> ## Documentation Index > Fetch the complete documentation index at: https://docs.surfais.com/llms.txt > Use this file to discover all available pages before exploring further. # The AI Readiness page > Your on-site, off-site, and earned readiness at a glance — and what the red items actually cost you. Every brand has an **AI Readiness** page: a live check of how well your site is set up for AI crawlers to reach it, read it, and cite it. It's separate from your [AIS Score](/metrics/ais-score) and your [audit](/audits/what-audits-check) — a fast, always-available snapshot you can re-run any time, rather than a scheduled deep assessment. ## What it checks The page runs a check against your homepage and reports back across three panels, in order of impact. ### On-site technical readiness This is the foundational layer: can AI crawlers actually reach and read your pages at all. Four signals, ordered by how much they matter: * **Server-side rendering** — whether your key content sits in the raw HTML a crawler fetches, or only appears after JavaScript runs. This is the single biggest technical lever: ChatGPT, Claude, and Perplexity fetch pages without executing JavaScript, so content that only renders client-side is effectively invisible to them, even though Google can still reach it by rendering the page fully. * **Homepage reachability** — whether our own scanner could fetch your homepage cleanly. * **Structured data present** — whether we found any schema.org markup on the homepage at all. See [Schema & structured data](/readiness/schema-structured-data). * **llms.txt** — whether a `/llms.txt` file exists, shown for completeness rather than because it's required. See [AI crawler access](/readiness/ai-crawler-access) for why. Each signal gets a plain status rather than a raw pass/fail: a genuine problem is flagged clearly, a missing-but-optional item is marked as such rather than as a failure, and anything we couldn't verify is labelled honestly instead of guessed at. ### AI crawler access Where we can read your `robots.txt`, we show every AI-related crawler we found a rule for and whether that rule is actually costing you visibility — full detail in [AI crawler access](/readiness/ai-crawler-access). ### Schema types found If we detected structured data, we list the specific schema.org types present on your homepage — Organization, Product, and so on. An empty result means we found no readable markup; re-run the check after adding or fixing yours to confirm it worked. ### Authority & off-site presence The final panel looks outward, at earned coverage and third-party presence — the strongest measured driver of AI citations overall: * **Earned editorial citations** — independent domains that AI platforms have actually cited when talking about you, pulled from your scan data. A zero here is flagged as the single highest-impact lever available to you, above any on-site fix, because independent coverage is what AI platforms lean on most when deciding what to cite. * **Off-site presence** — whether we found you on Wikipedia, Trustpilot, and Reddit or community discussion. Wikipedia is a particularly strong lever for ChatGPT visibility specifically. Absence is shown as an opportunity, not a failure. Both parts depend on scan or audit data existing for your brand — until you've run at least one, they show an honest "not assessed yet" rather than a false negative. ## Reading the red items Not every red item here is equally serious, and the page reflects that rather than flattening everything into one colour. The rule: **something is only flagged as critical when it genuinely costs you citations.** A crawler that fetches your pages for AI search results being blocked is critical — that's a direct citation loss. A crawler that only trains a future model being blocked is a legitimate, deliberate choice with no visibility cost, so it's never shown as a failure. Missing structured data or a missing llms.txt file are opportunities, not red failures — neither is something engines are confirmed to require. A page that treats every gap as equally urgent teaches you to ignore all of them. Surfais would rather tell you plainly which two or three things actually move the needle. ## Running a check Click **Run Check** at any time to re-scan your homepage — this typically finishes in under 30 seconds. The page shows when it was last checked, and offers a one-click re-run once results are more than a few days old, so you're never looking at a stale snapshot without knowing it. ## How this relates to your audit and your AIS Score The Readiness page is a live, narrow, always-on technical and off-site check of your homepage. Your [audit](/audits/what-audits-check) is a much deeper, periodic assessment across six weighted categories — including this same technical ground, but also your content, structural readability, and branded-share health — that produces the separate Readiness Score. Your [AIS Score](/metrics/ais-score) is different again: it's what AI platforms are actually saying about you right now, measured from live scans. Use this page for a quick, frequent gut check between audits. See [AIS Score vs Readiness Score](/metrics/ais-vs-readiness) for how the three numbers relate and why they're never merged into one. ## Related pages What structured data does for AI engines, and the types we look for. The crawlers that matter, and how to think about your robots.txt. The deeper, six-category assessment behind your Readiness Score. Why these are two different numbers, never merged into one.