Improve SEO: rich bill metadata, sitemap, robots, JSON-LD

[mikaalnaik]

Context

The bills app had solid baseline metadata (per-page OG/Twitter tags, dynamic OG images) but several gaps limited search visibility: bill pages had generic titles/descriptions, there was no sitemap, and the static robots.txt pointed at the wrong app's sitemap. This PR is a full SEO pass.

Changes

Content metadata

• Bill pages now use descriptive titles (C-11: Military Justice System Modernization Act · Builder MP) and meta descriptions derived from the real AI summary (markdown-stripped, truncated to ~155 chars), plus keywords from genres.
• Not-found bills return robots: noindex,nofollow.
• Memoized the bill resolver with React cache() (src/server/get-unified-bill.ts) so generateMetadata and the page component share a single fetch.

Discoverability

• New dynamic src/app/sitemap.ts lists home + every bill with lastModified, using absolute canonical URLs. Wrapped in try/catch so a data outage can't 500 it.
• Replaced static robots.txt (which referenced the wrong app's sitemap) with src/app/robots.ts pointing at the app's own sitemap and disallowing non-public routes.
• Extracted shared bill-merge logic into src/server/get-merged-bills.ts (used by home + sitemap).

Structured data

• New src/components/JsonLd.tsx helper. Bill pages emit schema.org Legislation + BreadcrumbList; home emits WebSite.

Hygiene

• Added src/utils/seo.ts (getCanonicalOrigin, stripMarkdown, truncate) for a single deterministic canonical origin (https://www.buildcanada.com); set metadataBase fallback.
• Removed redundant hand-rolled twitter:* meta tags across layout/home/bill pages.

Verification

• tsc --noEmit clean; biome check clean; unit tests pass (pre-commit hook).
• Ran the dev server and confirmed:
  • /bills/robots.txt and /bills/sitemap.xml render with canonical https://www.buildcanada.com/bills/... URLs.
  • Bill page: rich title, truncated description, keywords, correct canonical, exactly one twitter:card, 2 JSON-LD blocks.
  • Non-existent bill → <meta name="robots" content="noindex">.
  • Home page → WebSite JSON-LD.

Deployment note

Because the app runs under the /bills basePath, its robots.txt/sitemap.xml are served at /bills/.... Crawlers honor robots.txt only at the host root, so the root buildcanada.com/robots.txt should reference https://www.buildcanada.com/bills/sitemap.xml (or submit the sitemap directly in Search Console).

🤖 Generated with Claude Code