROW, PRACTICE_AREAS } from '../data/site';
+
+const ldImage = await getImage({
+  src: ogDefault,
+  format: 'jpeg',
+  width: 1200,
+  height: 630,
+});
+const graph = serviceGraph({
+  path: '/mediation/',
+  name: 'Commercial mediation',
+  serviceType: 'Mediation',
+  description:
+    'Sole-mediator appointments in commercial disputes. Half day or full ' +
+    'day, in person, by video, or shuttle.',
+  imageUrl: new URL(ldImage.src, Astro.site).href,
+});
+
+/* Formats are SERVICE PARAMETERS, not offerings — AGENTS.md Q43 put process
+   commitments of this class ("not facts about Pouya") on the publishable side.
+   The half day / full day / video / in-person set is D14's card; shuttle and
+   hybrid are docs/01 §`/mediation/` item 2. */
+const FORMATS = [
+  {
+    name: 'Full day',
+    body: 'The default for a matter with more than two parties, or where the documents need working through in the room.',
+  },
+  {
+    name: 'Half day',
+    body: 'Enough for a two-party matter where the issues are narrow and both sides arrive ready.',
+  },
+  {
+    name: 'Shuttle',
+    body: 'The parties never sit in the same room. Useful where the relationship has broken down far enough that a joint session would cost more than it returns.',
+  },
+  {
+    name: 'By video',
+    body: 'Same preparation, same length, same rate. Remote is not a discount format.',
+  },
+  {
+    name: 'Hybrid',
+    body: 'Principals in the room, an expert or an insurer joining remotely for the part that concerns them.',
+  },
+];
+---
+
+<BaseLayout
+  title="Commercial Mediation · Pouya Lajevardi · Q.Med · Toronto"
+  description="Sole mediator for commercial disputes in Toronto. Q.Med through ADRIC and ADRIO. Session formats, the ADRIC National Mediation Rules, and what to bring."
+  jsonLd={graph}
+>
+  {/* ---- 1. Hero -------------------------------------------------------- */}
+  <section class="section hero">
+    <div class="wrap">
+      <Eyebrow dot>Mediation</Eyebrow>
+      <h1 class="display hero-h">A mediator decides nothing.</h1>
+      <p class="hero-lede">
+        That is the point of the role, and it is what makes the day work. I run
+        the process, test each side's case against the documents, and keep it
+        moving until there is an agreement or there plainly will not be one.
+      </p>
+      <div class="hero-creds">
+        <CredentialRow slots={CREDENTIAL_ROW} />
+      </div>
+    </div>
+  </section>
+
+  {/* ---- 2. The role ---------------------------------------------------- */}
+  <section class="section section-alt reveal">
+    <div class="wrap">
+      <div class="section-head">
+        <SectionHeading eyebrow="The role" level={2}>
+          <span slot="heading">What I am, and what I am not.</span>
+        </SectionHeading>
+      </div>
+      <div class="prose">
+        <p>
+          I act as a neutral. I do not act for a party in a matter I take, and
+          each party should have their own legal advice.
+        </p>
+        <p>
+          What I bring to a commercial file is that I read the contract and the
+          technical record underneath it — the change orders, the schedule, the
+          model card, the interconnection study — rather than either side's
+          summary of them.
+        </p>
+      </div>
+    </div>
+  </section>
+
+  {/* ---- 3. Formats ----------------------------------------------------- */}
+  <section class="section reveal">
+    <div class="wrap">
+      <div class="section-head">
+        <SectionHeading
+          eyebrow="Formats"
+          level={2}
+          lede="Chosen with the parties at engagement, not assumed."
