diff --git a/docs/infra.md b/docs/infra.md
new file mode 100644
index 0000000..97be9a7
--- /dev/null
+++ b/docs/infra.md
@@ -0,0 +1,19 @@
+# Infrastructure notes (state that lives outside this repo)
+
+## Domains (Vercel dashboard → Settings → Domains)
+
+- Production domains: `hackimi.dev` (apex) and `www.hackimi.dev`.
+- **Canonical host: apex `https://hackimi.dev`** — every canonical tag, sitemap URL, robots Sitemap line, and JSON-LD url in this repo uses apex.
+- Redirect direction: `www` → 308 → apex. (Flipped 2026-08-\_\_ by Nima; before that the dashboard had apex → www, contradicting the code's canonicals.)
+- `hackimi.vercel.app` additionally 308s to apex via `next.config.js` (exact-host rule — preview URLs on \*.vercel.app are deliberately not matched).
+
+## Deployments
+
+- Production deploys are automatic but CI-gated (commit `1c92564`, 2026-06-02): GitHub Actions runs lint/types/Vitest/Playwright on every push, and a deploy job (`vercel build` + `vercel deploy --prebuilt --prod`) runs only on `main` after the tests pass.
+- `vercel.json` sets `git.deploymentEnabled.main: false` to disable Vercel's OWN git auto-deploy for `main` only — so production flows exclusively through the tested CI gate while PR/branch previews keep deploying via Vercel as normal. Merging to `main` therefore DOES deploy, via CI.
+
+## Search Console
+
+- Property: `https://hackimi.dev` (URL-prefix), meta-tag verification.
+- Token lives in `SITE_GOOGLE_SITE_VERIFICATION` (`src/lib/site.ts`); empty string disables the tag entirely.
+- After each SEO-relevant deploy: submit `sitemap.xml`, request indexing of changed URLs.
diff --git a/docs/superpowers/plans/2026-08-09-seo-name-query.md b/docs/superpowers/plans/2026-08-09-seo-name-query.md
new file mode 100644
index 0000000..6c61371
--- /dev/null
+++ b/docs/superpowers/plans/2026-08-09-seo-name-query.md
@@ -0,0 +1,1005 @@
+# SEO Name-Query Implementation Plan
+
+> **For agentic workers:** REQUIRED SUB-SKILL: Use superpowers:subagent-driven-development (recommended) or superpowers:executing-plans to implement this plan task-by-task. Steps use checkbox (`- [ ]`) syntax for tracking.
+
+**Goal:** Make hackimi.dev win the "Nima Hakimi" name query: recover the dead `/projects` ranking URL, name-led titles, entity-graph JSON-LD, favicon + branded OG card, honest sitemap, Search Console wiring.
+
+**Architecture:** All changes ride the existing metadata pipeline (`src/lib/site.ts` constants → `src/lib/metadata.ts` builders → per-page `metadata` exports). New surfaces: `next.config.js` redirects, a `JsonLd` component for structured-data blocks, a site footer, Next file-convention icons, and a committed asset-generator script.
+
+**Tech Stack:** Next.js 16 App Router (Turbopack), React 19, Vitest, Playwright, Prettier 3.8.3.
+
+**Spec:** `docs/superpowers/specs/2026-08-09-seo-name-query-design.md`
+
+## Global Constraints
+
+- Positioning wording in SEO surfaces: exactly `Developer & AI Engineer` (prose form: `developer and AI engineer`). **Visible site copy keeps "frontend"** — do NOT touch hero/about/contact/case-study visible text (`src/app/page.tsx` JSX, `about` prose, `ContactClient` blurb, `src/data/*` narratives).
+- Canonical host is apex `https://hackimi.dev` everywhere in code. Never introduce `www.`
+- TDD: every task starts with a failing test. Run tests with `npm run test` (Vitest) / `npm run test:e2e` (Playwright).
+- Playwright locally requires **unsandboxed Bash** and `ulimit -n 10240` (Chromium fails to launch sandboxed: "Permission denied (1100)"). Same for the asset-generator script (it launches Chromium).
+- Turbopack dev server's first cold run may fail 1 random e2e while compiling on demand — re-run before blaming a change; production build is stable.
+- Commits: conventional messages, no AI attribution, no session links, no Claude-Session trailer.
+- `npm run lint` must stay 0 errors / 0 warnings; `npm run check` (Prettier) must pass before each commit.
+
+---
+
+### Task 1: Permanent redirects — recover the dead /projects URL
+
+**Files:**
+
+- Modify: `next.config.js` (currently an empty `nextConfig = {}`)
+- Create: `e2e/seo.spec.ts`
+
+**Interfaces:**
+
+- Produces: `e2e/seo.spec.ts` — later tasks append describe-blocks to this file.
+
+- [ ] **Step 1: Write the failing e2e test**
+
+Create `e2e/seo.spec.ts`:
+
+```ts
+import { test, expect } from '@playwright/test';
+
+test.describe('redirects', () => {
+  test('/projects permanently redirects to /work', async ({ request }) => {
+    const res = await request.get('/projects', { maxRedirects: 0 });
+    expect(res.status()).toBe(308);
+    expect(res.headers()['location']).toBe('/work');
+  });
+
+  test('/projects/anything redirects to /work', async ({ request }) => {
+    const res = await request.get('/projects/old-page', { maxRedirects: 0 });
+    expect(res.status()).toBe(308);
+    expect(res.headers()['location']).toBe('/work');
+  });
+});
+```
+
+- [ ] **Step 2: Run to verify it fails**
+
+Run (unsandboxed, `ulimit -n 10240` first): `npx playwright test e2e/seo.spec.ts` Expected: FAIL — `/projects` returns 404, not 308.
+
+- [ ] **Step 3: Implement redirects**
+
+Replace `next.config.js` content:
+
+```js
+/** @type {import('next').NextConfig} */
+const nextConfig = {
+  async redirects() {
+    return [
+      // The pre-2026-06-02 site served /projects (Google's current #1 result
+      // for "Nima Hakimi developer"); the redesign renamed it to /work.
+      { source: '/projects', destination: '/work', permanent: true },
+      { source: '/projects/:path*', destination: '/work', permanent: true },
+      // Exact-host only: preview deployments on *.vercel.app must stay untouched.
+      {
+        source: '/:path*',
+        has: [{ type: 'host', value: 'hackimi.vercel.app' }],
+        destination: 'https://hackimi.dev/:path*',
+        permanent: true,
+      },
+    ];
+  },
+};
+
+module.exports = nextConfig;
+```
+
+- [ ] **Step 4: Run to verify it passes**
+
+Run: `npx playwright test e2e/seo.spec.ts` — Expected: PASS. If it fails unexpectedly after the config edit: Turbopack's persistent FS cache is known to go stale across `next.config.js` changes in this repo — `rm -rf .next`, restart the server, re-run BEFORE debugging the redirect rule itself. Also verify the host rule manually (dev server on :3000): `curl -s -o /dev/null -w '%{http_code} %{redirect_url}' -H 'Host: hackimi.vercel.app' http://localhost:3000/about` Expected: `308 https://hackimi.dev/about`. If the dev server doesn't honor the Host match, verify against `npm run build && npm run start` instead; if it only works in prod mode, note that in the task report — the rule itself is correct.
+
+- [ ] **Step 5: Commit**
+
+```bash
+git add next.config.js e2e/seo.spec.ts
+git commit -m "feat: permanent redirects for legacy /projects and vercel.app host"
+```
+
+---
+
+### Task 2: Positioning constants — Developer & AI Engineer
+
+**Files:**
+
+- Modify: `src/lib/site.ts`
+- Modify: `tests/unit/site.test.ts`
+
+**Interfaces:**
+
+- Produces: `SITE_ROLE = 'Developer & AI Engineer'`, `SITE_TITLE = 'Nima Hakimi — Developer & AI Engineer'` (em-dash), updated `SITE_DESCRIPTION`/`SITE_KEYWORDS`, and `SITE_SOCIAL_LINKS.orcid`. Tasks 3–7 consume these.
+
+- [ ] **Step 1: Update the pinned test to the new contract**
+
+In `tests/unit/site.test.ts` replace the SITE_TITLE test:
+
+```ts
+it('SITE_TITLE is `${SITE_AUTHOR} — ${SITE_ROLE}` (em-dash form)', () => {
+  expect(SITE_TITLE).toBe(`${SITE_AUTHOR} — ${SITE_ROLE}`);
+  expect(SITE_ROLE).toBe('Developer & AI Engineer');
+});
+```
+
+And add to the same describe:
+
+```ts
+it('social links include GitHub, LinkedIn and ORCID', () => {
+  expect(Object.keys(SITE_SOCIAL_LINKS).sort()).toEqual(['github', 'linkedin', 'orcid']);
+});
+
+it('keywords carry the new positioning, not "Frontend"', () => {
+  expect(SITE_KEYWORDS).toContain('AI Engineer');
+  expect(SITE_KEYWORDS.join(' ')).not.toMatch(/frontend/i);
+});
+```
+
+- [ ] **Step 2: Run to verify failure**
+
+Run: `npx vitest run tests/unit/site.test.ts` Expected: FAIL (pipe title, no orcid key, 'Frontend Developer' keyword).
+
+- [ ] **Step 3: Implement in `src/lib/site.ts`**
+
+```ts
+export const SITE_ROLE = 'Developer & AI Engineer';
+export const SITE_TITLE = `${SITE_AUTHOR} — ${SITE_ROLE}`;
+export const SITE_DESCRIPTION =
+  'Portfolio of Nima Hakimi, a developer and AI engineer building accessible, high-performance web applications.';
+export const SITE_KEYWORDS = ['Nima Hakimi', 'Developer', 'AI Engineer', 'Next.js', 'React', 'Oslo'];
+```
+
+And extend the social links object:
+
+```ts
+export const SITE_SOCIAL_LINKS = {
+  github: 'https://github.com/nimkimi',
+  linkedin: 'https://linkedin.com/in/nima-hakimi-387716175',
+  orcid: 'https://orcid.org/0009-0002-6656-2498',
+};
+```
+
+Leave `SITE_OG_IMAGE` untouched in this task (Task 7 changes it).
+
+- [ ] **Step 4: Run the full unit/server suite**
+
+Run: `npm run test` Expected: PASS — `tests/unit/metadata.test.ts` compares against the imported constants, so it rides through. If `buildPersonJsonLd`'s sameAs test fails (it pins `[github, linkedin]`), do NOT fix it here — that contract changes in Task 6; temporarily it still passes because the builder maps the two keys explicitly.
+
+- [ ] **Step 5: Commit**
+
+```bash
+git add src/lib/site.ts tests/unit/site.test.ts
+git commit -m "feat: reposition site constants to Developer & AI Engineer"
+```
+
+---
+
+### Task 3: Name-led homepage title
+
+**Files:**
+
+- Modify: `src/lib/metadata.ts` (add `buildHomeMetadata`)
+- Modify: `src/app/page.tsx:10-14` (metadata export only — no JSX changes)
+- Modify: `tests/unit/metadata.test.ts`, `tests/integration/pages.test.tsx:58-62`
+- Modify: `e2e/seo.spec.ts` (append)
+
+**Interfaces:**
+
+- Consumes: `SITE_TITLE`, `SITE_DESCRIPTION` from Task 2.
+- Produces: `buildHomeMetadata(): Metadata` in `src/lib/metadata.ts`.
+
+Background for the implementer: Next's `title.template` applies only to **child** segments — the root `page.tsx` never receives it, so its current `title: 'Portfolio'` renders as literally `<title>Portfolio</title>` in production. That is why the fix must set an absolute full-string title and why the e2e assertion (rendered layer) is the one that counts: the unit-level metadata object looked correct the whole time this bug shipped.
+
+- [ ] **Step 1: Write the failing tests**
+
+Append to `tests/unit/metadata.test.ts`:
+
+```ts
+describe('buildHomeMetadata', () => {
+  it('title is the absolute SITE_TITLE (root segment gets no template)', () => {
+    const meta = buildHomeMetadata();
+    expect(meta.title).toBe(SITE_TITLE);
+    expect(meta.openGraph?.title).toBe(SITE_TITLE);
+    expect(meta.twitter?.title).toBe(SITE_TITLE);
+  });
+
+  it('canonical and og.url are the bare SITE_URL', () => {
+    const meta = buildHomeMetadata();
+    expect(meta.alternates?.canonical).toBe(SITE_URL);
+    expect(meta.openGraph?.url).toBe(SITE_URL);
+  });
+});
+```
+
+(Import `buildHomeMetadata` in the existing import list.)
+
+In `tests/integration/pages.test.tsx`, replace the home metadata test (lines 58–62):
+
+```ts
+it('exports absolute name-led home metadata', async () => {
+  const { metadata } = await import('@/app/page');
+  const { SITE_TITLE } = await import('@/lib/site');
+  expect(metadata.title).toBe(SITE_TITLE);
+  expect(metadata.description).toContain('Nima Hakimi');
+});
+```
+
+Append to `e2e/seo.spec.ts`:
+
+```ts
+test.describe('rendered metadata', () => {
+  test('homepage title is the full name-led title', async ({ page }) => {
+    await page.goto('/');
+    await expect(page).toHaveTitle('Nima Hakimi — Developer & AI Engineer');
+  });
+
+  test('homepage canonical points at apex', async ({ page }) => {
+    await page.goto('/');
+    await expect(page.locator('link[rel="canonical"]')).toHaveAttribute('href', 'https://hackimi.dev');
+  });
+});
+```
+
+- [ ] **Step 2: Run to verify failures**
+
+Run: `npx vitest run tests/unit/metadata.test.ts tests/integration/pages.test.tsx` Expected: FAIL — `buildHomeMetadata` doesn't exist; home title is 'Portfolio'.
+
+- [ ] **Step 3: Implement**
+
+In `src/lib/metadata.ts`, after `buildPageMetadata`, add (reusing the module's existing `defaultImage`, `defaultTwitterImage`, `defaultKeywords`):
+
+```ts
+/**
+ * Home is the ONE page whose title must be absolute: Next's title.template
+ * only applies to child segments, so the root page.tsx never receives the
+ * "%s | Nima Hakimi" suffix — it must carry the full name-led string itself.
+ */
+export function buildHomeMetadata(): Metadata {
+  return {
+    title: SITE_TITLE,
+    description: SITE_DESCRIPTION,
+    keywords: defaultKeywords,
+    alternates: { canonical: SITE_URL },
+    openGraph: {
+      title: SITE_TITLE,
+      description: SITE_DESCRIPTION,
+      url: SITE_URL,
+      siteName: SITE_AUTHOR,
+      type: 'website',
+      images: [defaultImage],
+    },
+    twitter: {
+      card: 'summary_large_image',
+      title: SITE_TITLE,
+      description: SITE_DESCRIPTION,
+      images: [defaultTwitterImage],
+    },
+  };
+}
+```
+
+(`SITE_TITLE` and `SITE_DESCRIPTION` are already imported at the top of the module.)
+
+In `src/app/page.tsx`, replace the metadata export (lines 10–14) with:
+
+```ts
+export const metadata = buildHomeMetadata();
+```
+
+and change the import from `buildPageMetadata` to `buildHomeMetadata`. Touch nothing else in the file.
+
+- [ ] **Step 4: Run to verify passes**
+
+Run: `npm run test` then `npx playwright test e2e/seo.spec.ts` (unsandboxed, ulimit set). Expected: all PASS.
+
+Contingency: if the e2e title assertion fails with a DOUBLED suffix (`Nima Hakimi — Developer & AI Engineer | Nima Hakimi`), the root template did apply after all — switch `buildHomeMetadata`'s title to `title: { absolute: SITE_TITLE }` and update the two object-level title assertions to match. Live-production evidence says this won't happen; the e2e exists to catch exactly this case.
+
+- [ ] **Step 5: Commit**
+
+```bash
+git add src/lib/metadata.ts src/app/page.tsx tests/unit/metadata.test.ts tests/integration/pages.test.tsx e2e/seo.spec.ts
+git commit -m "feat: absolute name-led homepage title via buildHomeMetadata"
+```
+
+---
+
+### Task 4: Site footer with visible name line
+
+**Files:**
+
+- Create: `src/components/layout/SiteFooter.tsx`
+- Modify: `src/app/layout.tsx:30-33` (add footer inside `SmoothScroll`, after `main`)
+- Create: `tests/components/site-footer.test.tsx`
+- Modify: `e2e/seo.spec.ts` (append)
+
+**Interfaces:**
+
+- Consumes: `SITE_AUTHOR`, `SITE_ROLE`, `SITE_LOCATION` from `src/lib/site.ts`.
+- Produces: `SiteFooter` default-export server component, rendered on every page.
+
+- [ ] **Step 0: Pre-flight — check for existing `<footer>` elements**
+
+Run: `grep -rn '<footer' src/` Expected: zero hits (then `page.locator('footer')` below is unambiguous). If there ARE hits, give the new footer `data-site-footer` and use `page.locator('[data-site-footer]')` in the e2e instead — `footer`/`contentinfo` locators would throw strict-mode violations.
+
+- [ ] **Step 1: Write the failing tests**
+
+Create `tests/components/site-footer.test.tsx`:
+
+```tsx
+import { render, screen } from '@testing-library/react';
+import { describe, expect, it } from 'vitest';
+
+import SiteFooter from '@/components/layout/SiteFooter';
+import { SITE_AUTHOR, SITE_ROLE, SITE_LOCATION } from '@/lib/site';
+
+describe('SiteFooter', () => {
+  it('renders the © name-role-city line as visible text', () => {
+    render(<SiteFooter />);
+    const footer = screen.getByRole('contentinfo');
+    const year = new Date().getFullYear();
+    expect(footer).toHaveTextContent(`© ${year} ${SITE_AUTHOR} — ${SITE_ROLE}, ${SITE_LOCATION.city}`);
+  });
+});
+```
+
+Append to `e2e/seo.spec.ts` inside the `rendered metadata` describe:
+
+```ts
+test('footer carries the visible name line on every page', async ({ page }) => {
+  for (const path of ['/', '/about', '/work']) {
+    await page.goto(path);
+    await expect(page.locator('footer')).toContainText('Nima Hakimi — Developer & AI Engineer');
+  }
+});
+```
+
+- [ ] **Step 2: Run to verify failure**
+
+Run: `npx vitest run tests/components/site-footer.test.tsx` Expected: FAIL — module not found.
+
+- [ ] **Step 3: Implement**
+
+Create `src/components/layout/SiteFooter.tsx`:
+
+```tsx
+import { SITE_AUTHOR, SITE_ROLE, SITE_LOCATION } from '@/lib/site';
+
+/**
+ * Site-wide footer. Besides closing the page visually, this is the one
+ * crawlable spot outside the hero h1 where the full name appears as text
+ * on every route (the nav logo is an SVG with only an aria-label).
+ */
+export default function SiteFooter() {
+  return (
+    <footer className="mx-auto max-w-6xl px-4 sm:px-6">
+      <div className="mt-[clamp(3rem,8vh,5rem)] border-t border-white/10 py-8">
+        <p className="font-mono text-xs uppercase tracking-[0.12em] text-muted">
+          © {new Date().getFullYear()} {SITE_AUTHOR} — {SITE_ROLE}, {SITE_LOCATION.city}
+        </p>
+      </div>
+    </footer>
+  );
+}
+```
+
+In `src/app/layout.tsx`, import it (`import SiteFooter from '@/components/layout/SiteFooter';`) and render it after `<main>` inside `SmoothScroll`:
+
+```tsx
+<SmoothScroll>
+  <SiteNav />
+  <main className="max-w-6xl mx-auto px-4 sm:px-6 py-8">{children}</main>
+  <SiteFooter />
+</SmoothScroll>
+```
+
+- [ ] **Step 4: Run to verify passes**
+
+Run: `npm run test`, then `npx playwright test e2e/seo.spec.ts e2e/accessibility.spec.ts` (the axe smoke must stay green with the new landmark). Expected: PASS.
+
+- [ ] **Step 5: Commit**
+
+```bash
+git add src/components/layout/SiteFooter.tsx src/app/layout.tsx tests/components/site-footer.test.tsx e2e/seo.spec.ts
+git commit -m "feat: site footer with visible name line"
+```
+
+---
+
+### Task 5: Per-page metadata hygiene (contact OG bug, about description, slug not-found)
+
+**Files:**
+
+- Modify: `src/app/contact/page.tsx:6-12`
+- Modify: `src/app/about/page.tsx:8-13` (description string only)
+- Modify: `src/app/work/[slug]/page.tsx:17`
+- Modify: `tests/integration/contact-page.test.tsx:82-86`
+
+**Interfaces:**
+
+- Consumes: `buildPageMetadata` (existing).
+
+- [ ] **Step 1: Write the failing tests**
+
+In `tests/integration/contact-page.test.tsx`, replace the metadata test (lines 82–86):
+
+```ts
+it('exports contact metadata with its own OpenGraph object', async () => {
+  const { metadata } = await import('@/app/contact/page');
+  const { SITE_AUTHOR } = await import('@/lib/site');
+  expect(metadata.title).toBe('Contact');
+  expect(metadata.alternates?.canonical).toBe('https://hackimi.dev/contact');
+  // Regression: an absent openGraph made /contact inherit the homepage OG.
