Run the full technical SEO checklist on a URL: rendering (server-rendered vs JS shell), HTTPS/SSL + redirect, page essentials (title, meta description, h1, canonical, Open Graph, mobile viewport, image lazy-loading/WebP/dimensions), caching headers, structured data (JSON-LD), sitemap.xml, robots.txt, and AI-crawler access. Returns a Markdown report with ✅/⚠️/❌ per check (or structured JSON with output: 'json', or a shareable HTML file with output: 'html'). Set compare_to_baseline: true to diff against the previous run. Does not include Lighthouse (use run_lighthouse separately — it is slower).

Crawl a site (seeded from sitemap.xml, sitemap-index aware; falls back to following internal links) and audit up to max_pages pages with a polite delay. Each page is fetched once and shared across checks. Aggregates: site-wide issues (same problem on >50% of pages — fix once in the template), broken internal links (404s), multi-hop redirect chains, duplicate titles/meta descriptions, orphan sitemap URLs, and remaining per-page issues.

Fetch a page without executing JavaScript and detect whether it serves real server-rendered HTML or an empty client-side JS shell (hard for Google/AI to interpret).

Verify the site serves https:// with a valid certificate and that http:// permanently (301) redirects to https://.

Check a page's title, meta description, single h1, canonical tag, Open Graph tags, responsive viewport meta (mobile-first), image lazy-loading, WebP/AVIF usage, and explicit image width/height (CLS risk).

Inspect Cache-Control, content-encoding (gzip/brotli), ETag/Last-Modified headers and HTML document size for easy speed wins.

Find the site's sitemap the way search engines do (robots.txt Sitemap: lines, then /sitemap.xml, then sitemap-index.xml / sitemap_index.xml / wp-sitemap.xml), count entries, flag admin/cart/checkout/internal-search URLs, sample up to 10 URLs (from the first child sitemap for an index) to verify they return 200 without redirects, and check lastmod presence/freshness.

Fetch /robots.txt, verify it exists and is valid (not an HTML page or redirect), references the sitemap, and does not block the entire site.

Audit a page's hreflang annotations for multi-language or multi-country targeting (e.g. US + Canada): validates BCP-47 values (catching en_US, en-UK and other common mistakes), requires a self-referencing annotation, checks for x-default, flags duplicate/conflicting values and relative hrefs, and verifies agrees. With verify_return_links: true it fetches each alternate and confirms it links back — missing return links are the most common reason Google ignores an otherwise-correct hreflang cluster. A single-language site correctly reports that hreflang is not needed.

Extract and validate schema.org JSON-LD on a page (tolerates arrays and @graph). Checks Organization (incl. sameAs links), SoftwareApplication (name/description/applicationCategory/offers), FAQPage (Question/acceptedAnswer structure), and BreadcrumbList. Flags malformed JSON-LD blocks and provides a ready-to-paste Organization example when none exists.

Report the effective robots.txt allow/block status for GPTBot, ClaudeBot, Google-Extended, PerplexityBot, CCBot, and the * default (flagging blanket Disallow: / without per-bot overrides). Checks for /llms.txt, and flags noindex/nofollow in both the X-Robots-Tag response header and — critical failures invisible to a plain robots.txt check.

Analyze one or more URLs against clean-URL best practices: ≤ ~60 chars, lowercase, hyphens not underscores, no accents/special chars, no messy query params, shallow hierarchy. Offline — no requests are made.

Analyze where target keywords appear on a page: (and whether it's front-loaded), URL path, meta description, H1, H2 subheadings, first ~100 words, image alt text, and keyword density (flags both thin usage and stuffing). Matching is plural/stem-tolerant ('freight broker' matches 'freight brokers'). Pass keywords as an array (one report section per keyword); a single keyword string is also accepted. Returns ✅/⚠️/❌ per placement.

Run Google Lighthouse (performance + SEO categories) on a URL via npx. Reports scores and Core Web Vitals — LCP (≤ 2.5s), CLS (≤ 0.1), and INP (≤ 200ms; falls back to TBT when the lab run can't measure interactions) — plus top speed opportunities. Slow (~1-2 min); requires Chrome installed.