+        >
+          <span slot="heading">How the session runs.</span>
+        </SectionHeading>
+      </div>
+      <DefinitionGrid items={FORMATS} minColumn="22rem" />
+    </div>
+  </section>
+
+  {/* ---- 4. Rules ------------------------------------------------------- */}
+  <section class="section section-inverse reveal">
+    <div class="wrap">
+      <div class="section-head">
+        <SectionHeading eyebrow="Rules" level={2}>
+          <span slot="heading">Under whose rules.</span>
+        </SectionHeading>
+      </div>
+      <div class="prose">
+        {
+          /* "ADRIC National Mediation Rules" — the document's own title, from
+            docs/reference/adric-rules.md. docs/01 said "Model Mediation Rules",
+            which is not the name of anything ADRIC publishes; "Model" belongs to
+            the Model Dispute Resolution Clause inside the rules. Finding 1. */
+        }
+        <p>
+          <strong>The ADRIC National Mediation Rules.</strong> The ADR Institute of
+          Canada publishes them. They cover initiating a mediation and appointing
+          a mediator where the parties cannot agree on one, and the document carries
+          a code of conduct, a standard form agreement to mediate, and a model dispute
+          resolution clause for contracts. If the agreement already names them, the
+          process is defined before anyone calls me.
+        </p>
+        <p>
+          <strong>Or a protocol the parties write.</strong> Where the contract is
+          silent, or where the matter needs something a standard rule set does not
+          contemplate, the protocol is settled in writing before the session.
+        </p>
+      </div>
+    </div>
+  </section>
+
+  {/* ---- 5. What to bring ----------------------------------------------- */}
+  <section class="section reveal">
+    <div class="wrap">
+      <div class="section-head">
+        <SectionHeading eyebrow="Preparation" level={2}>
+          <span slot="heading">What to bring.</span>
+        </SectionHeading>
+      </div>
+      <ul class="bring" role="list">
+        <li>
+          <strong>A brief.</strong> The issues, the position taken on each, and what
+          has already been offered and refused. Exchanged in advance, so the session
+          starts informed rather than spending the morning getting there.
+        </li>
+        <li>
+          <strong>The documents the case turns on</strong> — not the production. If
+          a clause, a change order, or a test result decides an issue, put it in front
+          of me before the day.
+        </li>
+        <li>
+          <strong>Authority to settle.</strong> In the room, or reachable in real
+          time and expecting the call. A mediation that reaches terms and then adjourns
+          for instructions is a mediation at risk.
+        </li>
+      </ul>
+    </div>
+  </section>
+
+  {/* ---- 6. Confidentiality --------------------------------------------- */}
+  <section class="section section-alt reveal">
+    <div class="wrap">
+      <div class="section-head">
+        <SectionHeading eyebrow="Confidentiality" level={2}>
+          <span slot="heading">What stays in the room.</span>
+        </SectionHeading>
+      </div>
+      <div class="prose">
+        <p>
+          Confidentiality is set by the agreement to mediate, and that is signed
+          before the session rather than described on a website. The part worth
+          settling in it explicitly is the caucus: what a mediator may carry
+          from one room to the other, and what a party has to say to hold
+          something back. An agreement that leaves that implicit is the one that
+          produces an argument on the day.
+        </p>
+        {
+          /* The without-prejudice question is answered by pointing, not by
+            characterising legal effect. AGENTS.md §4 bars this repository from
+            concluding a proposition of law, and docs/03's `[unestablished]`
+            pattern says to write around the capacity question. */
+        }
+        <p>
+          Mediation is conducted on a without-prejudice basis. What that means
+          for a particular file, and what survives it, is a question for each
+          party's own counsel rather than for the neutral.
+        </p>
+      </div>
+    </div>
+  </section>
+
+  {/* ---- 7. Practice areas ---------------------------------------------- */}
+  <section class="section reveal">
+    <div class="wrap">
+      <div class="section-head">
+        <SectionHeading
+          eyebrow="Subject matter"
+          level={2}
+          lede="Where a commercial mediation usually turns on something technical."
+        >
+          <span slot="heading">Six areas.</span>
+        </SectionHeading>
+      </div>
+      <div class="grid-autofit areas-grid">
+        {
+          PRACTICE_AREAS.map((area) => (
+            <PracticeCard
+              href={`/practice/${area.slug}/`}
+              chip={area.chip}
+              title={area.name}
+              level={3}
+            >
+              {area.blurb}
+            </PracticeCard>
+          ))
+        }
+      </div>
+    </div>
+  </section>
+
+  {/* ---- 8. Fees --------------------------------------------------------- */}
+  <section class="section section-alt reveal">
+    <div class="wrap">
+      <div class="section-head">
+        <SectionHeading eyebrow="Fees" level={2}>
+          <span slot="heading">Published in full.</span>
+        </SectionHeading>
+      </div>
+      <div class="prose">
+        <p>
+          Half day and full day, with the preparation time bundled into each one
+          stated rather than folded into the hours. Additional parties, overtime
+          and the cancellation schedule are on the same page.
