jimtrenaryofallon.com — AI Site Grade

Analysis

Exclusive anthropic-ai Access While All Other Crawlers Are Blocked

The site jimtrenaryofallon.com operates a two-tier access policy that grants anthropic-ai full content (200, 420KB) while returning 403 Access Denied to every other entity tested — including GPTBot, ClaudeBot, PerplexityBot, Google-Extended, ChatGPT-User, Applebot-Extended, OAI-SearchBot, Bytespider, and even a standard browser. This is not a robots.txt directive; it is enforced at the Akamai CDN edge (edgesuite.net), which serves the 403 pages, while the anthropic-ai requests route to a separate nginx origin server. The non-www and www subdomains behave identically.

Crawler Access

No robots.txt or llms.txt file is reachable — both return 403. No sitemap is accessible. The site is hosted on dealer.com infrastructure (NS: ns1.dealer.com, ns2.dealer.com), a major automotive dealership platform. The Akamai WAF configuration appears to have an explicit allowlist for anthropic-ai's user-agent string, while blocking everything else including real human browsers. This means Google, Bing, and every major AI crawler except Anthropic's cannot index any page content. The site is effectively invisible to search engines and non-Anthropic AI models.

Cold-Knowledge Gap

The LLM's prior knowledge describes Jim Trenary as a multi-brand GM dealership group (Chevrolet, Buick, GMC, Cadillac) with a family-owned history and community involvement. The actual site, as served to anthropic-ai, reveals a single-franchise Chevrolet dealership (CSS class: single-franchise chevrolet). The title tag reads "Jim Trenary Chevrolet | Chevrolet Dealership in O'Fallon, MO" — no mention of Buick, GMC, or Cadillac. The prior knowledge is inaccurate about the brand portfolio, likely derived from older or broader references to the Trenary group rather than this specific location.

Content and Schema Posture

The homepage is a heavy JavaScript application (dealer.com DDC platform) with 420KB of HTML but minimal server-rendered visible text. The page loads dozens of JS bundles for inventory faceting, filtering, and listing. No JSON-LD schema was detected in the initial fetch. The heading structure contains only the "Access Denied" H1 for blocked users. The anthropic-ai response includes structured inventory data but the schema posture is weak — no AutoDealer, Organization, or LocalBusiness markup was found in the raw HTML head. The site relies entirely on client-side rendering for its content, creating a second barrier for crawlers that do get past the WAF.

External Signals

No external search results were returned for the domain or brand name across multiple queries. The Wayback Machine shows a snapshot exists but the content was not retrieved. The dealership appears to have zero indexed presence in standard web search, consistent with the blanket 403 blocking. The only AI model that can currently read this site is Anthropic's Claude, creating a monoculture dependency for AI-driven discovery.

Findings

1. Exclusive anthropic-ai access while all other crawlers are blocked High

   The site grants full content access (200, 420KB) to anthropic-ai but returns 403 Access Denied to all other tested bots and standard browsers, enforced at the Akamai CDN edge.

   What to change: Remove the Akamai WAF allowlist that blocks non-anthropic-ai user agents, or replace it with a permissive policy that allows all legitimate crawlers and human browsers.

2. No robots.txt or sitemap accessible High

   Both robots.txt and sitemap.xml return 403, preventing crawlers from discovering allowed paths or content structure.

   What to change: Serve a robots.txt that allows all crawlers and reference a valid sitemap.xml.

3. No llms.txt file available Medium

   The llms.txt endpoint returns 403, missing an opportunity to guide AI crawlers to key content.

   What to change: Create an llms.txt file listing important pages for AI crawlers.

4. Zero indexed presence in web search High

   Multiple web searches for the domain and brand returned no results, indicating the site is not indexed by Google or Bing due to the blanket 403 blocking.

   What to change: Allow Googlebot and Bingbot access to the site to enable indexing.

5. Heavy JavaScript application with minimal server-rendered content Medium

   The homepage is a heavy JS application (420KB HTML) with minimal visible text server-side, relying on client-side rendering for inventory data, creating a barrier for crawlers.

   What to change: Implement server-side rendering or pre-rendering for key content to ensure crawlers can read it without executing JavaScript.

6. No JSON-LD schema markup detected Medium

   The raw HTML head contains no AutoDealer, Organization, or LocalBusiness schema markup, reducing the site's ability to provide structured information to AI models.

   What to change: Add JSON-LD schema for AutoDealer, Organization, and LocalBusiness to the homepage.

7. LLM prior knowledge inaccurately describes brand portfolio Medium

   The LLM's prior knowledge states Jim Trenary operates a multi-brand GM dealership group, but the site reveals a single-franchise Chevrolet dealership, creating potential misinformation.

   What to change: Ensure the site content clearly states the brand portfolio and consider publishing an llms.txt or structured data to correct external knowledge.

8. Monoculture dependency on Anthropic for AI visibility High

   Only Anthropic's Claude can access the site, creating a single point of failure for AI-driven discovery and limiting exposure to other AI ecosystems.

   What to change: Allow access to other major AI crawlers such as GPTBot, Google-Extended, and PerplexityBot.

What's working

• Full content accessible to anthropic-ai crawler — The anthropic-ai crawler receives a 200 response with 420KB of HTML, including inventory data, enabling Claude to read the site.
• CDN and origin separation with Akamai and nginx — The site uses Akamai CDN for edge delivery and nginx for origin, providing scalability and performance benefits.
• Wayback Machine snapshot available — A snapshot exists for the domain, providing a historical record that could aid in recovery if needed.