beanauto.com — AI Site Grade

Analysis

Cloudflare Wall Silences a Major Dealership Group

The live beanauto.com returns HTTP 403 for every AI crawler tested — GPTBot, ClaudeBot, Google-Extended, PerplexityBot, ChatGPT-User, Applebot-Extended, and Bytespider all hit a Cloudflare JS-challenge wall that serves zero visible content. The site has no robots.txt, no llms.txt, and no sitemap accessible to any bot. The homepage title is the generic placeholder "Dealer Website" — not the brand name.

Crawler Access

All eleven bot user-agents tested received 403 responses from Cloudflare, with GPTBot and Google-Extended getting the explicit "Attention Required! | Cloudflare" challenge page. The browser baseline (Mozilla/5.0) also gets a 403 with a 236KB HTML shell containing only CSS font data and no readable content. The DNS records confirm Cloudflare is the CDN/WAF layer. Despite having an anthropic-domain-verification TXT record present (indicating the site explicitly verified with Anthropic for Claude usage), ClaudeBot still receives a 403.

Cold-Knowledge Gap

The LLM's prior knowledge about "Bean Auto" describes a small Portland, Oregon vintage European car repair shop specializing in Volvos and BMWs from the 1960s-1980s, with a shop dog mascot and a YouTube following. The actual site (visible only via Wayback Machine) is Bean Automotive Group, a family-owned Miami dealership group founded in 1969 by Gerald Bean and now led by Lori Bean, operating Kendall Toyota, West Kendall Toyota, Lexus of Kendall, and Lexus of West Kendall — selling over 1,400 new vehicles and employing 700+ associates. The cold knowledge is entirely wrong: wrong geography, wrong business type, wrong brands, wrong era.

Schema Posture

The Wayback snapshot reveals the site does have structured data: an AutoDealer schema with @id referencing the group, a WebSite schema, and a WebPage schema. The AutoDealer includes priceRange, telephone, and a PostalAddress for Miami, FL 33186. However, the streetAddress field is empty. The homepage uses a single H1 heading and multiple H2/H3 subheadings. FAQ and table answer signals are present in the archived content.

External Signals

The search engine returned zero external results for any query about Bean Automotive Group, its dealership names, its addresses, or its ownership — an extreme visibility vacuum. No Reddit threads, no review sites, no press mentions surfaced. The entire dealership group appears to have no indexed external footprint accessible via standard web search, compounding the AI-invisibility problem.

Findings

1. Cloudflare blocks all AI crawlers with HTTP 403 High

   Every AI crawler tested (GPTBot, ClaudeBot, Google-Extended, PerplexityBot, ChatGPT-User, Applebot-Extended, Bytespider) receives a 403 response from Cloudflare, serving zero visible content. Even a standard browser gets a 403 with a 236KB empty shell.

   What to change: Configure Cloudflare to allow AI crawlers (e.g., GPTBot, ClaudeBot, Google-Extended) by adjusting WAF rules or creating a bypass for known bot user-agents.

2. No robots.txt file accessible High

   The site returns a 403 for robots.txt, meaning no crawler directives are available. This prevents AI crawlers from learning which paths are allowed or disallowed.

   What to change: Create a robots.txt file that allows AI crawlers to access the site, e.g., by allowing GPTBot and Google-Extended.

3. No llms.txt file published Medium

   The site does not provide an llms.txt file, which is a recommended way to give LLMs guidance on content usage. This missing file reduces AI discoverability.

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

4. No sitemap accessible to AI crawlers High

   No sitemap was discovered for the domain, and the site is entirely blocked, so AI crawlers cannot discover the site's pages.

   What to change: Create and submit an XML sitemap to search engines, and ensure it is accessible to AI crawlers.

5. Homepage title is generic placeholder Medium

   The homepage title is 'Dealer Website' instead of the brand name 'Bean Automotive Group', which harms brand recognition and SEO.

   What to change: Update the homepage title tag to 'Bean Automotive Group | New & Used Car Dealership in Miami'.

6. LLM cold knowledge describes wrong business High

   LLM prior knowledge about 'Bean Auto' describes a small Portland vintage European car repair shop, not the actual Miami-based dealership group. This mismatch is caused by the site being invisible to AI crawlers.

   What to change: Allow AI crawlers to access the site so LLMs can index the correct business information.

7. Zero external search results for the brand High

   Web searches for 'Bean Automotive Group', its dealership names, addresses, and ownership returned zero results. No Reddit, reviews, or press mentions were found, indicating a complete lack of indexed external presence.

   What to change: Build external citations by claiming business listings on Google Business Profile, Yelp, DealerRater, and other automotive directories, and encourage customer reviews.

8. AutoDealer schema has empty streetAddress field Medium

   The AutoDealer structured data on the homepage includes a PostalAddress but the streetAddress field is empty, which reduces schema completeness and local SEO value.

   What to change: Populate the streetAddress field in the AutoDealer schema with the correct physical address.

9. ClaudeBot blocked despite anthropic-domain-verification TXT record High

   The site has an anthropic-domain-verification TXT record, indicating explicit verification for Claude usage, yet ClaudeBot still receives a 403 from Cloudflare.

   What to change: Ensure Cloudflare WAF rules allow ClaudeBot user-agent, given the domain verification is already in place.

What's working

• AutoDealer schema present on homepage — The homepage includes an AutoDealer schema with @id, priceRange, telephone, and PostalAddress, which helps search engines understand the business type.
• WebSite and WebPage schemas present — The homepage includes WebSite and WebPage structured data, providing basic page-level metadata.
• Anthropic domain verification TXT record present — The DNS TXT records include an anthropic-domain-verification entry, indicating the site has explicitly verified with Anthropic for Claude usage.
• Wayback Machine snapshot available with content — A Wayback Machine snapshot from March 2026 contains 966 words of readable content, including structured data and headings, proving the site has content worth indexing.