Land swept WIP: concept page titles and canonicals, SEO roadmap (#127)

Recovers wip/sweep-20260926. Concept pages set their own title and point
canonical at the static curriculum page, restoring the shell title and
canonical when the route changes. Adds the queued SEO roadmap (the repo stays
maintenance-only). The swept JSON-LD edit to index.html is not included: that
block is generated from the agent-surfaces registry.

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>

Author: sarthakagrawal927

docs/seo-sprint.md

@@ -0,0 +1,41 @@
+# swe-interview-prep SEO sprint roadmap
+
+> Note: this repo is **maintenance-only since 2026-07-10** (`STATUS.md`). These
+> phases are queued intent, not active work — resume only if the owner lifts
+> maintenance mode; until then the indexing loop below still runs passively.
+
+Satellite playbook (see `saas-maker/tooling/skills/seo-sprint`). Baseline
+(28d ending 2026-09-18): **8,274 imp · 2 clicks · pos 24 · 23 indexed / 300
+pending** — big impression volume on junk queries; real queries rank but get
+zero clicks.
+
+## Evidence
+
+- **Junk signal**: most top impressions are `"indocad" …` copied-content
+  searches at pos 2–3 — don't optimize for these; they're people hunting a
+  leaked package, not interview prep.
+- **Real signal**: `"explain_estimated_rows"` pos 7 — Postgres EXPLAIN
+  internals. Curriculum content is surfacing for genuine technical questions.
+
+## Phases
+
+- [ ] **Phase 1 — topic landing pages for real interview queries.** From the
+  curriculum: system design, Postgres/database internals, behavioral, coding
+  patterns. Each page = the actual curriculum section + practice questions,
+  not a stub. Target the query classes where pos is already 2–8.
+- [ ] **Phase 2 — question pages.** One page per curriculum question with a
+  real worked answer — the "<question> interview" long tail. Buildable from
+  `shared/data`.
+- [ ] **Phase 3 — company-tagged pages** ("<company> interview questions")
+  only where the curriculum genuinely covers that company's known pattern —
+  don't fabricate company specificity.
+- [ ] **Phase 4 — indexing + CTR.** 300 pending churn via the daily agent;
+  pages at pos 2–8 with 0 clicks need title/snippet work (the scoreboard
+  flags them as `striking-distance`).
+
+## Rules
+
+- Never write an answer the curriculum doesn't actually teach — interview
+  searchers spot fake prep instantly.
+- Register each phase: `seo-scoreboard.mjs register --project
+  swe-interview-prep --lane editorial|programmatic --summary "…"`.

src/pages/ConceptDetail.tsx

@@ -7,7 +7,7 @@ import {
   Lock,
   Sparkles,
 } from 'lucide-react';
-import { useState } from 'react';
+import { useEffect, useState } from 'react';
 import { Link, useParams } from 'react-router-dom';
 
 import ConceptLibrary from '../components/ConceptLibrary';
@@ -52,13 +52,38 @@ const RATINGS: { id: 'again' | 'hard' | 'good' | 'easy'; label: string; tone: st
   { id: 'easy', label: 'Easy', tone: 'emerald' },
 ];
 
+const CANONICAL_ORIGIN = 'https://learn.significanthobbies.com';
+
 export default function ConceptDetail() {
   const { id } = useParams();
   const concept = id ? CONCEPT_BY_ID[id] : undefined;
   const { mastery, review } = useConceptMastery();
   const { gateStatus } = useGates();
   const [copied, setCopied] = useState(false);
 
+  // The SPA shell serves a generic <title> and a site-root canonical. Give each
+  // concept its own title and point canonical at the static curriculum page so
+  // the duplicate /concepts/ and /curriculum/concepts/ URLs consolidate
+  // ranking signals. Restore both when the route changes.
+  useEffect(() => {
+    if (!concept) return;
+    const previousTitle = document.title;
+    let canonical = document.head.querySelector<HTMLLinkElement>('link[rel="canonical"]');
+    const previousHref = canonical?.getAttribute('href') ?? null;
+    if (!canonical) {
+      canonical = document.createElement('link');
+      canonical.rel = 'canonical';
+      document.head.appendChild(canonical);
+    }
+    document.title = `${concept.name} · SWE Prep`;
+    canonical.href = `${CANONICAL_ORIGIN}/curriculum/concepts/${concept.id}`;
+    return () => {
+      document.title = previousTitle;
+      if (previousHref === null) canonical.remove();
+      else canonical.setAttribute('href', previousHref);
+    };
+  }, [concept]);
+
   if (!concept) {
     return (
       <PageShell>