+  expect(metadata.openGraph?.title).toBe(`Contact | ${SITE_AUTHOR}`);
+  expect(metadata.openGraph?.url).toBe('https://hackimi.dev/contact');
+});
+```
+
+- [ ] **Step 2: Run to verify failure**
+
+Run: `npx vitest run tests/integration/contact-page.test.tsx` Expected: FAIL — `metadata.openGraph` is undefined.
+
+- [ ] **Step 3: Implement**
+
+`src/app/contact/page.tsx` — replace the hand-written metadata (lines 6–12) with:
+
+```ts
+import { buildPageMetadata } from '@/lib/metadata';
+
+export const metadata = buildPageMetadata({
+  title: 'Contact',
+  description: 'Send a message to Nima Hakimi through a secure contact form powered by reCAPTCHA.',
+  path: '/contact',
+});
+```
+
+(Drop the now-unused `import type { Metadata } from 'next';`.)
+
+`src/app/about/page.tsx` — description string (line 10–11) becomes:
+
+```ts
+'Nima Hakimi — developer and AI engineer with a designer’s eye, building accessible, expressive web interfaces. Experience, education, and skills.',
+```
+
+`src/app/work/[slug]/page.tsx` line 17 — unknown slug gets a real title instead of `{}`:
+
+```ts
+if (!c) return { title: 'Not found' };
+```
+
+- [ ] **Step 4: Run to verify passes**
+
+Run: `npm run test` — Expected: PASS (about's description is only asserted via `toContain('Nima Hakimi')`-style checks, which still hold).
+
+- [ ] **Step 5: Commit**
+
+```bash
+git add src/app/contact/page.tsx src/app/about/page.tsx "src/app/work/[slug]/page.tsx" tests/integration/contact-page.test.tsx
+git commit -m "fix: contact OG fallthrough, about description wording, slug not-found title"
+```
+
+---
+
+### Task 6: Entity graph — @id, ORCID, WebSite, ProfilePage, BreadcrumbList
+
+**Files:**
+
+- Modify: `src/lib/metadata.ts` (Person builder + three new builders)
+- Create: `src/components/seo/JsonLd.tsx`
+- Modify: `src/app/layout.tsx` (Person + WebSite via JsonLd), `src/app/about/page.tsx` (ProfilePage), `src/app/work/[slug]/page.tsx` (BreadcrumbList)
+- Modify: `tests/unit/metadata.test.ts`
+- Modify: `e2e/seo.spec.ts` (append)
+
+**Interfaces:**
+
+- Consumes: `SITE_SOCIAL_LINKS.orcid` (Task 2).
+- Produces: `PERSON_ID` const (`https://hackimi.dev/#person`); `buildWebSiteJsonLd()`, `buildProfilePageJsonLd()`, `buildBreadcrumbJsonLd(caseTitle: string, slug: string)`; `JsonLd({ data }: { data: object })` component.
+
+- [ ] **Step 1: Write the failing unit tests**
+
+In `tests/unit/metadata.test.ts`, replace the `buildPersonJsonLd` sameAs assertion and add:
+
+```ts
+describe('buildPersonJsonLd', () => {
+  it('carries a stable @id and all three sameAs anchors', () => {
+    const ld = buildPersonJsonLd();
+    expect(ld['@id']).toBe(`${SITE_URL}/#person`);
+    expect(ld.sameAs).toEqual([SITE_SOCIAL_LINKS.github, SITE_SOCIAL_LINKS.linkedin, SITE_SOCIAL_LINKS.orcid]);
+  });
+});
+
+describe('entity graph builders', () => {
+  it('WebSite names the site after the author and references the person', () => {
+    const ld = buildWebSiteJsonLd();
+    expect(ld['@type']).toBe('WebSite');
+    expect(ld.name).toBe(SITE_AUTHOR);
+    expect(ld.url).toBe(SITE_URL);
+    expect(ld.publisher).toEqual({ '@id': `${SITE_URL}/#person` });
+  });
+
+  it('ProfilePage points its mainEntity at the person @id', () => {
+    const ld = buildProfilePageJsonLd();
+    expect(ld['@type']).toBe('ProfilePage');
+    expect(ld.url).toBe(`${SITE_URL}/about`);
+    expect(ld.mainEntity).toEqual({ '@id': `${SITE_URL}/#person` });
+  });
+
+  it('BreadcrumbList walks Home → Work → case', () => {
+    const ld = buildBreadcrumbJsonLd('Be My Guide', 'be-my-guide');
+    expect(ld['@type']).toBe('BreadcrumbList');
+    expect(ld.itemListElement).toEqual([
+      { '@type': 'ListItem', 'position': 1, 'name': 'Home', 'item': SITE_URL },
+      { '@type': 'ListItem', 'position': 2, 'name': 'Work', 'item': `${SITE_URL}/work` },
+      { '@type': 'ListItem', 'position': 3, 'name': 'Be My Guide', 'item': `${SITE_URL}/work/be-my-guide` },
+    ]);
+  });
+});
+```
+
+(Extend the import list with the three new builders. Keep the existing `builds the expected Person schema` test but update its `sameAs` line to the three-element array.)
+
+- [ ] **Step 2: Run to verify failure**
+
+Run: `npx vitest run tests/unit/metadata.test.ts` — Expected: FAIL, builders undefined.
+
+- [ ] **Step 3: Implement builders in `src/lib/metadata.ts`**
+
+```ts
+export const PERSON_ID = `${SITE_URL}/#person` as const;
+
+export function buildWebSiteJsonLd() {
+  return {
+    '@context': 'https://schema.org',
+    '@type': 'WebSite',
+    'name': SITE_AUTHOR,
+    'url': SITE_URL,
+    'publisher': { '@id': PERSON_ID },
+  } as const;
+}
+
+export function buildProfilePageJsonLd() {
+  return {
+    '@context': 'https://schema.org',
+    '@type': 'ProfilePage',
+    'url': `${SITE_URL}/about`,
+    'mainEntity': { '@id': PERSON_ID },
+  } as const;
+}
+
+export function buildBreadcrumbJsonLd(caseTitle: string, slug: string) {
+  return {
+    '@context': 'https://schema.org',
+    '@type': 'BreadcrumbList',
+    'itemListElement': [
+      { '@type': 'ListItem', 'position': 1, 'name': 'Home', 'item': SITE_URL },
+      { '@type': 'ListItem', 'position': 2, 'name': 'Work', 'item': `${SITE_URL}/work` },
+      { '@type': 'ListItem', 'position': 3, 'name': caseTitle, 'item': `${SITE_URL}/work/${slug}` },
+    ],
+  } as const;
+}
+```
+
+In `buildPersonJsonLd`, add `'@id': PERSON_ID,` directly under `'@type': 'Person',` and change the sameAs line to:
+
+```ts
+'sameAs': [SITE_SOCIAL_LINKS.github, SITE_SOCIAL_LINKS.linkedin, SITE_SOCIAL_LINKS.orcid],
+```
+
+Create `src/components/seo/JsonLd.tsx`:
+
+```tsx
+/** Renders one schema.org block. Server component — emits static JSON, no hydration. */
+export default function JsonLd({ data }: { data: object }) {
+  return (
+    <script
+      type="application/ld+json"
+      dangerouslySetInnerHTML={{ __html: JSON.stringify(data) }}
+      suppressHydrationWarning
+    />
+  );
+}
+```
+
+Wire-in:
+
+- `src/app/layout.tsx`: replace the inline `<script type="application/ld+json" …>` block with `<JsonLd data={buildPersonJsonLd()} />` and `<JsonLd data={buildWebSiteJsonLd()} />` (both where the old script sat, before the grain div). Update imports accordingly; the local `jsonLd` variable goes away.
+- `src/app/about/page.tsx`: inside the returned `<article>`, first child: `<JsonLd data={buildProfilePageJsonLd()} />`.
+- `src/app/work/[slug]/page.tsx`: inside the returned `<article>`, first child: `<JsonLd data={buildBreadcrumbJsonLd(c.title, c.slug)} />` (place after the `notFound()` guard so `c` is non-null).
+
+- [ ] **Step 4: Append the rendered-layer e2e test**
+
+Add to `e2e/seo.spec.ts`:
+
+```ts
+test.describe('structured data', () => {
+  async function ldTypes(page: import('@playwright/test').Page) {
+    const blocks = await page.locator('script[type="application/ld+json"]').allTextContents();
+    return blocks.map((b) => JSON.parse(b)['@type']);
+  }
+
+  test('every page carries Person + WebSite; about and case pages add their own', async ({ page }) => {
+    await page.goto('/');
+    expect((await ldTypes(page)).sort()).toEqual(['Person', 'WebSite']);
+
+    await page.goto('/about');
+    expect((await ldTypes(page)).sort()).toEqual(['Person', 'ProfilePage', 'WebSite']);
+
+    await page.goto('/work/be-my-guide');
+    expect((await ldTypes(page)).sort()).toEqual(['BreadcrumbList', 'Person', 'WebSite']);
+  });
+});
+```
+
+- [ ] **Step 5: Run to verify passes**
+
+Run: `npm run test` and `npx playwright test e2e/seo.spec.ts`. Expected: PASS. (JSON.parse doubles as a validity check on every block.)
+
+- [ ] **Step 6: Commit**
+
+```bash
+git add src/lib/metadata.ts src/components/seo/JsonLd.tsx src/app/layout.tsx src/app/about/page.tsx "src/app/work/[slug]/page.tsx" tests/unit/metadata.test.ts e2e/seo.spec.ts
+git commit -m "feat: entity graph JSON-LD (person @id, WebSite, ProfilePage, breadcrumbs)"
+```
+
+---
+
+### Task 7: Brand assets — favicon set and OG card ⚠️ TASTE GATE
+
+**Files:**
+
+- Create: `scripts/generate-brand-assets.mjs`, `src/app/icon.svg`, `src/app/apple-icon.png`, `public/og.png`
+- Modify: `src/lib/site.ts` (`SITE_OG_IMAGE` → og.png; new `SITE_PERSON_IMAGE`), `src/lib/metadata.ts` (Person image), `tests/unit/site.test.ts`, `tests/unit/metadata.test.ts`
+- Modify: `e2e/seo.spec.ts` (append)
+
+**Interfaces:**
+
+- Consumes: Monogram geometry from `src/components/brand/Monogram.tsx` (120×120 viewBox; N: `M20 100 L20 20 L54 100 L54 20`; H: `M66 20 L66 100 M66 60 L100 60 M100 20 L100 100`; strokeWidth 8, round caps/joins).
+- Produces: `SITE_PERSON_IMAGE` (the face photo, kept for Person JSON-LD), `SITE_OG_IMAGE` now `${SITE_URL}/og.png`.
+
+- [ ] **Step 1: Write the failing tests**
+
+`tests/unit/site.test.ts`, add:
+
+```ts
+it('OG image is the branded card; person image is the photo', async () => {
+  const { SITE_OG_IMAGE, SITE_PERSON_IMAGE, SITE_URL } = await import('@/lib/site');
+  expect(SITE_OG_IMAGE).toBe(`${SITE_URL}/og.png`);
+  expect(SITE_PERSON_IMAGE).toBe(`${SITE_URL}/bigSmile.JPEG`);
+});
+```
+
+`tests/unit/metadata.test.ts`, add inside the `buildPersonJsonLd` describe:
+
+```ts
+it('uses the person photo, not the OG card, as the Person image', () => {
+  expect(buildPersonJsonLd().image).toBe(SITE_PERSON_IMAGE);
+});
+```
+
+(add `SITE_PERSON_IMAGE` to that file's imports from `@/lib/site`.)
+
+`e2e/seo.spec.ts`, append inside `rendered metadata`:
+
+```ts
+test('favicon links are emitted', async ({ page }) => {
+  await page.goto('/');
+  await expect(page.locator('link[rel="icon"]').first()).toHaveAttribute('href', /icon/);
+});
+
+test('og image resolves to a real asset', async ({ page, request }) => {
+  await page.goto('/');
+  const og = await page.locator('meta[property="og:image"]').getAttribute('content');
+  expect(og).toContain('/og.png');
+  const res = await request.get('/og.png');
+  expect(res.status()).toBe(200);
+});
+```
+
+- [ ] **Step 2: Run to verify failures**
+
+Run: `npx vitest run tests/unit/site.test.ts tests/unit/metadata.test.ts` — Expected: FAIL (`SITE_PERSON_IMAGE` missing).
+
+- [ ] **Step 3: Author the icon and the generator script**
+
+Create `src/app/icon.svg` (Next serves this as the favicon automatically):
+
+```svg
+<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 120 120">
+  <rect width="120" height="120" rx="26" fill="#0E0E10"/>
+  <g fill="none" stroke="#C6FF3D" stroke-width="11" stroke-linecap="round" stroke-linejoin="round" transform="translate(60 60) scale(0.78) translate(-60 -60)">
+    <path d="M20 100 L20 20 L54 100 L54 20"/>
+    <path d="M66 20 L66 100 M66 60 L100 60 M100 20 L100 100"/>
+  </g>
+</svg>
+```
+
+Create `scripts/generate-brand-assets.mjs` (Playwright is already a devDependency; this renders two HTML canvases and screenshots them):
+
+```js
+// Generates src/app/apple-icon.png (180x180) and public/og.png (1200x630)
+// from the NH monogram. Rerun after any monogram/palette change:
+//   ulimit -n 10240 && node scripts/generate-brand-assets.mjs
+import { chromium } from '@playwright/test';
+import { readFileSync } from 'node:fs';
+
+const iconSvg = readFileSync(new URL('../src/app/icon.svg', import.meta.url), 'utf8');
+
+const MONOGRAM = `
+  <g fill="none" stroke="#C6FF3D" stroke-width="8" stroke-linecap="round" stroke-linejoin="round">
+    <path d="M20 100 L20 20 L54 100 L54 20"/>
+    <path d="M66 20 L66 100 M66 60 L100 60 M100 20 L100 100"/>
+  </g>`;
+
+const ogHtml = `<!doctype html><html><body style="margin:0">
+  <div style="width:1200px;height:630px;background:#0E0E10;display:grid;place-items:center">
+    <svg viewBox="0 0 120 120" width="380" height="380">${MONOGRAM}</svg>
+  </div></body></html>`;
+
+const appleHtml = `<!doctype html><html><body style="margin:0">
+  <div style="width:180px;height:180px;display:grid;place-items:center;background:#0E0E10">
+    ${iconSvg.replace('<svg ', '<svg width="180" height="180" ')}
+  </div></body></html>`;
+
+const browser = await chromium.launch();
+const page = await browser.newPage();
+
+await page.setViewportSize({ width: 1200, height: 630 });
+await page.setContent(ogHtml);
+await page.locator('div').first().screenshot({ path: 'public/og.png' });
+
+await page.setViewportSize({ width: 180, height: 180 });
+await page.setContent(appleHtml);
+await page.locator('div').first().screenshot({ path: 'src/app/apple-icon.png' });
+
+await browser.close();
+console.log('wrote public/og.png and src/app/apple-icon.png');
+```
+
+Run it (unsandboxed Bash, `ulimit -n 10240` first): `node scripts/generate-brand-assets.mjs` Verify dims: `sips -g pixelWidth -g pixelHeight public/og.png src/app/apple-icon.png` → 1200×630 and 180×180.
+
+- [ ] **Step 4: Switch the constants**
+
+`src/lib/site.ts`:
+
+```ts
+export const SITE_OG_IMAGE = `${SITE_URL}/og.png` as const;
+export const SITE_PERSON_IMAGE = `${SITE_URL}/bigSmile.JPEG` as const;
+```
+
+`src/lib/metadata.ts`: import `SITE_PERSON_IMAGE` and use it for the Person builder's `image` field (the OG `defaultImage` keeps using `SITE_OG_IMAGE`, whose declared 1200×630 now matches the actual file).
+
+- [ ] **Step 5: Run to verify passes**
+
+Run: `npm run test` and `npx playwright test e2e/seo.spec.ts` — Expected: PASS.
+
+- [ ] **Step 6: ⚠️ TASTE GATE — STOP and show Nima (ORCHESTRATOR-ONLY step)**
+
+A subagent implementer CANNOT perform this step — it must return after Step 5 WITHOUT committing, reporting the generated asset paths. The orchestrator then sends Nima `public/og.png`, `src/app/apple-icon.png`, and a browser-tab-size rendering of `src/app/icon.svg` (screenshot of the actual favicon in a tab, or a 32px render), plays the approval sound, and waits. **Nothing merges past this point without his explicit approval of both assets.** Iterate on his feedback (stroke weight, corner radius, monogram scale) by editing the SVG/script and re-running the generator.
+
+- [ ] **Step 7: Commit (after approval)**
+
+```bash
+git add src/app/icon.svg src/app/apple-icon.png public/og.png scripts/generate-brand-assets.mjs src/lib/site.ts src/lib/metadata.ts tests/unit/site.test.ts tests/unit/metadata.test.ts e2e/seo.spec.ts
+git commit -m "feat: NH monogram favicon set and branded OG card"
+```
+
+---
+
+### Task 8: Honest sitemap + idiomatic robots.ts
+
+**Files:**
+
+- Modify: `src/app/sitemap.ts`, `tests/server/sitemap.test.ts`
+- Create: `src/app/robots.ts`
+- Delete: `src/app/robots.txt/` (whole folder)
+- Rewrite: `tests/server/robots.test.ts`
+
+- [ ] **Step 1: Write the failing tests**
+
+Rewrite `tests/server/sitemap.test.ts`'s last test (replace the changeFrequency/lastModified one):
+
+```ts
+it('entries are url-only — no fabricated lastModified/changeFrequency/priority', () => {
+  for (const entry of sitemap()) {
+    expect(Object.keys(entry)).toEqual(['url']);
+  }
+});
+```
+
+Also update the two priority assertions (`priority 1` / `priority 0.6` tests): drop the priority expectations, keep the URL-shape ones — rename to "uses the bare base URL (no trailing slash) for '/'" and "uses the full path for a non-root route".
+
+Rewrite `tests/server/robots.test.ts`:
+
+```ts
+import { describe, expect, it } from 'vitest';
+
+import robots from '@/app/robots';
+import { SITE_URL } from '@/lib/site';
+
+describe('robots', () => {
+  it('allows all crawlers everything and points at the sitemap', () => {
+    expect(robots()).toEqual({
+      rules: { userAgent: '*', allow: '/' },
+      sitemap: `${SITE_URL}/sitemap.xml`,
+    });
+  });
+});
+```
+
+- [ ] **Step 2: Run to verify failure**
+
+Run: `npx vitest run tests/server/` — Expected: FAIL (no `@/app/robots` module; sitemap entries carry extra keys).
+
+- [ ] **Step 3: Implement**
+
+Replace `src/app/sitemap.ts`:
+
+```ts
+import type { MetadataRoute } from 'next';
+import work from '@/data/work';
+import { SITE_URL } from '@/lib/site';
+
+const routes = ['/', '/about', '/work', '/contact', ...work.map((c) => `/work/${c.slug}`)] as const;
+
+/**
+ * URL list only, deliberately: the previous lastModified stamped `new Date()`
+ * on every build (a false freshness signal on all URLs), and Google ignores
+ * changeFrequency/priority. No signal beats a wrong signal.
+ */
+export default function sitemap(): MetadataRoute.Sitemap {
+  return routes.map((route) => ({ url: `${SITE_URL}${route === '/' ? '' : route}` }));
+}
+```
+
+Create `src/app/robots.ts`:
+
+```ts
+import type { MetadataRoute } from 'next';
+import { SITE_URL } from '@/lib/site';
+
+export default function robots(): MetadataRoute.Robots {
+  return {
+    rules: { userAgent: '*', allow: '/' },
+    sitemap: `${SITE_URL}/sitemap.xml`,
+  };
+}
+```
+
+Delete the folder `src/app/robots.txt/` (`git rm -r src/app/robots.txt`).
+
+- [ ] **Step 4: Run to verify passes**
+
+Run: `npm run test`, then boot the dev server and check both endpoints render: `curl -s http://localhost:3000/robots.txt` → `User-Agent: *`, `Allow: /`, `Sitemap: https://hackimi.dev/sitemap.xml` `curl -s http://localhost:3000/sitemap.xml` → 7 `<loc>` entries, no `<lastmod>`.
+
+- [ ] **Step 5: Commit**
+
+```bash
+git add src/app/sitemap.ts src/app/robots.ts tests/server/   # robots.txt deletion already staged by git rm
+git commit -m "refactor: url-only sitemap, robots.ts file convention, SITE_URL single-sourced"
+```
+
+---
+
+### Task 9: Search Console verification slot + infra docs
+
+**Files:**
+
+- Modify: `src/lib/site.ts`, `src/lib/metadata.ts` (`buildRootMetadata`), `tests/unit/metadata.test.ts`
+- Create: `docs/infra.md`
+
+- [ ] **Step 1: Write the failing test**
+
+Append to `tests/unit/metadata.test.ts`'s `buildRootMetadata` describe:
+
+```ts
+it('emits no verification block while the GSC token is empty', () => {
+  // SITE_GOOGLE_SITE_VERIFICATION is '' until Nima creates the property;
+  // an empty <meta name="google-site-verification"> must never ship.
+  expect(buildRootMetadata().verification).toBeUndefined();
+});
+```
+
+- [ ] **Step 2: Run to verify failure**
+
+Run: `npx vitest run tests/unit/metadata.test.ts` — Expected: FAIL only if the implementation adds an empty block; if it passes trivially (no `verification` key exists yet), proceed — the test pins the contract for when the token lands.
+
+- [ ] **Step 3: Implement**
+
+`src/lib/site.ts`:
+
+```ts
+/** Google Search Console meta-tag token. Empty until the property is created. */
+export const SITE_GOOGLE_SITE_VERIFICATION = '';
+```
+
+`src/lib/metadata.ts` — in `buildRootMetadata`'s returned object, add:
+
+```ts
+...(SITE_GOOGLE_SITE_VERIFICATION
+  ? { verification: { google: SITE_GOOGLE_SITE_VERIFICATION } }
+  : {}),
+```
+
+(and import the constant).
+
+Create `docs/infra.md`:
+
+```markdown
+# Infrastructure notes (state that lives outside this repo)
+
+## Domains (Vercel dashboard → Settings → Domains)
+
+- Production domains: `hackimi.dev` (apex) and `www.hackimi.dev`.
