# KI-Bereitschaft von gov.uk

> Gemessen auf der Startseite. Gesamtnote: Gut.

Punktzahl 71/100, Note Gut. Gemessen 2026-09-04T22:50:39.215+00:00, Engine ax-audit@4.1.0.

## Prüfungen

### Inhalt — 67/100

#### Content Negotiation — 0/100

- **FAIL** Homepage does not serve Markdown via content negotiation
  Got text/html (HTTP 200) for "Accept: text/markdown"
  Serve a Markdown representation of your pages when agents request "Accept: text/markdown". Agents like Claude Code and Cursor ask for it, and Markdown cuts token usage by roughly 80% against HTML. Cloudflare ("Markdown for Agents") and Vercel can enable this without code changes.
  https://axrush.com/guides/content-negotiation#not-supported
- **WARN** No <link rel="alternate" type="text/markdown"> fallback found on the homepage
  If you cannot enable content negotiation, advertise a Markdown version with <link rel="alternate" type="text/markdown" href="/index.md"> so agents can discover it.
  https://axrush.com/guides/content-negotiation#no-alternate

#### Agent-Bedienbarkeit — 100/100

- **PASS** 120/120 interactive elements have an accessible name
- **PASS** 5/5 form controls are labelled
- **PASS** Method note: this reads markup, not a rendered accessibility tree
  Labels attached by script and roles computed at runtime are invisible here, so treat low proportions as a prompt to check the real tree rather than as a count. Every finding is also a plain accessibility defect.

#### Structured Data — 0/100

- **FAIL** No JSON-LD structured data found
  Add a <script type="application/ld+json"> block in your HTML <head> with schema.org structured data describing your site, organization, or person.
  https://axrush.com/guides/structured-data#not-found

#### HTML Rendering — 100/100

- **PASS** Server-rendered content detected (799 words, 5476 chars of visible text)
- **PASS** Text-to-markup ratio is healthy (6.4%)
- **PASS** Semantic landmarks present (main, section, header, footer, nav)
- **PASS** Single <h1> heading: "The best place to find government services and information"
- **PASS** 4/4 <img> tags have alt attributes

#### SEO Basics — 82/100

- **WARN** <title> is too short (17 chars): "Welcome to GOV.UK"
  Lengthen the title to 20-70 characters with a clear topic indicator.
  https://axrush.com/guides/seo-basics#short-title
- **WARN** Meta description is short (68 chars)
  Expand the description to 70-160 characters with the page's main value proposition.
  https://axrush.com/guides/seo-basics#short-description
- **PASS** Canonical URL: https://www.gov.uk/
- **PASS** <html lang="en">
- **PASS** UTF-8 charset declared
- **PASS** Viewport meta present: "width=device-width, initial-scale=1"

### Zugang — 95/100

#### TLS / HTTPS — 90/100

- **PASS** Site is served over HTTPS
- **PASS** HTTP requests redirect to HTTPS
- **PASS** HSTS max-age=31536000
- **WARN** HSTS does not include subdomains
  Add includeSubDomains to apply HSTS across api., docs., etc. Required for preload list submission.
  https://axrush.com/guides/tls-https#hsts-no-subdomains
- **WARN** HSTS has preload directive but does not satisfy preload-list requirements
  Preload requires max-age >= 31536000 and includeSubDomains. See https://hstspreload.org.
  https://axrush.com/guides/tls-https#hsts-preload-invalid

#### Agent Access — 100/100

- **PASS** All 10 core AI crawler user-agents receive the same page as a regular client

#### KI-Direktiven — 100/100

- **PASS** Homepage is indexable
- **PASS** 1 data-nosnippet region(s) — scoped exclusions, main content still quotable
- **PASS** No directive restricts how AI assistants may use this page

#### HTTP-Hygiene — 80/100

- **WARN** A missing page redirects (301) instead of returning 404
  Location: https://www.gov.uk/ax-audit-probe-c5kvwyd1
  A redirect on a nonexistent path hides the error. Return 404 or 410 so a client can tell the difference.
  https://axrush.com/guides/http-hygiene#soft-404-redirect
- **PASS** Homepage answers after 1 redirect
  301 → https://www.gov.uk/
- **PASS** HEAD requests are supported
- **PASS** Content-Type: text/html with charset

#### Crawl Efficiency — 100/100

- **PASS** Response compressed with Brotli (br)
- **PASS** Cache validator present (ETag)
- **PASS** Conditional request returns 304 Not Modified
- **PASS** Homepage size is reasonable (83.5 KB decompressed)
- **WARN** Roughly 1,369 tokens of content in 21,364 tokens of response (94% markup)
  Estimated at four characters per token.
  An agent pays to receive the markup and then discards it. Serving Markdown on Accept negotiation is the direct fix.
  https://axrush.com/guides/crawl-efficiency#markup-overhead
- **PASS** Homepage responded in 20ms

