vust

AI Search Visibility / Free Audit

Free GEO Audit: Is Your Page Visible to AI?

Paste a URL into @vustSEObot on Telegram and get six visible checks of crawler access and extraction-ready page structure, plus AI-suggested drafts. The report explains its evidence and limitations; it does not guarantee ranking or citation.

3 free audits/day. No account.Every finding is a fact you can verify yourself
6 deterministic checks10 named AI crawlersFree, no signup — 3/day

The Honest Frame

A GEO score measures eligibility, never a guaranteed citation

This audit space isn't empty — dashboards like Web Aloha, Stackra, Siftly and a crowded field of llms.txt checkers already do a version of this. VUST's actual difference is where it lives: no account, no tab to keep open, no email capture. You paste a URL into Telegram and the report comes back in the same chat, capped at 3 free audits a day.

Six deterministic checks you can verify yourself, plus AI drafts clearly labeled as drafts — never a ranking promise.

Method / result evidence

What the report observes, and what it cannot prove

Tier 1 fetches one public URL and its domain-level robots.txt and llms.txt, then applies deterministic parsers and visible heuristics. Tier 2 reads the fetched page content and proposes drafts. No search-engine ranking, citation-frequency, or whole-site measurement is performed.

CheckEvidence in the resultPractical actionLimitation
AI-crawler accessEach named crawler is allowed or blocked; a matched robots.txt rule is quoted when found.Review the exact rule before changing crawler policy.Access does not prove crawling, indexing, ranking, or citation.
Structured dataJSON-LD presence, parsed @type values, and invalid-block count.Repair invalid JSON and keep schema aligned with visible facts.Parsing does not validate every claim or search-engine eligibility rule.
Answer structureLead-answer, question-heading, FAQ, heading, and visible-word signals.Move a concise answer up and make sections easier to scan.These are extraction-readiness heuristics, not an outcome benchmark.
Title and descriptionPresence and character counts for title and meta description.Add or tighten missing metadata.Length alone does not establish quality or relevance.
llms.txtWhether /llms.txt returned successfully.Treat it as optional context packaging.It has zero score weight and does not control access, indexing, or ranking.
AI suggestionsA proposed answer block, FAQs, JSON-LD draft, and topic gaps.Fact-check and edit every draft before publishing.Suggestions are generated text, not observed facts.
Based on
One fetched URL, deterministic parsers, and separately labeled AI drafts.
Best for
Finding a concrete first set of page-level fixes.
Not ideal for
Whole-site monitoring, citation tracking, or an exportable evidence log.
In short
Use the score to inspect evidence, not to predict a citation.
Specimens

What one check catches

Three of the six Tier-1 checks, and the concrete difference between failing and passing them.

AI-crawler access

Invisible to AI answers

robots.txt has "User-agent: PerplexityBot\nDisallow: /" — quietly added months ago, never revisited.

Readable

No disallow rule for PerplexityBot, ClaudeBot, Google-Extended, GPTBot or CCBot — the audit names the exact line, or confirms none exists.

Structured data

Nothing to parse

Zero <script type="application/ld+json"> blocks on the page, or a block that fails JSON.parse silently.

Machine-readable

Valid FAQPage or Article JSON-LD present — the audit lists which @type values it actually found, and which recommended types are missing.

Answer-block presence

Buried

The page's actual answer sits in paragraph six, under three paragraphs of scene-setting.

Liftable

A concise 1–3 sentence answer sits near the top — the exact chunk an AI answer engine can quote without reading the whole page.

Illustrative example

What a sample audit report looks like

This is a fictional page's report to show the SHAPE of the output — a mix of passes and fails, exactly how the real bot returns it. It is not a real vust.ai audit result; run your own URL through @vustSEObot to see an actual one.

example.com/blog/some-guide55/100 GEO-readiness
  • AI-crawler access

    PerplexityBot blocked by "Disallow: /blog/" — ClaudeBot, Google-Extended, GPTBot, CCBot all allowed.

  • Structured data

    Article + BreadcrumbList JSON-LD found and valid. FAQPage missing (recommended).

  • Answer-block presence

    No concise lead answer detected — the page opens with a 4-sentence intro before the first fact.

  • Chunkability

    6 H2s, 3 phrased as questions, FAQ section detected.

  • Freshness + basics

    Title, meta description and canonical present. No dateModified found.

  • llms.txt

    GET /llms.txt returned 404 — not present at the domain root.

Sample Tier-2 suggestion (AI draft, review before use)

Answer-block draft to paste as the opening paragraph: "This guide covers X in three steps: [step 1], [step 2], [step 3] — each explained below with the specifics most people search for."

Plus 4 suggested FAQ Q&A pairs with ready FAQPage JSON-LD, and a short list of related sub-topics the page doesn't cover yet — every Tier-2 item is a draft to edit, not a fact to publish as-is.