+- **Canonical host: apex `https://hackimi.dev`** — every canonical tag, sitemap URL, robots Sitemap line, and JSON-LD url in this repo uses apex.
+- Redirect direction: `www` → 308 → apex. (Flipped 2026-08-\_\_ by Nima; before that the dashboard had apex → www, contradicting the code's canonicals.)
+- `hackimi.vercel.app` additionally 308s to apex via `next.config.js` (exact-host rule — preview URLs on \*.vercel.app are deliberately not matched).
+
+## Deployments
+
+- `vercel.json` sets `git.deploymentEnabled.main: false` — merging to `main` does NOT auto-deploy. Production deploys are triggered manually (dashboard or `vercel --prod`).
+
+## Search Console
+
+- Property: `https://hackimi.dev` (URL-prefix), meta-tag verification.
+- Token lives in `SITE_GOOGLE_SITE_VERIFICATION` (`src/lib/site.ts`); empty string disables the tag entirely.
+- After each SEO-relevant deploy: submit `sitemap.xml`, request indexing of changed URLs.
+```
+
+(The redirect-direction line contains a deliberate `2026-08-__` blank — Nima fills the date when he flips it; leave as-is.)
+
+- [ ] **Step 4: Run to verify passes**
+
+Run: `npm run test` — Expected: PASS.
+
+- [ ] **Step 5: Commit**
+
+```bash
+git add src/lib/site.ts src/lib/metadata.ts tests/unit/metadata.test.ts docs/infra.md
+git commit -m "feat: search console verification slot; document domain/deploy state"
+```
+
+---
+
+### Task 10: Full verification + PR
+
+- [ ] **Step 1: Full local gate**
+
+Run, in order, all from the repo root:
+
+```bash
+npm run lint        # must be 0 errors / 0 warnings
+npm run check       # Prettier clean
+npm run test        # full Vitest suite
+npm run build       # production build succeeds
+```
+
+Then the full e2e suite against a production server (unsandboxed, `ulimit -n 10240`):
+
+```bash
+npx playwright test
+```
+
+Expected: everything green. A single flaky failure on a cold Turbopack dev server is a known non-signal — but this run is against the prod build, which is stable; failures here are real.
+
+- [ ] **Step 2: Verify no stray "frontend" leaked into SEO surfaces**
+
+```bash
+grep -rn 'rontend' src/lib/ src/app/sitemap.ts src/app/robots.ts next.config.js
+```
+
+Expected: zero hits. (Visible-copy hits in `page.tsx`/`about`/`ContactClient`/`src/data/` are correct and expected — those stay.)
+
+- [ ] **Step 3: Push and open PR**
+
+Push (its own command, per git-push-guard): `git push -u origin seo/name-query`
+
+PR against `main`, title "SEO: win the name query". Body: summary of the seven workstreams, a plain statement that the OG card and favicon were taste-approved (no links of any kind to sessions/conversations), and the post-merge checklist for Nima (Vercel apex flip → manual deploy → GSC property + token → sitemap submit → recrawl requests → LinkedIn/GitHub profile updates). No AI attribution anywhere. Note: `gh` requires unsandboxed Bash on this machine (sandboxed TLS to api.github.com fails).
+
+- [ ] **Step 4: Post-deploy verification (after Nima merges + deploys)**
+
+```bash
+curl -s -o /dev/null -w '%{http_code} %{redirect_url}\n' https://www.hackimi.dev/projects   # 308 → /work
+curl -s -o /dev/null -w '%{http_code} %{redirect_url}\n' https://hackimi.vercel.app/        # 308 → https://hackimi.dev/
+curl -s https://www.hackimi.dev/ | grep -o '<title>[^<]*</title>'                            # name-led title
+curl -s https://www.hackimi.dev/sitemap.xml | grep -c '<loc>'                                # 7
+```
+
+And after his Vercel flip: `curl -s -o /dev/null -w '%{http_code} %{redirect_url}\n' https://www.hackimi.dev/` → `308 https://hackimi.dev/`. Update `docs/infra.md`'s flip date. Move the ROADMAP.md row to Done.
diff --git a/docs/superpowers/specs/2026-08-09-seo-name-query-design.md b/docs/superpowers/specs/2026-08-09-seo-name-query-design.md
new file mode 100644
index 0000000..704d6f3
--- /dev/null
+++ b/docs/superpowers/specs/2026-08-09-seo-name-query-design.md
@@ -0,0 +1,117 @@
+# SEO: win the name query — design spec
+
+**Date:** 2026-08-09 **Goal:** hackimi.dev is what people find when they search for Nima Hakimi. **Branch:** `seo/name-query`
+
+## Acceptance criteria
+
+1. **#1 for qualified name queries** — "Nima Hakimi developer", "Nima Hakimi AI engineer", "Nima Hakimi utvikler" — measured in Google Search Console once set up.
+2. **hackimi.dev on page 1 for bare "Nima Hakimi" searched from Norway.**
+3. Global bare-name ranking is the long-game direction, not a gate: page 1 for the bare query is currently fully occupied by the Convoso CEO Nima Hakimi (press, speaker profiles, podcasts, LinkedIn since 2006). No code change can beat that authority head-on; only third-party mentions can, over time.
+
+## SERP evidence (recon 2026-08-09)
+
+- Bare "Nima Hakimi": hackimi.dev absent from page 1 (all slots: Convoso CEO + LinkedIn directory + Wikipedia surname pages). 6–7 distinct people share the name online.
+- "Nima Hakimi developer": `hackimi.dev/projects` ranks **#1** — and **404s** (route renamed to `/work` in the 2026-06-02 redesign, no redirect). Homepage ranks #5.
+- "Nima Hakimi utvikler": positions 6 and 9.
+- "Nima Hakimi" + NAV/frontend: positions 1–2. The problem is specific to the unqualified query.
+- `hackimi.vercel.app` is live, indexable (200, no noindex), and surfaces in SERPs — a duplicate of production.
+- Search vantage was non-Norwegian; Norwegian results are likely already better. GSC will give real data.
+
+## Decisions (grill round, settled 2026-08-09)
+
+| Decision | Outcome |
+| --- | --- |
+| Canonical host | **Apex `hackimi.dev`.** Nima flips Vercel's primary domain in the dashboard (his hands); code already points at apex everywhere and stays. Dashboard state gets documented in-repo. |
+| Positioning wording | **"Developer & AI Engineer"** in all SEO surfaces this spec touches (title, jobTitle, descriptions, footer). Visible site copy (hero, about prose, case studies, contact blurb) keeps "frontend" for now — content repositioning is a separate later job Nima owns. Temporary mismatch accepted. |
+| Search Console | Not set up yet. Code ships a verification-token slot; Nima creates the property and supplies the token (separate tiny commit). |
+| OG image | Simple branded card: **NH monogram on near-black with acid-lime accent**, static 1200×630. No name/role text on the card (Nima's call). Taste-gated: he sees the built image before it ships. |
+| Favicon | Built from the NH monogram. Taste-gated the same way. |
+| Out of scope | Blog/content marketing; per-page dynamic OG images (`docs/features/dynamic-og-images.md` stays unexecuted — abandoned-direction artifact); any visible-copy repositioning. |
+
+## Workstream 1 — Recover dead ranking equity (ships first)
+
+`next.config.js` permanent redirects:
+
+- `/projects` → `/work` (the currently-#1, currently-404 URL; the only content route the old site had — verified via `git log --diff-filter=D`, no `/projects/:slug` children, no pages-router history).
+- `/projects/:path*` → `/work` (insurance against indexed variants).
+- Host-based: **exact host** `hackimi.vercel.app` → `https://hackimi.dev/:path*`, permanent. Exact match only — `*.vercel.app` preview deployments must stay untouched.
+
+**Acceptance:** `curl -I https://www.hackimi.dev/projects` → 308 to `/work`; e2e asserts the redirect locally; vercel.app redirect verified post-deploy (host matching hard to e2e locally).
+
+## Workstream 2 — Canonical consistency
+
+- Nima (dashboard): set `hackimi.dev` as the primary domain in Vercel so www 308s to apex, matching every code signal. Currently it's inverted (apex 308s to www while all canonicals/sitemap/JSON-LD say apex — conflicting signals).
+- Code: consolidate the 3× independently hardcoded `https://hackimi.dev` — `src/app/sitemap.ts` and the robots implementation import `SITE_URL` from `src/lib/site.ts`.
+- Convert `src/app/robots.txt/route.ts` (route-handler folder) to the idiomatic `src/app/robots.ts` (`MetadataRoute.Robots`). Same output: allow all, sitemap line.
+- Document the Vercel domain setup (primary domain, redirect direction, who owns it) in `docs/infra.md` so the dashboard state is no longer invisible to version control.
+
+## Workstream 3 — Name-led metadata
+
+Files: `src/lib/site.ts`, `src/lib/metadata.ts`, `src/app/page.tsx`, `src/app/about/page.tsx`, `src/app/contact/page.tsx`, root `layout.tsx` (footer).
+
+- `SITE_ROLE` = `'Developer & AI Engineer'`; `SITE_TITLE` = `` `${SITE_AUTHOR} — ${SITE_ROLE}` `` (em-dash form; reads better in tabs/SERPs than the pipe).
+- **Homepage title becomes the explicit full string "Nima Hakimi — Developer & AI Engineer".** Root cause being fixed: Next's `title.template` applies only to child segments, so the root `page.tsx`'s `title: 'Portfolio'` rendered as literally `<title>Portfolio</title>` — no name on the site's most important page. The fix must be asserted at the rendered layer (see Testing) because the metadata _object_ was always "correct" while the resolved output was wrong.
+- `SITE_DESCRIPTION` and the home/about meta descriptions drop "frontend developer" for "developer and AI engineer" phrasing. `SITE_KEYWORDS` → `['Nima Hakimi', 'Developer', 'AI Engineer', 'Next.js', 'React', 'Oslo']` (Google ignores the tag; kept harmless and consistent).
+- **Footer (new, site-wide):** visible name line — `© <year> Nima Hakimi — Developer & AI Engineer, Oslo`. Today the name exists as visible text only in the hero h1; this adds one crawlable, every-page occurrence. Small server component, near-black/muted styling consistent with the design system, no navigation duties.
+- **Contact page** reroutes through `buildPageMetadata()` — fixes the shipping bug where its unspecified `openGraph` falls through to the root layout's homepage OG object.
+- `work/[slug]/generateMetadata` returns a real not-found title instead of `{}` on unknown slugs (minor hygiene).
+
+## Workstream 4 — Entity graph (structured data)
+
+Files: `src/lib/metadata.ts` (builders), `src/lib/site.ts` (links), page layouts.
+
+- Person JSON-LD gains `@id: 'https://hackimi.dev/#person'` and ORCID (`https://orcid.org/0009-0002-6656-2498`, already public on his GitHub profile) added to `sameAs` alongside GitHub + LinkedIn. `jobTitle` follows `SITE_ROLE`.
+- New **WebSite** block (all pages): name "Nima Hakimi", url apex, `publisher`/`author` referencing the Person `@id`. This is the block Google ties to site-name display in results.
+- New **ProfilePage** block on `/about` with `mainEntity` → Person `@id`.
+- New **BreadcrumbList** on `/work/[slug]`: Home → Work → case title.
+- All emitted as separate `<script type="application/ld+json">` blocks; builders unit-tested like `buildPersonJsonLd` already is.
+
+## Workstream 5 — Assets (taste-gated)
+
+- **Favicon:** none exists today — Google renders a generic globe next to every hackimi.dev result. Add `src/app/icon.svg` (NH monogram, acid-lime on near-black) + `src/app/apple-icon.png` (180×180) via Next file conventions. Legacy `favicon.ico` fallback if the svg-only setup leaves gaps.
+- **OG card:** replace `bigSmile.JPEG` (1536×2048 portrait declared as 1200×630 landscape — platforms may crop/reject) with a committed static `public/og.png`, 1200×630: NH monogram on near-black, acid-lime accent, simple. Declared dims match actual dims.
+- **Gate:** both assets rendered and screenshotted for Nima's approval before merge. The monogram SVG already exists in the codebase (`Monogram` component) — reuse its paths.
+
+## Workstream 6 — Sitemap honesty
+
+`src/app/sitemap.ts`: today every URL stamps `lastModified: new Date()` on every build — a false freshness signal on all 7 URLs. Google ignores `changeFrequency`/`priority` entirely. The sitemap becomes a clean URL list (no lastmod/changefreq/priority): no signal beats a false signal, and hand-maintained dates would rot. `tests/server/sitemap.test.ts` updated to pin the new shape.
+
+## Workstream 7 — Search Console wiring
+
+- `verification: { google: GSC_VERIFICATION_TOKEN }` slot in `buildRootMetadata`, constant in `site.ts`, filled when Nima supplies the token (empty/absent until then — must not emit an empty tag).
+- After deploy + verification, GSC gives the real index inventory (the `site:` operator probes were unreliable), recrawl requests for `/`, `/work`, `/projects` (to pick up the redirect), and the measurement layer for the acceptance criteria.
+
+## Off-site checklist (Nima's hands — delivered with the PR)
+
+1. **Vercel:** flip primary domain to `hackimi.dev` (Settings → Domains). One minute. Tell me when done so I verify the redirect direction flipped.
+2. **Google Search Console:** create property for `https://hackimi.dev` (URL-prefix), choose meta-tag verification, send me the token. After deploy: submit `sitemap.xml`, request indexing of `/`, `/work`, `/projects`.
+3. **LinkedIn:** add `hackimi.dev` to the profile website/contact section; confirm public-profile visibility is on. Your LinkedIn never surfaced in any name search — that slot is currently occupied by other Nima Hakimis.
+4. **GitHub nit:** put `https://hackimi.dev` in the dedicated website field (API shows `blog: ""`; it's currently only in social links).
+5. **Long game (no action now):** third-party mentions — talks, articles, podcasts — are the only lever that eventually contests the bare-name query globally.
+
+## Testing
+
+TDD throughout, repo conventions (`tests/unit`, `tests/server`, `tests/integration`, `e2e/`).
+
+- **Rendered-layer first:** new `e2e/seo.spec.ts` — the homepage-title bug was invisible to unit tests (object right, resolution wrong), so the RED tests assert what a crawler sees: exact homepage `<title>`, canonical hrefs per page, `/projects` → `/work` redirect status, presence + JSON-parseability of each JSON-LD block per page type, favicon link tags, footer name line visible.
+- **Unit:** metadata builders (new titles/descriptions/verification slot), new JSON-LD builders (WebSite/ProfilePage/BreadcrumbList, `@id` wiring), `resolveUrl` unchanged.
+- **Server:** sitemap new shape; `robots.ts` output equivalent to the old route handler (allow all + sitemap line).
+- **Post-deploy verification (not in CI):** curl the production redirects (www→apex after the flip, vercel.app→apex, /projects→/work), Google Rich Results test on `/` and one case study.
+
+## Sequencing
+
+1. Redirects (equity recovery — the dead #1 URL)
+2. Name-led metadata + footer
+3. Entity graph JSON-LD
+4. Assets (favicon, OG card) — taste gates
+5. Sitemap/robots consolidation
+6. GSC wiring + `docs/infra.md`
+7. PR → merge → deploy → his checklist → post-deploy verification → ROADMAP.md Done entry
+
+Single branch `seo/name-query`, PR to `main`. No AI attribution anywhere in commits/PR (standing rule).
+
+## Non-goals
+
+- Changing visible site copy/positioning (his separate later job).
+- Blog, per-page dynamic OG images, content marketing.
+- Beating the Convoso CEO on the bare global query as an acceptance criterion.
diff --git a/e2e/seo.spec.ts b/e2e/seo.spec.ts
new file mode 100644
index 0000000..5cff335
--- /dev/null
+++ b/e2e/seo.spec.ts
@@ -0,0 +1,70 @@
+import { test, expect } from '@playwright/test';
+
+test.describe('redirects', () => {
+  test('/projects permanently redirects to /work', async ({ request }) => {
+    const res = await request.get('/projects', { maxRedirects: 0 });
+    expect(res.status()).toBe(308);
+    expect(res.headers()['location']).toBe('/work');
+  });
+
+  test('/projects/anything redirects to /work', async ({ request }) => {
+    const res = await request.get('/projects/old-page', { maxRedirects: 0 });
+    expect(res.status()).toBe(308);
+    expect(res.headers()['location']).toBe('/work');
+  });
+});
+
+test.describe('rendered metadata', () => {
+  test('homepage title is the full name-led title', async ({ page }) => {
+    await page.goto('/');
+    await expect(page).toHaveTitle('Nima Hakimi — Developer & AI Engineer');
+  });
+
+  test('homepage canonical points at apex', async ({ page }) => {
+    await page.goto('/');
+    await expect(page.locator('link[rel="canonical"]')).toHaveAttribute('href', 'https://hackimi.dev');
+  });
+
+  test('child-page canonical points at its own apex URL', async ({ page }) => {
+    await page.goto('/about');
+    await expect(page.locator('link[rel="canonical"]')).toHaveAttribute('href', 'https://hackimi.dev/about');
+  });
+
+  test('footer carries the visible name line on every page', async ({ page }) => {
+    for (const path of ['/', '/about', '/work']) {
+      await page.goto(path);
+      await expect(page.locator('footer')).toContainText('Nima Hakimi — Developer & AI Engineer');
+    }
+  });
+
+  test('favicon links are emitted', async ({ page }) => {
+    await page.goto('/');
+    await expect(page.locator('link[rel="icon"]').first()).toHaveAttribute('href', /icon/);
+  });
+
+  test('og image resolves to a real asset', async ({ page, request }) => {
+    await page.goto('/');
+    const og = await page.locator('meta[property="og:image"]').getAttribute('content');
+    expect(og).toContain('/og.png');
+    const res = await request.get('/og.png');
+    expect(res.status()).toBe(200);
+  });
+});
+
+test.describe('structured data', () => {
+  async function ldTypes(page: import('@playwright/test').Page) {
+    const blocks = await page.locator('script[type="application/ld+json"]').allTextContents();
+    return blocks.map((b) => JSON.parse(b)['@type']);
+  }
+
+  test('every page carries Person + WebSite; about and case pages add their own', async ({ page }) => {
+    await page.goto('/');
+    expect((await ldTypes(page)).sort()).toEqual(['Person', 'WebSite']);
+
+    await page.goto('/about');
+    expect((await ldTypes(page)).sort()).toEqual(['Person', 'ProfilePage', 'WebSite']);
+
+    await page.goto('/work/be-my-guide');
+    expect((await ldTypes(page)).sort()).toEqual(['BreadcrumbList', 'Person', 'WebSite']);
+  });
+});
diff --git a/next.config.js b/next.config.js
index 658404a..568fd04 100644
--- a/next.config.js
+++ b/next.config.js
@@ -1,4 +1,20 @@
 /** @type {import('next').NextConfig} */
-const nextConfig = {};
+const nextConfig = {
+  async redirects() {
+    return [
+      // The pre-2026-06-02 site served /projects (Google's current #1 result
+      // for "Nima Hakimi developer"); the redesign renamed it to /work.
+      { source: '/projects', destination: '/work', permanent: true },
+      { source: '/projects/:path*', destination: '/work', permanent: true },
+      // Exact-host only: preview deployments on *.vercel.app must stay untouched.
+      {
+        source: '/:path*',
+        has: [{ type: 'host', value: 'hackimi.vercel.app' }],
+        destination: 'https://hackimi.dev/:path*',
+        permanent: true,
+      },
+    ];
+  },
+};
 