### Auffindbarkeit — 58/100

#### LLMs.txt — 0/100

- **FAIL** /llms.txt not found
  HTTP 404
  Create a /llms.txt file at your site root following the llmstxt.org specification. It should be a Markdown file starting with "# Your Site Name" and include a description, sections, and links.
  https://axrush.com/guides/llms-txt#not-found

#### Robots.txt — 72/100

- **PASS** /robots.txt exists
- **WARN** 1/12 core AI crawlers configured
  GPTBot — Blocking it keeps your content out of OpenAI model training. It does not affect ChatGPT search citations.
ClaudeBot — Blocking it keeps your content out of Claude model training.
Google-Extended — Controls Gemini training and grounding in Gemini Apps and Vertex AI. It does NOT remove your site from AI Overviews or AI Mode, which follow Googlebot and the snippet directives.
Applebot-Extended — Opts your content out of Apple foundation-model training without affecting Siri or Spotlight.
Amazonbot — General crawler whose content may be used to train Amazon AI models.
Bytespider — Collects training data for ByteDance models. Compliance with robots.txt is disputed.
CCBot — Builds the open Common Crawl corpus that many models train on. Blocking it is the single broadest training opt-out.
OAI-SearchBot — Blocking it removes your site from ChatGPT search answers and citations.
Claude-SearchBot — Blocking it removes your site from the index Claude cites when it searches the web.
PerplexityBot — Blocking it removes your site from Perplexity answers and citations. Not used for model training.
ChatGPT-User — Fetches a page because a ChatGPT user asked for that URL. Blocking it breaks link-following in conversations.
  Add explicit User-agent entries for the missing crawlers with Allow: / for each one.
  https://axrush.com/guides/robots-txt#missing-crawlers
- **WARN** 1 AI crawler(s) have partial path restrictions
  meta-externalagent
  These crawlers have Disallow rules on specific paths. For full AI access, use only "Allow: /" and let the wildcard User-agent: * handle path restrictions.
  https://axrush.com/guides/robots-txt#partial-restrictions
- **PASS** Sitemap directive present
- **WARN** No Content-Signal directive found (optional)
  Declare how crawlers may use your content after access with the Content Signals Policy, e.g.: Content-Signal: search=yes, ai-train=no. Known signals: search, ai-input, ai-train, plus the optional use=immediate|reference|full. Generate yours at contentsignals.org.
  https://axrush.com/guides/robots-txt#missing-content-signals
- **WARN** 1/57 known AI crawlers have explicit rules
  Add explicit User-agent entries for more AI crawlers to maximize discoverability.
  https://axrush.com/guides/robots-txt#low-coverage

#### Meta Tags — 40/100

- **WARN** No AI meta tags (ai:*) found
  Add AI meta tags to your HTML <head>: <meta name="ai:summary" content="Brief description">, <meta name="ai:content_type" content="website">, <meta name="ai:author" content="Your Name">.
  https://axrush.com/guides/meta-tags#no-ai-meta
- **WARN** No rel="alternate" link to llms.txt in HTML
  Add to your <head>: <link rel="alternate" type="text/plain" href="/llms.txt" title="LLM-optimized content">
  https://axrush.com/guides/meta-tags#no-llms-alternate
- **WARN** No rel="alternate" link to the Agent Card in HTML
  Add to your <head>: <link rel="alternate" type="application/json" href="/.well-known/agent-card.json" title="Agent Card">
  https://axrush.com/guides/meta-tags#no-agent-alternate
- **WARN** No rel="me" identity links found
  Add rel="me" links to verify your identity across platforms: <link rel="me" href="https://github.com/yourname">, <link rel="me" href="https://twitter.com/yourname">.
  https://axrush.com/guides/meta-tags#no-rel-me
- **WARN** OpenGraph required tags missing: og:url, og:type
  Add these meta tags: <meta property="og:url" content="...">, <meta property="og:type" content="...">.
  https://axrush.com/guides/meta-tags#og-required-missing
- **WARN** No Twitter Card meta tags found
  Add twitter:card, twitter:title, twitter:description, and twitter:image so X / Threads / Bluesky / Discord agents render link previews correctly.
  https://axrush.com/guides/meta-tags#no-twitter

#### Sitemap — 100/100

- **PASS** Sitemap located: https://www.gov.uk/sitemap.xml
- **PASS** Content-Type is XML (application/xml)
- **PASS** Sitemap-index references 35 child sitemap(s)
- **PASS** 3/3 sample child sitemap(s) reachable
- **PASS** Sample yielded 75000 URL(s) across 3 child(ren)
- **PASS** Newest <lastmod> is recent (4 day(s) ago)