+        </p>
+      </div>
+      <div class="cta">
+        <Button href="/fees/" variant="ghost">The rate card &rarr;</Button>
+      </div>
+    </div>
+  </section>
+
+  <ContactBand />
+</BaseLayout>
+
+<style>
+  .hero {
+    padding-block-start: var(--space-9);
+  }
+  .hero-h {
+    margin-block: var(--space-4) var(--space-5);
+    font-size: var(--text-6xl);
+  }
+  .hero-lede {
+    max-inline-size: 58ch;
+    font-size: var(--text-lg);
+    line-height: var(--leading-body);
+    color: var(--text-secondary);
+  }
+  .hero-creds {
+    margin-block-start: var(--space-8);
+  }
+
+  .cta {
+    margin-block-start: var(--space-6);
+  }
+
+  .bring {
+    /* No `list-style: none` or `padding: 0` here — `global.css` applies both to
+       `ul[role='list']`, and a second copy is a second thing to keep true. */
+    display: grid;
+    gap: var(--space-5);
+    max-inline-size: var(--width-prose);
+  }
+  .bring li {
+    padding-inline-start: var(--space-5);
+    border-inline-start: 1px solid var(--rule);
+    line-height: var(--leading-body);
+  }
+
+  /* `.grid-autofit` (global.css) carries the columns and the `min()` guard;
+     this only sets the gap. One home for that reasoning, not five. */
+  .areas-grid {
+    gap: var(--space-5);
+  }
+</style>

src/styles/global.css

@@ -321,6 +321,42 @@ a:hover {
 .section-head {
   margin-block-end: var(--space-7);
 }
+
+/* THE AUTO-FIT GRID GUARD, IN ONE PLACE. `adversarial-reviewer` counted five
+   copies of the same five-line comment across three pages at step 4 — 25 lines
+   explaining 5 identical declarations — plus a sixth, differently worded copy on
+   `/about/`. That is this repo's own "a second copy is a second thing to keep
+   true", and D19's comment rule, both breached by a paragraph about a guard.
+
+   THE GUARD: `min(Nrem, 100%)`, never a bare rem. A bare rem floor is a HARD
+   minimum, so at a 200% default font size (root 32px — a real accessibility
+   setting, not page zoom) an 18rem floor becomes 576px and the track refuses to
+   shrink. Measured on `/`: 234px of document overflow at a 390px viewport, down
+   to 3px once the three grids took `min()`. docs/02 §Accessibility floor carries
+   the full table.
+
+   Set `--grid-min` on the element; default 18rem.
+
+   ⚠️ NOT YET THE ONLY HOME, and the remaining copies are listed rather than
+   claimed swept. `git grep -n 'auto-fit' -- src`, 2026-08-28:
+
+     src/pages/index.astro:875   minmax(min(18rem, 100%), 1fr)
+     src/pages/index.astro:916   minmax(min(13rem, 100%), 1fr)
+     src/pages/about.astro:965   minmax(min(18rem, 100%), 1fr)
+     src/pages/about.astro:1007  minmax(min(16rem, 100%), 1fr)
+
+   Those four are on pages that already shipped; converting them after the
+   review cap would be an unreviewed change to live CSS. They move onto this
+   class at step 5, when `/practice/*` is in the same files. Until then a
+   correction here reaches three call sites, not seven — say that rather than
+   letting a reader believe the guard has one home. */
+.grid-autofit {
+  display: grid;
+  grid-template-columns: repeat(
+    auto-fit,
+    minmax(min(var(--grid-min, 18rem), 100%), 1fr)
+  );
+}
 .section {
   padding-block: var(--section-y);
 }