 module.exports = nextConfig;
diff --git a/public/og.png b/public/og.png
new file mode 100644
index 0000000..5517519
Binary files /dev/null and b/public/og.png differ
diff --git a/scripts/generate-brand-assets.mjs b/scripts/generate-brand-assets.mjs
new file mode 100644
index 0000000..e05f779
--- /dev/null
+++ b/scripts/generate-brand-assets.mjs
@@ -0,0 +1,37 @@
+// Generates src/app/apple-icon.png (180x180) and public/og.png (1200x630)
+// from the NH monogram. Rerun after any monogram/palette change:
+//   ulimit -n 10240 && node scripts/generate-brand-assets.mjs
+import { chromium } from '@playwright/test';
+import { readFileSync } from 'node:fs';
+
+const iconSvg = readFileSync(new URL('../src/app/icon.svg', import.meta.url), 'utf8');
+
+const MONOGRAM = `
+  <g fill="none" stroke="#C6FF3D" stroke-width="8" stroke-linecap="round" stroke-linejoin="round">
+    <path d="M20 100 L20 20 L54 100 L54 20"/>
+    <path d="M66 20 L66 100 M66 60 L100 60 M100 20 L100 100"/>
+  </g>`;
+
+const ogHtml = `<!doctype html><html><body style="margin:0">
+  <div style="width:1200px;height:630px;background:#0E0E10;display:grid;place-items:center">
+    <svg viewBox="0 0 120 120" width="380" height="380">${MONOGRAM}</svg>
+  </div></body></html>`;
+
+const appleHtml = `<!doctype html><html><body style="margin:0">
+  <div style="width:180px;height:180px;display:grid;place-items:center;background:#0E0E10">
+    ${iconSvg.replace('<svg ', '<svg width="180" height="180" ')}
+  </div></body></html>`;
+
+const browser = await chromium.launch();
+const page = await browser.newPage();
+
+await page.setViewportSize({ width: 1200, height: 630 });
+await page.setContent(ogHtml);
+await page.locator('div').first().screenshot({ path: 'public/og.png' });
+
+await page.setViewportSize({ width: 180, height: 180 });
+await page.setContent(appleHtml);
+await page.locator('div').first().screenshot({ path: 'src/app/apple-icon.png' });
+
+await browser.close();
+console.log('wrote public/og.png and src/app/apple-icon.png');
diff --git a/src/app/about/page.tsx b/src/app/about/page.tsx
index 3a3cae5..8d7db59 100644
--- a/src/app/about/page.tsx
+++ b/src/app/about/page.tsx
@@ -1,14 +1,15 @@
 import about from '@/data/about';
-import { buildPageMetadata } from '@/lib/metadata';
+import { buildPageMetadata, buildProfilePageJsonLd } from '@/lib/metadata';
 import { SITE_EMAIL } from '@/lib/site';
 import Reveal from '@/components/motion/Reveal';
 import UnderlineLink from '@/components/motion/UnderlineLink';
 import MagneticButton from '@/components/motion/MagneticButton';
+import JsonLd from '@/components/seo/JsonLd';
 
