Is your website ready for humans and machines?

xcheck audits a page across four dimensions. The findings tell you what to do: code quotes, copy-paste fixes and self-test instructions.

View an example report

Report by email

Enter your website and your email address. We send you a confirmation link: one click starts your scan and you receive the full report. Exactly one scan is free. No further obligations.

Address of the page to audit, e.g. example.com

Already have an account? Sign in

  • Agent readiness

    Do AI assistants like ChatGPT understand your content? And which questions does your page answer?

  • Accessibility

    Can people with disabilities use your page? Checked against WCAG 2.2, including AI judgement and the screen reader experience.

  • SEO

    Does classic search find you? Titles, descriptions, structured data, trust signals.

  • Speed

    How fast does your page load and respond? Measured in the lab and for real users.

Why these four dimensions

Agent readiness (AEO)

More and more visitors ask an AI first. ChatGPT, Perplexity, and Google answer the question directly and cite their sources. Those answers only include what machines can read and verify cleanly. If you are missing there, you lose reach without noticing.

Accessibility

Accessibility decides whether people with disabilities can use your site: with a screen reader, by keyboard, with high contrast. Since June 2025, the German accessibility act (BFSG) requires this from many digital offerings. The report shows you where your site stands.

SEO

Search engines remain the largest plannable channel. Clean titles, structured data, descriptive link texts, and reachable content decide whether your site gets found before anyone knows your name.

Speed

Load time is user experience. Every extra second costs visitors, on mobile first. Google measures Core Web Vitals in the field and factors them into rankings.

Four checks, one foundation

The four dimensions are connected because they share the same foundation: clean, structured content.

An alt text describes an image for blind people. The same text makes the image readable for search engines and AI assistants. A clear heading hierarchy guides screen reader users through the page and shows machines what matters. Structured data helps Google classify your content and gives AI assistants citable facts. Build accessibly and you build machine-readable. Build machine-readable and you get found and cited.

That is why xcheck checks the four dimensions together. One fix often pays into two or three checks at once. The report shows you where.

How it works

  1. Enter URL and email

    You provide your site address and your email address. No account, no password.

  2. Confirm

    One click on the link in your inbox starts the scan. Exactly one scan is free.

  3. Read the report

    After a few minutes your full report is ready: findings, code citations, copy-ready fixes.

Measures what AI search actually asks

AI search fans every prompt out into many sub-questions: price, comparison, freshness, reviews. xcheck derives these fan-out questions for your core topics and shows, with evidenced passages, which ones your page answers and which remain open. Every open question is a concrete content task.

xcheck also checks whether your page carries the traits that otherwise make AI answers cite comparison portals: concrete prices, comparison tables, your own alternatives content, rating markup and a visible last-updated date.

From finding straight to ticket

Every report exports: as a ticket bundle for Jira, Linear, or Plane (matching each tool’s importer), as Markdown or JSON, or as compact context for your LLM. You pick the findings; xcheck bundles one ticket per criterion with all locations, severity, fix recommendation, and code excerpt.

Jira · Linear · Plane · Markdown · JSON · LLM context

Automate with the API

Start scans and fetch reports as JSON, straight from your CI pipeline or your own tools. Create an API key in your account, send it as a bearer token, done. The full OpenAPI specification is public.

Read the API documentation

Reliably checked

  • AI with evidence

    Parts of the audit use AI (Anthropic Claude). Every AI finding cites the code location that backs it. No evidence, no finding.

  • Hosted in Germany

    xcheck runs on Hetzner in Germany. Processing follows the GDPR.

  • ISO-certified operator

    xcheck is a product by mtc.berlin, certified to ISO 27001 and ISO 9001. We have been building software in Berlin for over 25 years.

Frequently asked questions

What does mtc.berlin xcheck check?

xcheck checks a web page in four dimensions: agent readiness (do AI assistants understand your content?), accessibility per WCAG, SEO, and speed. You receive a report with findings, code citations, and concrete fixes.

What does xcheck cost?

One scan is free: confirm your email address, the scan runs, the report is yours. No further obligations. For teams and recurring checks, write to hey@xcheck.dev.

How does the free scan work?

You enter a URL and your email address. After clicking the confirmation link, the scan starts. The report opens after a few minutes and stays reachable via its link.

What is agent readiness (AEO)?

Agent readiness measures whether AI assistants like ChatGPT or Perplexity can understand, use, and cite your page: structured data, machine-readable content, answered user questions. In short: AEO, answer engine optimization.

Is the report legal advice?

No. The findings are guidance, for example on accessibility per WCAG. The report does not provide a binding legal assessment, for example under the German BFSG.

Where is my data stored?

On Hetzner servers in Germany. For parts of the audit, page content goes to AI services (Anthropic; OpenAI only for content moderation), safeguarded by EU standard contractual clauses.

Is xcheck available for agencies and large site portfolios?

Yes. mtc.berlin offers bulk scans across entire site portfolios, white-label reports in your own branding, and team access with organizations and roles. Write to hey@xcheck.dev and we will get back to you with a proposal.

Can I move findings into Jira, Linear, or Plane?

Yes. Every report exports ticket-ready findings: a CSV matching each tool’s importer (Jira CSV import, Linear CLI importer, Plane CSV import) or a Markdown bundle for pasting. Each ticket carries severity, locations with code excerpt, a fix recommendation, and the report link.

Is there an API?

Yes. With a personal API key you start scans via HTTP (POST /api/v1/scans) and fetch finished reports as JSON. The documentation with examples and the OpenAPI specification lives at xcheck.dev/docs/api.