Site Health Audit: SEO, TLS, llms.txt and Broken Links with a 0-100 Score

Paste a list of URLs. Get a score from 0 to 100 for each, with every lost point explained.

For each URL, this Actor checks the things that silently break: expiring certificates, redirect chains, missing tags, invalid structured data, a blocked robots.txt, dead links. It reports each one in plain English, ready for a client report or an automated alert.

No login, no browser, no proxies. It uses plain HTTP, so it is fast and cheap.

Who it is for

• SEO consultants and agencies who need a quick technical check across many client pages
• Developers and DevOps who want a scheduled check for certificate expiry and broken pages
• Site owners preparing for AI search, who want to see llms.txt, robots rules and structured data in one place

What it checks on every URL

Area	Details
Availability	Final status, full redirect chain with each hop's status, loop detection, response time
TLS certificate	Valid or not, the reason if not, expiry date, days left, issuer, protocol
SEO tags	Title, meta description, canonical, meta robots, hreflang, H1 count, language, viewport
Sharing	Open Graph tags
Structured data	JSON-LD types found, and any blocks with invalid JSON
Crawl setup	robots.txt rules, sitemap presence, whether robots.txt blocks all bots
AI readiness	llms.txt presence, ignoring sites that answer every path with an HTML page
Links	Broken links on that page, up to your cap, with anchor text
Score	0-100 with a grade and a list of reasons, each with its penalty

How to use it

1. Paste your URLs.
2. Choose how many links per page to test. Set it to 0 for a much faster run.
3. Run it, or put it on a schedule to catch problems such as an expiring certificate early.

{
  "urls": [{ "url": "https://www.example.com/" }, { "url": "https://www.example.com/pricing" }],
  "maxLinksPerPage": 50,
  "maxPages": 100
}

Output

One row per URL. Shortened sample from a real run:

{
  "inputUrl": "https://www.iana.org/",
  "finalUrl": "https://www.iana.org/",
  "finalStatus": 200,
  "redirectChain": [],
  "responseTimeMs": 133,
  "tls": { "checked": true, "valid": true, "protocol": "TLSv1.3", "daysUntilExpiry": 79 },
  "page": {
    "title": "Internet Assigned Numbers Authority",
    "metaDescription": null,
    "canonical": "https://iana.org/",
    "hreflang": [],
    "openGraph": {},
    "jsonLd": { "types": [], "blocks": 0, "invalidBlocks": 0 }
  },
  "robotsTxt": { "status": 200, "present": true, "sitemaps": [], "blocksAllBots": false },
  "sitemap": { "present": false },
  "llmsTxt": { "present": false, "status": 404 },
  "links": { "found": 23, "checked": 23, "ok": 20, "redirected": 0, "brokenCount": 0, "broken": [], "unverified": [{ "url": "https://www.icann.org/", "anchor": "ICANN", "status": null, "error": "The operation was aborted" }] },
  "score": 82,
  "grade": "B",
  "reasons": [
    { "check": "meta", "penalty": 6, "severity": "warning", "message": "The meta description is missing." },
    { "check": "open-graph", "penalty": 3, "severity": "info", "message": "Open Graph title or image is missing, so shared links look bare." },
    { "check": "sitemap", "penalty": 3, "severity": "warning", "message": "No sitemap found in robots.txt or at /sitemap.xml." },
    { "check": "canonical", "penalty": 2, "severity": "info", "message": "The canonical points elsewhere: https://iana.org/" },
    { "check": "structured-data", "penalty": 2, "severity": "info", "message": "No JSON-LD structured data." },
    { "check": "language", "penalty": 1, "severity": "info", "message": "The html element has no lang attribute." },
    { "check": "llms", "penalty": 1, "severity": "info", "message": "No llms.txt (an optional file that guides AI tools)." }
  ]
}

How the score works

Every page starts at 100 and loses points for specific problems, so you can always see why. Grades: A is 90+, B is 75+, C is 60+, D is 40+, F is below that.

Problem	Points lost
Unreachable site	Score is 0
HTTP 4xx or 5xx	60
Redirect loop	30
Invalid TLS certificate	25
Not served over HTTPS	20
Certificate expires in under 7 days / under 21 days	15 / 5
Broken links	5 for the first, then 2 more each, up to 15
Missing title	10
Noindex	10
Slow response (over 3 s / over 1.5 s)	8 / 3
Missing meta description	6
Several redirect hops	5
Missing canonical / invalid JSON-LD	4 each
Missing sitemap, H1, viewport or Open Graph	3 each
Smaller issues, such as title length, no llms.txt, no lang	1 or 2

The score is a checklist of technical hygiene. It is not a prediction of search ranking.

Honest link results

A link counts as broken only on clear evidence: HTTP 404, 410, a 5xx other than 503, a domain that does not exist, or a refused connection. Sites often answer automated checks with 400, 403, 429 or 503 even when the page works for people, so those are listed as unverified and do not cost points. Links whose site's robots.txt disallows us are skipped and counted.

Pricing

Pay per event.

Event	Price	When
Run started	$0.002	Once per run, per GB of memory
URL audited	$0.008	One URL audited and scored, including a dead site, which is a real finding

Invalid URLs, private addresses and URLs blocked by robots.txt are free.

Examples: 10 URLs cost $0.082. 100 URLs cost $0.80. Link checks are included in the price, up to 500 links per page. Set a maximum cost per run and the Actor stops cleanly at that limit.

Limits

• Only the URLs you supply are audited. Hard cap of 1,000 URLs per run.
• Link checking is one hop deep, in page order, and stops after 120 seconds per page. The output says how many links were left unchecked.
• Response time is one measurement from Apify's servers, so it is a rough signal and can vary between runs. It is not a Core Web Vitals test.
• HTML is read as served. Tags or links added later by JavaScript are not seen.
• Pages over 3 MB are read up to that point, and bodyTruncated is set.

What it does not do

• It does not crawl your site or discover pages. If you want that, use a crawler.
• It does not run a browser, so there are no Lighthouse or Core Web Vitals scores.
• It does not report backlinks, keywords, rankings or domain authority.
• It honours robots.txt. If a site disallows this Actor's fetching (LuanLabSiteHealthAudit), or its robots.txt returns a server error, that URL is reported as not audited and is not charged. To audit such a page, allow this user agent on your own site.
• It does not bypass firewalls, bot protection or CAPTCHAs. A page behind one is scored on what it returned.
• It does not collect e-mail addresses or other contact details.
• It refuses private and internal addresses.

Support

Think a score is wrong? Open an issue on the Issues tab with the URL and the reason you disagree.