 export const metadata = buildPageMetadata({
   title: 'About',
   description:
-    'Nima Hakimi — frontend developer with a designer’s eye, building accessible, expressive web interfaces. Experience, education, and skills.',
+    'Nima Hakimi — developer and AI engineer with a designer’s eye, building accessible, expressive web interfaces. Experience, education, and skills.',
   path: '/about',
 });
 
@@ -24,6 +25,7 @@ function SectionLabel({ children }: { children: React.ReactNode }) {
 export default function AboutPage() {
   return (
     <article className="py-12 sm:py-16">
+      <JsonLd data={buildProfilePageJsonLd()} />
       {/* Intro — developer-first identity, design taste as the edge */}
       <header className="max-w-5xl">
         <Reveal>
diff --git a/src/app/apple-icon.png b/src/app/apple-icon.png
new file mode 100644
index 0000000..82267da
Binary files /dev/null and b/src/app/apple-icon.png differ
diff --git a/src/app/contact/page.tsx b/src/app/contact/page.tsx
index 83f7e8e..877baea 100644
--- a/src/app/contact/page.tsx
+++ b/src/app/contact/page.tsx
@@ -1,15 +1,13 @@
-import type { Metadata } from 'next';
+import { buildPageMetadata } from '@/lib/metadata';
 import ContactClient from './ContactClient';
 
