devonvale.com — AI Site Grade

Analysis

Cloudflare JS Challenge Blocks All AI Crawlers — Site Is Invisible to Every Major Bot

The live devonvale.com returns HTTP 403 with a Cloudflare JS challenge wall to every tested user-agent — GPTBot, ClaudeBot, PerplexityBot, Google-Extended, ChatGPT-User, Applebot-Extended, Bytespider, and even a plain browser — making the site entirely unreadable by any AI crawler or search engine spider.

Crawler Access

Every bot tested receives the same Cloudflare "Verifying your connection" JS challenge page. The robots.txt and llms.txt endpoints also return 403 with the same challenge wall — neither file is accessible. The site is hosted on Cloudflare (A record 23.227.38.65, Shopify-managed) with aggressive bot mitigation enabled. No sitemap is reachable. The Wayback Machine snapshot from February 2025 shows the actual site content, confirming the live site is not a parked domain but a functioning Shopify store that has locked out all automated access.

Cold-Knowledge Gap

The LLM describes Devonvale as a South African mining investment holding company linked to the Menar group, with interests in coal and chrome. The actual site is a UK-based contract bakery in Devon that bakes flapjacks, cereal bars, cakes, and crumbles for retail, foodservice, vending, and sports nutrition clients. The owner is Martin Loader. The tagline is "Baking Better." The phone number is +44(0)1404 549980. The email is [email protected]. The LLM has zero awareness of the real business and instead hallucinates an entirely different company in a different continent and industry.

Schema Posture

The archived homepage and all subpages contain zero JSON-LD schema of any type — no Organization, Bakery, Product, BreadcrumbList, or LocalBusiness markup. The site runs on Shopify but has no structured data for AI engines to parse. The heading structure is flat: the homepage uses H2 for all section titles ("Welcome to Devonvale", "Baking today", "Flapjacks & Cereal Bars") with no H1 present, and the About page uses H2 for its main heading.

External Signals

The site has no detectable external mentions across web search results — no reviews, no Reddit threads, no press coverage, no directory listings, no social media presence beyond a LinkedIn link on the site itself. The domain appears in zero search results for any combination of its actual business keywords. This total absence of off-domain signals means AI models have no third-party sources to correct the hallucinated mining company narrative.

Findings

1. Cloudflare JS challenge blocks all AI crawlers High

   Every tested bot receives a 403 with a Cloudflare JS challenge wall, making the site entirely unreadable by any AI crawler or search engine spider. Robots.txt and llms.txt are also inaccessible.

   What to change: Disable the JS challenge for known AI crawler user-agents or serve a static HTML version to bots. Ensure robots.txt and llms.txt are publicly accessible.

2. LLM hallucinates South African mining company instead of UK bakery High

   The LLM describes Devonvale as a South African mining investment holding company linked to the Menar group, while the actual site is a UK-based contract bakery. This hallucination is reinforced by the site's invisibility to crawlers.

   What to change: Make the site accessible to AI crawlers and publish structured data to correct the LLM's knowledge.

3. No JSON-LD schema on any page High

   The archived homepage and subpages contain zero JSON-LD schema of any type, missing Organization, Bakery, Product, BreadcrumbList, or LocalBusiness markup.

   What to change: Add JSON-LD structured data for Organization, Bakery, Product, and LocalBusiness to all pages.

4. No external mentions across web search results High

   The site has no detectable external mentions, reviews, directory listings, or social media presence beyond a LinkedIn link on the site itself. This total absence of off-domain signals prevents AI models from correcting the hallucinated narrative.

   What to change: Build external signals through directory listings, press releases, social media profiles, and customer reviews.

5. Robots.txt and llms.txt return 403 High

   Both robots.txt and llms.txt endpoints return 403 with the same Cloudflare challenge wall, preventing crawlers from reading any access rules.

   What to change: Ensure robots.txt and llms.txt are publicly accessible without a JS challenge.

6. Sitemap returns 403 High

   The sitemap endpoint returns 403, preventing crawlers from discovering site URLs.

   What to change: Make the sitemap publicly accessible and ensure it lists all important pages.

7. Flat heading structure with no H1 Medium

   The homepage uses H2 for all section titles with no H1 present, and the About page uses H2 for its main heading, which weakens semantic structure for crawlers.

   What to change: Add a single H1 heading on each page and use a proper heading hierarchy.

8. No llms.txt file published Medium

   The llms.txt endpoint returns 403, meaning no guidance is provided to LLMs about which content to use.

   What to change: Create an llms.txt file with a summary of the site and links to key pages.