AI Search Citation Readiness Checker (GEO / AEO audit)

Check how ready your pages are to be found, understood and cited by AI answer engines: ChatGPT search, Perplexity, Google AI Overviews / Gemini, Copilot and Claude. Give it a list of URLs (your own site, a client's site or any public page). For each page you get a 0-100 citation readiness score, an A-D grade and a prioritised fix list.

What it checks

Area	Checks
AI crawler access	robots.txt verdict per bot for this exact URL: GPTBot, OAI-SearchBot, ChatGPT-User, Google-Extended, PerplexityBot, ClaudeBot, CCBot, Applebot-Extended, Bingbot (RFC 9309 wildcard matching)
llms.txt	present at /llms.txt, starts with # Name, has a > summary, ## sections, Markdown links
Structured data	schema.org types from JSON-LD and microdata, invalid JSON-LD, Organization/Person, Article, FAQPage, HowTo, BreadcrumbList
Trust signals	author (meta, rel=author, schema), datePublished / dateModified, <time> tags
Indexability	noindex (meta + X-Robots-Tag), canonical (self / other / missing), HTTPS, lang
Chunkability	H1, words per heading, question-style headings, paragraph length, lists and tables, thin content
Speed proxies	HTML response time, HTML size

Input

• startUrls: pages to audit (up to 500).
• maxPagesPerSite: 1 = only the listed pages; higher = follow internal links on the same host.

Output

One dataset item per page (citationScore, grade, aiCrawlerAccess, aiCrawlersBlocked, llmsTxt, schemaTypes, authorSignal, datePublished, chunk metrics, issues[] with severity and a concrete fix). Views: Readiness and Fixes. Failed URLs and reasons are in the key-value store record ERRORS, counts in SUMMARY.

Example input

{
  "startUrls": [
    "https://www.iana.org/help/example-domains"
  ],
  "maxPagesPerSite": 1
}

Example output (one record, some fields shortened)

Real record from a test run on the input above (dates and values change over time).

{
  "url": "https://www.iana.org/help/example-domains",
  "httpStatus": 200,
  "title": "Example Domains",
  "wordCount": 185,
  "aiCrawlerAccess": {
    "GPTBot": true,
    "OAI-SearchBot": true,
    "ChatGPT-User": true,
    "Google-Extended": true,
    "PerplexityBot": true,
    "ClaudeBot": true,
    "CCBot": true,
    "Applebot-Extended": true,
    "Bingbot": true
  },
  "aiCrawlersBlocked": [],
  "llmsTxt": {
    "present": false,
    "hasH1": false,
    "hasSummary": false,
    "links": 0,
    "sections": 0
  },
  "schemaTypes": [],
  "hasFaqSchema": false,
  "authorSignal": false,
  "issues": [
    {
      "code": "llms-txt-missing",
      "severity": "medium",
      "fix": "No /llms.txt. Add a Markdown file: '# Site name', '> one-line summary', then '## Section' lists of '[Title](url): note' links to key pages."
    },
    {
      "code": "schema-missing",
      "severity": "high",
      "fix": "No schema.org markup. Add JSON-LD (Organization/WebSite site-wide; Article, Product, FAQPage or HowTo per page)."
    }
  ],
  "citationScore": 55,
  "grade": "C"
}

Limits

• It checks technical signals only (robots rules, llms.txt, schema, structure). It does not query ChatGPT, Perplexity or Gemini, so it cannot tell you whether you are cited today.
• Pages that need JavaScript to render their main content are read as served HTML, so client-side text and schema can be missed.
• Robots verdicts cover GPTBot, Google-Extended, PerplexityBot and the other bots listed in the output; new crawler names need an update.
• Crawling stays on the start host and is capped by maxPagesPerSite; Google and social-network hosts are refused.

Pricing

$0.009 per page audited (pay per event). No start fee. robots.txt and llms.txt lookups are free. Pages that fail (HTTP errors, non-HTML, robots-disallowed, timeouts) are not charged. Your maximum cost per run is always respected.

Responsible use

• robots.txt is respected for every page (fail-closed); at most 1 request per second per host; crawl-delay and Retry-After honoured. No logins, no proxies, no anti-bot bypass. Search engines and social networks are refused.
• Only technical metadata is stored; no emails, phone numbers or body text.
• A high score means the page is technically easy to cite; it does not guarantee citations, which depend on the engines' own ranking.