 export const runtime = 'nodejs';
 
-export const metadata: Metadata = {
+export const metadata = buildPageMetadata({
   title: 'Contact',
   description: 'Send a message to Nima Hakimi through a secure contact form powered by reCAPTCHA.',
-  alternates: {
-    canonical: 'https://hackimi.dev/contact',
-  },
-};
+  path: '/contact',
+});
 
 export default function ContactPage() {
   const siteKey = process.env.NEXT_PUBLIC_RECAPTCHA_SITE_KEY || process.env.RECAPTCHA_SITE_KEY || '';
diff --git a/src/app/icon.svg b/src/app/icon.svg
new file mode 100644
index 0000000..50ceedc
--- /dev/null
+++ b/src/app/icon.svg
@@ -0,0 +1,7 @@
+<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 120 120">
+  <rect width="120" height="120" rx="26" fill="#0E0E10"/>
+  <g fill="none" stroke="#C6FF3D" stroke-width="11" stroke-linecap="round" stroke-linejoin="round" transform="translate(60 60) scale(0.78) translate(-60 -60)">
+    <path d="M20 100 L20 20 L54 100 L54 20"/>
+    <path d="M66 20 L66 100 M66 60 L100 60 M100 20 L100 100"/>
+  </g>
+</svg>
diff --git a/src/app/layout.tsx b/src/app/layout.tsx
index 4404f92..b7ed578 100644
--- a/src/app/layout.tsx
+++ b/src/app/layout.tsx
@@ -1,10 +1,12 @@
 import Preloader from '@/components/intro/Preloader';
+import SiteFooter from '@/components/layout/SiteFooter';
 import SiteNav from '@/components/layout/SiteNav';
 import SmoothScroll from '@/components/motion/SmoothScroll';
+import JsonLd from '@/components/seo/JsonLd';
 import './globals.css';
 import type { Viewport } from 'next';
 import { clashDisplay, satoshi, geistMono } from '@/styles/fonts';
-import { buildPersonJsonLd, buildRootMetadata } from '@/lib/metadata';
+import { buildPersonJsonLd, buildRootMetadata, buildWebSiteJsonLd } from '@/lib/metadata';
 
 export const metadata = buildRootMetadata();
 
@@ -15,21 +17,17 @@ export const viewport: Viewport = {
 };
 
 export default function RootLayout({ children }: { children: React.ReactNode }) {
-  const jsonLd = buildPersonJsonLd();
-
   return (
     <html lang="en" className={`${clashDisplay.variable} ${satoshi.variable} ${geistMono.variable}`}>
       <body className="min-h-screen bg-dark text-ink font-sans">
         <Preloader />
-        <script
-          type="application/ld+json"
-          dangerouslySetInnerHTML={{ __html: JSON.stringify(jsonLd) }}
-          suppressHydrationWarning
-        />
+        <JsonLd data={buildPersonJsonLd()} />
+        <JsonLd data={buildWebSiteJsonLd()} />
         <div className="grain" aria-hidden />
         <SmoothScroll>
           <SiteNav />
           <main className="max-w-6xl mx-auto px-4 sm:px-6 py-8">{children}</main>
+          <SiteFooter />
         </SmoothScroll>
       </body>
     </html>
diff --git a/src/app/page.tsx b/src/app/page.tsx
index 061e473..428b832 100644
--- a/src/app/page.tsx
+++ b/src/app/page.tsx
@@ -1,4 +1,4 @@
-import { buildPageMetadata } from '@/lib/metadata';
+import { buildHomeMetadata } from '@/lib/metadata';
 import Reveal from '@/components/motion/Reveal';
 import MagneticButton from '@/components/motion/MagneticButton';
 import UnderlineLink from '@/components/motion/UnderlineLink';
@@ -7,11 +7,7 @@ import PlaygroundSection from '@/components/playground/PlaygroundSection';
 import work from '@/data/work';
 import { SITE_EMAIL, SITE_SOCIAL_LINKS } from '@/lib/site';
 
-export const metadata = buildPageMetadata({
-  title: 'Portfolio',
-  description: 'Discover the work, skills, and contact information of frontend developer Nima Hakimi.',
-  path: '/',
-});
+export const metadata = buildHomeMetadata();
 
 export default function Home() {
   return (
diff --git a/src/app/robots.ts b/src/app/robots.ts
new file mode 100644
index 0000000..c2ca105
--- /dev/null
+++ b/src/app/robots.ts
@@ -0,0 +1,9 @@
+import type { MetadataRoute } from 'next';
+import { SITE_URL } from '@/lib/site';
+
+export default function robots(): MetadataRoute.Robots {
+  return {
+    rules: { userAgent: '*', allow: '/' },
+    sitemap: `${SITE_URL}/sitemap.xml`,
+  };
+}
diff --git a/src/app/robots.txt/route.ts b/src/app/robots.txt/route.ts
deleted file mode 100644
index 56279ec..0000000
--- a/src/app/robots.txt/route.ts
+++ /dev/null
@@ -1,10 +0,0 @@
-const baseUrl = 'https://hackimi.dev';
-
-export function GET() {
-  return new Response(`User-agent: *\nAllow: /\nSitemap: ${baseUrl}/sitemap.xml\n`, {
-    headers: {
-      'Content-Type': 'text/plain; charset=utf-8',
-      'Cache-Control': 'public, max-age=86400',
-    },
-  });
-}
diff --git a/src/app/sitemap.ts b/src/app/sitemap.ts
index 42e8763..67feec8 100644
--- a/src/app/sitemap.ts
+++ b/src/app/sitemap.ts
@@ -1,17 +1,14 @@
 import type { MetadataRoute } from 'next';
 import work from '@/data/work';
-
-const baseUrl = 'https://hackimi.dev';
+import { SITE_URL } from '@/lib/site';
 
 const routes = ['/', '/about', '/work', '/contact', ...work.map((c) => `/work/${c.slug}`)] as const;
 