Practical use cases

Who runs a free GEO audit

Content/SEO teams
Check whether a page is actually blocked from PerplexityBot, ClaudeBot, Google-Extended, GPTBot, CCBot or the other named AI crawlers before assuming a citation problem is a content problem.
The exact robots.txt line, or confirmation that all 10 named crawlers are allowed — a fact, not a guess.
Bloggers and solo publishers
Get a mechanical read on whether a post has the structured data, answer-block and FAQ shape AI answers actually lift.
A transparent, itemized score plus a ranked list of what to fix first.
Devs shipping a new page
Run a pre-launch check for llms.txt, JSON-LD validity and basic freshness meta before the page goes live.
Six deterministic checks, each independently verifiable in the page's own HTML.
Anyone who wants a draft fix, not just a diagnosis
Get an AI-suggested answer-block rewrite and FAQ Q&A with ready JSON-LD for a page that's structurally weak.
Copy-paste drafts clearly labeled as suggestions to edit, never facts to publish as-is.
How it works01–07

Six deterministic checks, plus AI-suggested drafts

  1. 01

    1. AI-crawler access

    Parses robots.txt + meta-robots and checks named crawlers by user-agent — PerplexityBot, ClaudeBot, Google-Extended, GPTBot, CCBot among others — reporting each as allowed or blocked individually, quoting the deciding rule.

  2. 02

    2. Structured data

    Parses every JSON-LD block on the page, flags which schema types (FAQPage, Article, Product, HowTo, Organization) are present, missing or invalid — a broken block that silently does nothing is called out by name.

  3. 03

    3. Answer-block presence

    Checks whether a concise, direct answer sits near the top of the page, or whether it's buried under several paragraphs of intro before the first fact.

  4. 04

    4. Chunkability

    Reads heading structure — H1/H2 phrased as questions or entities, FAQ sections, paragraph and list density — the shape that makes a page easy for an AI answer engine to lift cleanly.

  5. 05

    5. Freshness + basics

    Confirms dateModified, title, meta description and canonical tag are present and reasonable length — the basics answer engines still read.

  6. 06

    6. llms.txt

    Checks whether GET {domain}/llms.txt returns 200 at the domain root.

  7. 07

    Then: AI-suggested drafts (Tier-2)

    An LLM reads the page and drafts a rewritten answer-block, 3-5 FAQ Q&A pairs with ready FAQPage JSON-LD, and a short list of entity/topic gaps — always labeled suggestions, never guarantees.

Same tool · in Telegram@vustSEObot

Run your free GEO audit now

Paste any URL into @vustSEObot and get the six-check report back in the same Telegram chat — free, no signup.

Open in Telegram
Quality & trust

Honest about what this measures

A score is eligibility, never a citation guarantee

The composite score is a transparent sum of the six checkable Tier-1 items, with the full breakdown shown — it measures whether the page is structured to be citable, not a ranking and not a promise ChatGPT or Perplexity will actually cite it.

This space isn't empty — the wedge is where it lives

GEO-audit checkers already exist as web dashboards (llms.txt checkers alone are a crowded sub-niche). VUST's real difference: no signup, no email, no dashboard tab — paste a URL into Telegram, get the report in the same chat.

Tier-1 facts vs. Tier-2 drafts, always labeled

Every Tier-1 finding quotes something you can verify yourself (the robots.txt line, the missing schema type). Tier-2 output is explicitly a draft to review and edit, never a fact and never a ranking claim.

Free-only, fail-open, no batch mode yet

10 audits/day per user, no paid tier exists. A system error never blocks you. There's no whole-site crawl and no live citation probe yet — one URL at a time, today.

FAQ

Frequently asked questions

What does the free GEO audit actually check?

Six deterministic Tier-1 checks it can verify itself: whether 10 named AI crawlers (GPTBot, OAI-SearchBot, ChatGPT-User, PerplexityBot, Perplexity-User, ClaudeBot, Google-Extended, CCBot, Applebot-Extended, Amazonbot) are allowed by robots.txt, whether structured data (JSON-LD) is present and valid, whether there's an answer-block near the top, how chunkable the heading structure is, whether the page has basic freshness signals (title, meta description, dateModified), and whether llms.txt exists at the domain root. It also offers Tier-2 AI-suggested drafts — an answer-block rewrite, FAQ Q&A with ready JSON-LD, and entity-gap suggestions — always labeled as suggestions, never guarantees.

Is this GEO audit tool actually free?

Yes — @vustSEObot has no paid tier yet. It's free-only, fail-open, capped at 3 audits per day per user so a system error never blocks you. There's no signup, no email, no credit card — you paste a URL in Telegram and get the report back in the same chat.

How is this different from other GEO/AEO audit tools?

Plenty of GEO-audit checkers already exist as web dashboards — this isn't the first. The actual difference is where it lives: no account, no web app tab to keep open, no email capture. You paste a URL into a Telegram chat you probably already have open and get the report back inline. That's the entire wedge — mechanically, the six Tier-1 checks are comparable to what other GEO tools examine.

Does a good GEO score guarantee ChatGPT or Perplexity will cite my page?

No, and treat any tool that claims otherwise with suspicion. The score measures GEO eligibility — is the page structured so an AI answer engine COULD cite it — never a ranking or a guarantee of being cited. Being readable and well-structured makes citation more likely; it can't force it.

What's the difference between the Tier-1 checks and the AI suggestions?

Tier-1 is deterministic: robots.txt parsing, JSON-LD parsing, heading/word counts — pure facts you can verify yourself by opening the same page. Tier-2 is an LLM (Anthropic Sonnet) reading your content and drafting a rewritten answer-block, FAQ pairs with ready JSON-LD, and topics your page doesn't cover yet. Tier-2 output is always marked as a draft to review, not a fact.

Can I audit an entire site, not just one URL?

Not today — the free tier audits one URL at a time, pasted directly into the chat. A batch/whole-site crawl is a design idea, not something the bot does right now; don't expect it until it actually ships.

Ready when you are

Is your page visible to AI search?

Six deterministic checks plus AI-suggested fixes — free, in Telegram, 10 audits a day, no signup.