AI 검색 준비도 진단

URL을 넣으면 llms.txt·구조화데이터·AI봇 허용·메타 등 AI 검색(GEO) 준비 상태를 점수로 진단합니다.

Enter a URL and the server fetches the page to check llms.txt · structured data (JSON-LD) · AI-crawler access · meta/OG · HTTPS · FAQ schema, scoring readiness out of 100 for citation by AI search engines.

Search is no longer just ten blue links. Generative engines like ChatGPT, Gemini and Perplexity read the web, summarize it, and build an answer — and which sites get cited as sources inside that answer is the new visibility game. Shaping a site so AI can read and cite it well is called GEO (Generative Engine Optimization).

Enter the address of the page you want to audit and the server fetches it directly, checks whether the key signals for being cited by AI search are in place, and scores it out of 100. Each item shows pass or gap, along with what to fix.

How it works

Signals checked and weights

The server reads the target page HTML plus /robots.txt and /llms.txt, judging by rules. Browsers cannot read other sites directly (CORS), so this audit runs server-side.

Item	Weight	Why it matters
HTTPS	15	AI and search do not trust unencrypted sites.
Structured data (JSON-LD)	18	Lets machines understand your topic, brand and content as facts.
llms.txt	15	An index file that tells AI crawlers your summary and key links.
AI-crawler access	14	If robots.txt blocks GPTBot, ClaudeBot and others, you cannot be cited at all.
Meta description	8	Page summary, used in search and AI cards.
Open Graph (OG)	8	Makes titles and images render well in shares and AI cards.
FAQ schema	8	Question–answer format is easy for AI to quote directly.
title tag	8	The page name; without it, the topic is hard to judge.
canonical	6	Names the primary URL when the same content lives at several addresses.

Reading the score

Weights of passing items are summed to a 0–100 score with an A–F grade. Above 85 is well prepared; below 50 means several key signals are missing. The hint on each failing item is your improvement priority.

Limits

Pages behind a login or rendered by JavaScript may show empty HTML to the server and score lower than reality. This is a rule-based audit — it does not actually query an AI chatbot to confirm citation.

Terms explained

GEO: Generative Engine Optimization — shaping a site so AI engines read and cite it well when they build answers.
llms.txt: A text file at the site root that gives AI crawlers a summary and important links, acting as an index.
Structured data (JSON-LD): Data written in a standard format so machines can read a page's meaning (topic, author, Q&A, etc.).
robots.txt: A file that tells crawlers what they may access; blocking AI bots removes you from citation.
canonical: A link tag naming the primary URL when the same content exists at multiple addresses.
Open Graph (OG): Meta tags for the title, description and image used in share and preview cards.

Frequently asked questions

How is GEO different from SEO?

SEO optimizes for clicks in search results; GEO optimizes for being cited as a source inside AI-generated answers. They overlap, but GEO leans harder on machine-readable form — structured data, llms.txt, and AI-crawler access.

If my score is low, will AI never show my site?

The score reflects readiness, not a hard block. A low score does not make citation impossible, but more signals raise the odds that AI understands and cites your content accurately.

Can I audit other people's sites?

Yes, for public pages. It is useful for comparing what signals a competitor or reference site has in place. Pages that require a login cannot be audited.

Do I really need llms.txt?

It is not mandatory, but it helps AI crawlers grasp your site quickly. The convention is new, so its absence is not a big penalty — but having it helps.

Are results stored?

No. The URL and results are not saved; the page is fetched and checked fresh each time.