+/**
+ * URL list only, deliberately: the previous lastModified stamped `new Date()`
+ * on every build (a false freshness signal on all URLs), and Google ignores
+ * changeFrequency/priority. No signal beats a wrong signal.
+ */
 export default function sitemap(): MetadataRoute.Sitemap {
-  const lastModified = new Date();
-
-  return routes.map((route) => ({
-    url: `${baseUrl}${route === '/' ? '' : route}`,
-    lastModified,
-    changeFrequency: 'monthly',
-    priority: route === '/' ? 1 : 0.6,
-  }));
+  return routes.map((route) => ({ url: `${SITE_URL}${route === '/' ? '' : route}` }));
 }
diff --git a/src/app/work/[slug]/page.tsx b/src/app/work/[slug]/page.tsx
index 3c3383e..8cd6fee 100644
--- a/src/app/work/[slug]/page.tsx
+++ b/src/app/work/[slug]/page.tsx
@@ -1,10 +1,11 @@
 import { notFound } from 'next/navigation';
-import { buildPageMetadata } from '@/lib/metadata';
+import { buildBreadcrumbJsonLd, buildPageMetadata } from '@/lib/metadata';
 import Reveal from '@/components/motion/Reveal';
 import UnderlineLink from '@/components/motion/UnderlineLink';
 import CaseHero from '@/components/work/CaseHero';
 import CaseSection from '@/components/work/CaseSection';
 import { CaseCover, CaseGallery } from '@/components/work/CaseGallery';
+import JsonLd from '@/components/seo/JsonLd';
 import work from '@/data/work';
 
 export function generateStaticParams() {
@@ -14,7 +15,7 @@ export function generateStaticParams() {
 export async function generateMetadata({ params }: { params: Promise<{ slug: string }> }) {
   const { slug } = await params;
   const c = work.find((w) => w.slug === slug);
-  if (!c) return {};
+  if (!c) return { title: 'Not found' };
   return buildPageMetadata({
     title: c.title,
     description: c.summary,
@@ -33,6 +34,7 @@ export default async function CasePage({ params }: { params: Promise<{ slug: str
 
   return (
     <article className="pb-[clamp(3rem,10vh,7rem)]">
+      <JsonLd data={buildBreadcrumbJsonLd(c.title, c.slug)} />
       <div className="pt-[clamp(1.5rem,5vh,3rem)]">
         <UnderlineLink href="/work" className="font-mono text-xs uppercase tracking-[0.12em] text-muted">
           ← All work
diff --git a/src/components/layout/SiteFooter.tsx b/src/components/layout/SiteFooter.tsx
new file mode 100644
index 0000000..d8a0885
--- /dev/null
+++ b/src/components/layout/SiteFooter.tsx
@@ -0,0 +1,18 @@
+import { SITE_AUTHOR, SITE_ROLE, SITE_LOCATION } from '@/lib/site';
+
+/**
+ * Site-wide footer. Besides closing the page visually, this is the one
+ * crawlable spot outside the hero h1 where the full name appears as text
+ * on every route (the nav logo is an SVG with only an aria-label).
+ */
+export default function SiteFooter() {
+  return (
+    <footer className="mx-auto max-w-6xl px-4 sm:px-6">
+      <div className="mt-[clamp(3rem,8vh,5rem)] border-t border-white/10 py-8">
+        <p className="font-mono text-xs uppercase tracking-[0.12em] text-muted">
+          © {new Date().getFullYear()} {SITE_AUTHOR} — {SITE_ROLE}, {SITE_LOCATION.city}
+        </p>
+      </div>
+    </footer>
+  );
+}
diff --git a/src/components/seo/JsonLd.tsx b/src/components/seo/JsonLd.tsx
new file mode 100644
index 0000000..3e70715
--- /dev/null
+++ b/src/components/seo/JsonLd.tsx
@@ -0,0 +1,10 @@
+/** Renders one schema.org block. Server component — emits static JSON, no hydration. */
+export default function JsonLd({ data }: { data: object }) {
+  return (
+    <script
+      type="application/ld+json"
+      dangerouslySetInnerHTML={{ __html: JSON.stringify(data) }}
+      suppressHydrationWarning
+    />
+  );
+}
diff --git a/src/lib/metadata.ts b/src/lib/metadata.ts
index dc4725c..4297a25 100644
--- a/src/lib/metadata.ts
+++ b/src/lib/metadata.ts
@@ -4,9 +4,11 @@ import {
   SITE_DESCRIPTION,
   SITE_EMAIL,
   SITE_EMPLOYER,
+  SITE_GOOGLE_SITE_VERIFICATION,
   SITE_KEYWORDS,
   SITE_LOCATION,
   SITE_OG_IMAGE,
+  SITE_PERSON_IMAGE,
   SITE_ROLE,
   SITE_SOCIAL_LINKS,
   SITE_TITLE,
@@ -104,6 +106,7 @@ export function buildRootMetadata(): Metadata {
     alternates: {
       canonical: SITE_URL,
     },
+    ...(SITE_GOOGLE_SITE_VERIFICATION ? { verification: { google: SITE_GOOGLE_SITE_VERIFICATION } } : {}),
   };
 }
 
@@ -136,10 +139,41 @@ export function buildPageMetadata({ title, description, path, keywords, ogImage
   };
 }
 
+/**
+ * Home is the ONE page whose title must be absolute: Next's title.template
+ * only applies to child segments, so the root page.tsx never receives the
+ * "%s | Nima Hakimi" suffix — it must carry the full name-led string itself.
+ */
+export function buildHomeMetadata(): Metadata {
+  return {
+    title: SITE_TITLE,
+    description: SITE_DESCRIPTION,
+    keywords: defaultKeywords,
+    alternates: { canonical: SITE_URL },
+    openGraph: {
+      title: SITE_TITLE,
+      description: SITE_DESCRIPTION,
+      url: SITE_URL,
+      siteName: SITE_AUTHOR,
+      type: 'website',
+      images: [defaultImage],
+    },
+    twitter: {
+      card: 'summary_large_image',
+      title: SITE_TITLE,
+      description: SITE_DESCRIPTION,
+      images: [defaultTwitterImage],
+    },
+  };
+}
+
+export const PERSON_ID = `${SITE_URL}/#person` as const;
+
 export function buildPersonJsonLd() {
   return {
     '@context': 'https://schema.org',
     '@type': 'Person',
+    '@id': PERSON_ID,
     'name': SITE_AUTHOR,
     'url': SITE_URL,
     'jobTitle': SITE_ROLE,
@@ -153,8 +187,39 @@ export function buildPersonJsonLd() {
       'addressCountry': SITE_LOCATION.country,
     },
     'email': `mailto:${SITE_EMAIL}`,
-    'image': SITE_OG_IMAGE,
-    'sameAs': [SITE_SOCIAL_LINKS.github, SITE_SOCIAL_LINKS.linkedin],
+    'image': SITE_PERSON_IMAGE,
+    'sameAs': [SITE_SOCIAL_LINKS.github, SITE_SOCIAL_LINKS.linkedin, SITE_SOCIAL_LINKS.orcid],
+  } as const;
+}
+
+export function buildWebSiteJsonLd() {
+  return {
+    '@context': 'https://schema.org',
+    '@type': 'WebSite',
+    'name': SITE_AUTHOR,
+    'url': SITE_URL,
+    'publisher': { '@id': PERSON_ID },
+  } as const;
+}
+
+export function buildProfilePageJsonLd() {
+  return {
+    '@context': 'https://schema.org',
+    '@type': 'ProfilePage',
+    'url': `${SITE_URL}/about`,
+    'mainEntity': { '@id': PERSON_ID },
+  } as const;
+}
+
+export function buildBreadcrumbJsonLd(caseTitle: string, slug: string) {
+  return {
+    '@context': 'https://schema.org',
+    '@type': 'BreadcrumbList',
+    'itemListElement': [
+      { '@type': 'ListItem', 'position': 1, 'name': 'Home', 'item': SITE_URL },
+      { '@type': 'ListItem', 'position': 2, 'name': 'Work', 'item': `${SITE_URL}/work` },
+      { '@type': 'ListItem', 'position': 3, 'name': caseTitle, 'item': `${SITE_URL}/work/${slug}` },
+    ],
   } as const;
 }
 
diff --git a/src/lib/site.ts b/src/lib/site.ts
index 1a0ac11..0ef8b0f 100644
--- a/src/lib/site.ts
+++ b/src/lib/site.ts
@@ -1,10 +1,10 @@
 export const SITE_URL = 'https://hackimi.dev';
 export const SITE_AUTHOR = 'Nima Hakimi';
-export const SITE_ROLE = 'Frontend Developer';
-export const SITE_TITLE = `${SITE_AUTHOR} | ${SITE_ROLE}`;
+export const SITE_ROLE = 'Developer & AI Engineer';
+export const SITE_TITLE = `${SITE_AUTHOR} — ${SITE_ROLE}`;
 export const SITE_DESCRIPTION =
-  'Portfolio of Nima Hakimi, a frontend developer focused on accessible, high-performance web applications built with modern technologies.';
-export const SITE_KEYWORDS = ['Nima Hakimi', 'Frontend Developer', 'Next.js', 'React', 'Oslo'];
+  'Portfolio of Nima Hakimi, a developer and AI engineer building accessible, high-performance web applications.';
+export const SITE_KEYWORDS = ['Nima Hakimi', 'Developer', 'AI Engineer', 'Next.js', 'React', 'Oslo'];
 export const SITE_EMAIL = 'nima@hackimi.dev';
 export const SITE_LOCATION = {
   city: 'Oslo',
@@ -16,5 +16,9 @@ export const SITE_EMPLOYER = {
 export const SITE_SOCIAL_LINKS = {
   github: 'https://github.com/nimkimi',
   linkedin: 'https://linkedin.com/in/nima-hakimi-387716175',
+  orcid: 'https://orcid.org/0009-0002-6656-2498',
 };
-export const SITE_OG_IMAGE = `${SITE_URL}/bigSmile.JPEG` as const;
+export const SITE_OG_IMAGE = `${SITE_URL}/og.png` as const;
+export const SITE_PERSON_IMAGE = `${SITE_URL}/bigSmile.JPEG` as const;
+/** Google Search Console meta-tag token. Empty until the property is created. */
+export const SITE_GOOGLE_SITE_VERIFICATION = '';
diff --git a/tests/components/site-footer.test.tsx b/tests/components/site-footer.test.tsx
new file mode 100644
index 0000000..3332eec
--- /dev/null
+++ b/tests/components/site-footer.test.tsx
@@ -0,0 +1,14 @@
+import { render, screen } from '@testing-library/react';
+import { describe, expect, it } from 'vitest';
+
+import SiteFooter from '@/components/layout/SiteFooter';
+import { SITE_AUTHOR, SITE_ROLE, SITE_LOCATION } from '@/lib/site';
+
+describe('SiteFooter', () => {
+  it('renders the © name-role-city line as visible text', () => {
+    render(<SiteFooter />);
+    const footer = screen.getByRole('contentinfo');
+    const year = new Date().getFullYear();
+    expect(footer).toHaveTextContent(`© ${year} ${SITE_AUTHOR} — ${SITE_ROLE}, ${SITE_LOCATION.city}`);
+  });
+});
diff --git a/tests/integration/contact-page.test.tsx b/tests/integration/contact-page.test.tsx
index a042eb3..0cc8dcb 100644
--- a/tests/integration/contact-page.test.tsx
+++ b/tests/integration/contact-page.test.tsx
@@ -79,9 +79,13 @@ describe('Contact page (/contact)', () => {
     expect(contactClientSpy).toHaveBeenCalledWith(expect.objectContaining({ siteKey: 'public-key' }));
   });
 
-  it('exports static contact metadata', async () => {
+  it('exports contact metadata with its own OpenGraph object', async () => {
     const { metadata } = await import('@/app/contact/page');
+    const { SITE_AUTHOR } = await import('@/lib/site');
     expect(metadata.title).toBe('Contact');
     expect(metadata.alternates?.canonical).toBe('https://hackimi.dev/contact');
+    // Regression: an absent openGraph made /contact inherit the homepage OG.
+    expect(metadata.openGraph?.title).toBe(`Contact | ${SITE_AUTHOR}`);
+    expect(metadata.openGraph?.url).toBe('https://hackimi.dev/contact');
   });
 });
diff --git a/tests/integration/pages.test.tsx b/tests/integration/pages.test.tsx
index 0b63e0e..ae8b700 100644
--- a/tests/integration/pages.test.tsx
+++ b/tests/integration/pages.test.tsx
@@ -55,9 +55,10 @@ describe('Home page (/)', () => {
     }
   });
 
