Skip to content

SEO.md

Latest commit

 

History

History
88 lines (70 loc) · 4.33 KB

SEO.md

File metadata and controls

88 lines (70 loc) · 4.33 KB
Raw

SEO audit and publishing

Improvement campaign (26 October 2026)

The implemented campaign, local measurements, limitations and four-week Search Console follow-up are recorded in the campaign report.

The build now checks all exported HTML, canonical URLs, reciprocal language alternates, unique titles/descriptions, H1, social metadata, JSON-LD and sitemap coverage. Search utility pages and aliases are deliberately excluded from the sitemap. Six French foundation records provide a complete introductory route. The five tool pages link to relevant explanations and those explanations link back to investigation tools. Existing public routes are preserved.

npm run verify:content
npm test
npm run build
npm run check:seo -- --report /tmp/devutils-seo-report.json
# With the production build served on port 5179 and Chrome available:
CHROME_PATH=/path/to/chrome npm run check:seo:browser
# After publishing the changes:
npm run check:seo:live

CHECK_BASE_URL and CHECK_OUTPUT configure the browser checks. SITE_URL configures the expected origin for build and live checks. Live verification is read-only and is intended to run after deployment. A local pass is not evidence that the production site has already received the changes.

Audit (24 October 2026)

GitHub Pages API confirms https://devutils-org.com/ as the production domain. The existing site already provides prerendered HTML, unique bilingual page metadata, semantic headings, internal navigation and social images. Its default canonical and Open Graph origin was still github.io. It had no robots.txt or sitemap. The static ATFQ book was outside the page registry and had no canonical or description metadata.

Implementation

src/site.ts is the single deployment origin, optionally overridden with SITE_URL=https://example.org npm run build. Only origins are accepted because the site uses root-relative navigation. DNS and hosting settings stay as configured.

src/shell.ts emits canonical and bilingual hreflang links (including x-default), Open Graph metadata, and Organization, WebPage and breadcrumb JSON-LD. No review scores, pricing, article dates or product capabilities are invented.

After prerendering, scripts/seo.ts verifies registered pages and social images, enriches exported ATFQ book pages, and generates dist/sitemap.xml and dist/robots.txt. Existing .html book URLs remain unchanged. New registered pages and exported book chapters enter the sitemap automatically. No speculative lastmod dates or priority values are emitted.

Run npm test and npm run build. Publish dist/ through the existing Pages workflow. Check production /robots.txt, /sitemap.xml, and canonical tags after deployment; production is not changed by building this isolated checkout.

Search Console (account access required)

  1. Open https://search.google.com/search-console and add the domain property devutils-org.com (or use an existing verified property).
  2. Complete Google's ownership verification if needed; keep its verification record. The website DNS configuration itself does not prove Search Console ownership.
  3. After deployment, submit https://devutils-org.com/sitemap.xml in Sitemaps.
  4. Inspect the homepage, a tool page, its French translation and a book chapter. Use the live test to check fetchability and request indexing where appropriate.
  5. Monitor indexing exclusions, canonical selection and performance. Submission helps discovery but does not guarantee indexing or a ranking.

References:

Content and performance follow-up

Preserve the current static rendering and lightweight dependency set. Check Core Web Vitals on production with real traffic; local builds do not prove performance scores. Continue linking tools to their actual documentation and reports.

The broader writing brief describes a future blog and reusable authoring layouts. There are no genuine blog articles in the current site, so this SEO change creates neither fabricated articles nor an empty RSS feed. When real articles are added, register their canonical routes, include author/date metadata from their source, use Article/TechArticle data and generate RSS from that same content registry.