#### HTTP Headers — 85/100

- **PASS** All 7 security headers present
- **WARN** Link header present but does not reference AI discovery files
  Add AI discovery entries to your Link header: Link: </llms.txt>; rel="alternate"; type="text/plain", </.well-known/agent-card.json>; rel="alternate"; type="application/json"
  https://axrush.com/guides/http-headers#no-ai-discovery
- **WARN** No machine-readable discovery relations beyond llms.txt and the Agent Card
  describedby — llms.txt v2 uses this relation to point a page at the llms.txt that covers it.
api-catalog — RFC 9727: the catalog of APIs this publisher offers.
service-desc — RFC 8631: a machine-readable API description.
service-doc — RFC 8631: human documentation for the API.
ai-catalog — Draft: the AI catalog listing agent cards and MCP server cards.
c2pa-manifest — C2PA 2.4: content provenance for media on the page.
license — RSL and other machine-readable licensing terms.
  Advertise what you publish with Link relations so agents stop guessing paths. Add the ones that apply, for example: Link: </llms.txt>; rel="describedby", </.well-known/api-catalog>; rel="api-catalog". Informational in 3.x: this does not affect your score.
  https://axrush.com/guides/http-headers#discovery-relations

### Protokolle — 50/100

#### Agent Card (A2A) — N. z.

- **PASS** No agent-facing surface — an Agent Card does not apply to this site
  No API, MCP server or existing card was found. An Agent Card advertises capabilities another agent can invoke; a site that offers none has nothing to put in it. Run with --profile agent to audit as though it did.

#### OpenAPI Spec — N. z.

- **PASS** No API surface — API discovery does not apply to this site
  No description, catalog, service-desc relation or developer area was found. Run with --profile api to audit as though the site offered one.

#### MCP (Model Context Protocol) — N. z.

- **PASS** No MCP server — MCP discovery does not apply to this site
  A server card describes an MCP server so agents can find it. A site that runs none has nothing to advertise. Run with --profile mcp to audit as though it did.

#### KI-Katalog — 0/100 (berichtet, wertet nicht)

- **WARN** No agent resource catalog found
  Checked robots.txt Agentmap: directive, Link header rel="ai-catalog", <link rel="ai-catalog">, well-known path and /.well-known/ai-catalog.json, /.well-known/ard.json. Both specifications are drafts.
  A catalog is one document listing everything an agent can call here — agent cards, MCP servers, APIs, skills — so a client stops probing four conventions to find out. Worth publishing once you have more than one of those. Informational: both ai-catalog.json and ard.json are still drafts, so this never affects your score.
  https://axrush.com/guides/ai-catalog#not-found

#### Agent-Skills — N. z.

- **PASS** No developer-facing surface — skills do not apply to this site
  No documentation links, llms.txt, or API description found. Skills describe procedures an agent follows; a site with no procedures to teach has nothing to publish.

#### WebMCP — 100/100 (berichtet, wertet nicht)

- **WARN** 3 form(s) on the page, none declared as agent tools
  WebMCP is a W3C Community Group draft in a Chrome origin trial. It is not a standard and adoption is minimal, so this is a forward-looking note, not a defect.
  A declared form is called by an agent rather than driven pixel by pixel. Add toolname and tooldescription to the forms worth automating — search, filter, subscribe — and toolparamdescription to each field.
  https://axrush.com/guides/webmcp#no-annotations

#### Commerce-Discovery — N. z.

- **PASS** No commerce surface — agentic-commerce discovery does not apply to this site
  No Product or Offer structured data, cart links, or product price tags found.

#### Auth-Discovery — N. z.

- **PASS** Nothing on this site requires authorization — auth discovery does not apply
  No API description, API catalog, MCP server card or commerce profile found.

### Richtlinien — 51/100

#### Security.txt — 100/100

- **PASS** /.well-known/security.txt exists
- **PASS** Required field "Contact" present
- **PASS** Required field "Expires" present
- **PASS** Expires date is in the future (2026-12-01)
- **PASS** 2/5 optional fields present

#### RSL License — 0/100

- **FAIL** No RSL license discovery found
  Checked robots.txt License directive, Link header, and <link rel="license" type="application/rsl+xml">
  Declare machine-readable licensing terms for your content with Really Simple Licensing. Add to robots.txt: License: https://your-site.com/license.xml — then publish the RSL document. See https://rslstandard.org.
  https://axrush.com/guides/rsl#not-found

#### Nutzungsrichtlinie — 40/100

- **WARN** No machine-readable usage policy declared
  Checked robots.txt Content-Signal and Content-Usage, the Content-Usage and content-signal response headers, an RSL licence, TDMRep, and the noai meta directive.
  State your terms where they can be read without a lawyer. The lowest-effort option is a Content-Signal line in robots.txt: Content-Signal: search=yes, ai-input=yes, ai-train=no. Absence is neutral, not permission — but it also gives you nothing to point at.
  https://axrush.com/guides/usage-policy#no-policy

---

Markdown-Darstellung von https://axrush.com/report/gov.uk