-  it('exports page metadata derived from buildPageMetadata', async () => {
+  it('exports absolute name-led home metadata', async () => {
     const { metadata } = await import('@/app/page');
-    expect(metadata.title).toBe('Portfolio');
+    const { SITE_TITLE } = await import('@/lib/site');
+    expect(metadata.title).toBe(SITE_TITLE);
     expect(metadata.description).toContain('Nima Hakimi');
   });
 });
@@ -188,12 +189,12 @@ describe('Work case page (/work/[slug])', () => {
     expect(meta.alternates?.canonical).toContain(`/work/${c.slug}`);
   });
 
-  it('generateMetadata returns an empty object for an unknown slug', async () => {
+  it('generateMetadata returns a not-found title for an unknown slug', async () => {
     const { generateMetadata } = await import('@/app/work/[slug]/page');
 
     const meta = await generateMetadata({
       params: Promise.resolve({ slug: 'does-not-exist' }),
     });
-    expect(meta).toEqual({});
+    expect(meta).toEqual({ title: 'Not found' });
   });
 });
diff --git a/tests/server/robots.test.ts b/tests/server/robots.test.ts
index 287cbe1..6fb1e30 100644
--- a/tests/server/robots.test.ts
+++ b/tests/server/robots.test.ts
@@ -1,22 +1,13 @@
 import { describe, expect, it } from 'vitest';
 
-import { GET } from '@/app/robots.txt/route';
-
-describe('robots.txt GET', () => {
-  it('returns a Response', () => {
-    expect(GET()).toBeInstanceOf(Response);
-  });
-
-  it('serves the expected robots directives in the body', async () => {
-    const body = await GET().text();
-    expect(body).toContain('User-agent: *');
-    expect(body).toContain('Allow: /');
-    expect(body).toContain('Sitemap: https://hackimi.dev/sitemap.xml');
-  });
-
-  it('sets the text/plain content type and a cache header', () => {
-    const res = GET();
-    expect(res.headers.get('Content-Type')).toBe('text/plain; charset=utf-8');
-    expect(res.headers.get('Cache-Control')).toBe('public, max-age=86400');
+import robots from '@/app/robots';
+import { SITE_URL } from '@/lib/site';
+
+describe('robots', () => {
+  it('allows all crawlers everything and points at the sitemap', () => {
+    expect(robots()).toEqual({
+      rules: { userAgent: '*', allow: '/' },
+      sitemap: `${SITE_URL}/sitemap.xml`,
+    });
   });
 });
diff --git a/tests/server/sitemap.test.ts b/tests/server/sitemap.test.ts
index 2045f68..b4f18b7 100644
--- a/tests/server/sitemap.test.ts
+++ b/tests/server/sitemap.test.ts
@@ -12,18 +12,16 @@ describe('sitemap', () => {
     expect(entries).toHaveLength(STATIC_ROUTES.length + work.length);
   });
 
-  it("uses the bare base URL (no trailing slash) and priority 1 for '/'", () => {
+  it("uses the bare base URL (no trailing slash) for '/'", () => {
     const root = sitemap().find((e) => e.url === BASE_URL);
     expect(root).toBeDefined();
     expect(root?.url).toBe('https://hackimi.dev');
-    expect(root?.priority).toBe(1);
   });
 
-  it('uses the full path and priority 0.6 for a non-root route', () => {
+  it('uses the full path for a non-root route', () => {
     const about = sitemap().find((e) => e.url === `${BASE_URL}/about`);
     expect(about).toBeDefined();
     expect(about?.url).toBe('https://hackimi.dev/about');
-    expect(about?.priority).toBe(0.6);
   });
 
   it('includes every work slug as /work/<slug>', () => {
@@ -33,10 +31,9 @@ describe('sitemap', () => {
     }
   });
 
-  it("gives every entry changeFrequency 'monthly' and a lastModified Date", () => {
+  it('entries are url-only — no fabricated lastModified/changeFrequency/priority', () => {
     for (const entry of sitemap()) {
-      expect(entry.changeFrequency).toBe('monthly');
-      expect(entry.lastModified).toBeInstanceOf(Date);
+      expect(Object.keys(entry)).toEqual(['url']);
     }
   });
 });
diff --git a/tests/unit/metadata.test.ts b/tests/unit/metadata.test.ts
index 8fac988..3d02146 100644
--- a/tests/unit/metadata.test.ts
+++ b/tests/unit/metadata.test.ts
@@ -1,5 +1,15 @@
 import { describe, it, expect } from 'vitest';
-import { buildRootMetadata, buildPageMetadata, buildPersonJsonLd, resolveUrl, METADATA_BASE } from '@/lib/metadata';
+import {
+  buildRootMetadata,
+  buildPageMetadata,
+  buildHomeMetadata,
+  buildPersonJsonLd,
+  buildWebSiteJsonLd,
+  buildProfilePageJsonLd,
+  buildBreadcrumbJsonLd,
+  resolveUrl,
+  METADATA_BASE,
+} from '@/lib/metadata';
 import {
   SITE_URL,
   SITE_AUTHOR,
@@ -9,7 +19,9 @@ import {
   SITE_EMAIL,
   SITE_EMPLOYER,
   SITE_OG_IMAGE,
+  SITE_PERSON_IMAGE,
   SITE_SOCIAL_LINKS,
+  SITE_GOOGLE_SITE_VERIFICATION,
 } from '@/lib/site';
 
 describe('resolveUrl', () => {
@@ -111,6 +123,21 @@ describe('buildPageMetadata', () => {
   });
 });
 
+describe('buildHomeMetadata', () => {
+  it('title is the absolute SITE_TITLE (root segment gets no template)', () => {
+    const meta = buildHomeMetadata();
+    expect(meta.title).toBe(SITE_TITLE);
+    expect(meta.openGraph?.title).toBe(SITE_TITLE);
+    expect(meta.twitter?.title).toBe(SITE_TITLE);
+  });
+
+  it('canonical and og.url are the bare SITE_URL', () => {
+    const meta = buildHomeMetadata();
+    expect(meta.alternates?.canonical).toBe(SITE_URL);
+    expect(meta.openGraph?.url).toBe(SITE_URL);
+  });
+});
+
 describe('buildRootMetadata', () => {
   it('metadataBase is a URL of SITE_URL', () => {
     const meta = buildRootMetadata();
@@ -131,6 +158,15 @@ describe('buildRootMetadata', () => {
     expect(meta.openGraph).toBeDefined();
     expect(meta.twitter).toBeDefined();
   });
+
+  it('emits the verification block only when the GSC token is set', () => {
+    // SITE_GOOGLE_SITE_VERIFICATION is '' until Nima creates the property;
+    // an empty <meta name="google-site-verification"> must never ship. The
+    // assertion is conditional so the future token commit stays green too.
+    expect(buildRootMetadata().verification).toEqual(
+      SITE_GOOGLE_SITE_VERIFICATION ? { google: SITE_GOOGLE_SITE_VERIFICATION } : undefined
+    );
+  });
 });
 
 describe('buildPersonJsonLd', () => {
@@ -143,6 +179,43 @@ describe('buildPersonJsonLd', () => {
     expect(ld.jobTitle).toBe(SITE_ROLE);
     expect(ld.worksFor.name).toBe(SITE_EMPLOYER.name);
     expect(ld.email).toBe(`mailto:${SITE_EMAIL}`);
-    expect(ld.sameAs).toEqual([SITE_SOCIAL_LINKS.github, SITE_SOCIAL_LINKS.linkedin]);
+    expect(ld.sameAs).toEqual([SITE_SOCIAL_LINKS.github, SITE_SOCIAL_LINKS.linkedin, SITE_SOCIAL_LINKS.orcid]);
+  });
+
+  it('carries a stable @id and all three sameAs anchors', () => {
+    const ld = buildPersonJsonLd();
+    expect(ld['@id']).toBe(`${SITE_URL}/#person`);
+    expect(ld.sameAs).toEqual([SITE_SOCIAL_LINKS.github, SITE_SOCIAL_LINKS.linkedin, SITE_SOCIAL_LINKS.orcid]);
+  });
+
+  it('uses the person photo, not the OG card, as the Person image', () => {
+    expect(buildPersonJsonLd().image).toBe(SITE_PERSON_IMAGE);
+  });
+});
+
+describe('entity graph builders', () => {
+  it('WebSite names the site after the author and references the person', () => {
+    const ld = buildWebSiteJsonLd();
+    expect(ld['@type']).toBe('WebSite');
+    expect(ld.name).toBe(SITE_AUTHOR);
+    expect(ld.url).toBe(SITE_URL);
+    expect(ld.publisher).toEqual({ '@id': `${SITE_URL}/#person` });
+  });
+
+  it('ProfilePage points its mainEntity at the person @id', () => {
+    const ld = buildProfilePageJsonLd();
+    expect(ld['@type']).toBe('ProfilePage');
+    expect(ld.url).toBe(`${SITE_URL}/about`);
+    expect(ld.mainEntity).toEqual({ '@id': `${SITE_URL}/#person` });
+  });
+
+  it('BreadcrumbList walks Home → Work → case', () => {
+    const ld = buildBreadcrumbJsonLd('Be My Guide', 'be-my-guide');
+    expect(ld['@type']).toBe('BreadcrumbList');
+    expect(ld.itemListElement).toEqual([
+      { '@type': 'ListItem', 'position': 1, 'name': 'Home', 'item': SITE_URL },
+      { '@type': 'ListItem', 'position': 2, 'name': 'Work', 'item': `${SITE_URL}/work` },
+      { '@type': 'ListItem', 'position': 3, 'name': 'Be My Guide', 'item': `${SITE_URL}/work/be-my-guide` },
+    ]);
   });
 });
diff --git a/tests/unit/site.test.ts b/tests/unit/site.test.ts
index 94c55a8..d0a6eef 100644
--- a/tests/unit/site.test.ts
+++ b/tests/unit/site.test.ts
@@ -15,14 +15,21 @@ describe('site constants', () => {
     expect(url.protocol).toBe('https:');
   });
 
-  it('SITE_TITLE is `${SITE_AUTHOR} | ${SITE_ROLE}`', () => {
-    expect(SITE_TITLE).toBe(`${SITE_AUTHOR} | ${SITE_ROLE}`);
+  it('SITE_TITLE is `${SITE_AUTHOR} — ${SITE_ROLE}` (em-dash form)', () => {
+    expect(SITE_TITLE).toBe(`${SITE_AUTHOR} — ${SITE_ROLE}`);
+    expect(SITE_ROLE).toBe('Developer & AI Engineer');
   });
 
   it('SITE_OG_IMAGE starts with SITE_URL', () => {
     expect(SITE_OG_IMAGE.startsWith(SITE_URL)).toBe(true);
   });
 
+  it('OG image is the branded card; person image is the photo', async () => {
+    const { SITE_OG_IMAGE, SITE_PERSON_IMAGE, SITE_URL } = await import('@/lib/site');
+    expect(SITE_OG_IMAGE).toBe(`${SITE_URL}/og.png`);
+    expect(SITE_PERSON_IMAGE).toBe(`${SITE_URL}/bigSmile.JPEG`);
+  });
+
   it('social links are valid URLs', () => {
     Object.values(SITE_SOCIAL_LINKS).forEach((link) => {
       expect(() => new URL(link)).not.toThrow();
@@ -30,7 +37,16 @@ describe('site constants', () => {
     });
   });
 
+  it('social links include GitHub, LinkedIn and ORCID', () => {
+    expect(Object.keys(SITE_SOCIAL_LINKS).sort()).toEqual(['github', 'linkedin', 'orcid']);
+  });
+
   it('SITE_KEYWORDS is non-empty', () => {
     expect(SITE_KEYWORDS.length).toBeGreaterThan(0);
   });
+
+  it('keywords carry the new positioning, not "Frontend"', () => {
+    expect(SITE_KEYWORDS).toContain('AI Engineer');
+    expect(SITE_KEYWORDS.join(' ')).not.toMatch(/frontend/i);
+  });
 });