mcdanielsautogroup.com — AI Site Grade

Analysis

Anthropic-Only Access: The Site That Blocks Everything Except Claude

The site mcdanielsautogroup.com returns 403 Access Denied to every major AI crawler and every browser — GPTBot, Google-Extended, PerplexityBot, ClaudeBot, ChatGPT-User, and even a standard browser UA all receive the same Akamai-hosted "Access Denied" page. The sole exception is anthropic-ai, which gets a 200 with 1.1MB of HTML served from an nginx backend, bypassing the Akamai WAF entirely. This creates a bizarre inversion: only Anthropic's crawler can read the site, while OpenAI, Google, Perplexity, and every human visitor are locked out.

Crawler Access

The site sits behind Akamai (edgesuite.net) with a WAF rule that blocks all traffic except anthropic-ai. The robots.txt and llms.txt both return 403 — they are not merely absent but actively denied. The sitemap at /sitemap.xml is also 403 for browsers but returns 200 with 211KB of XML to anthropic-ai, revealing hundreds of inventory pages. DNS points to dealer.com nameservers, and the HTML confirms a Dealer.com (DDC) platform with multi-franchise support for Acura, Audi, Porsche, Subaru, Volkswagen. The site redirects from the bare domain to www.mcdanielsautogroup.com for the allowed bot.

Content & Schema

The homepage title reads "Used Cars Columbia SC | Charleston SC | Newnan GA" — three locations across two states. The page is a heavy JS-rendered shell (1.1MB HTML, dozens of bundled JavaScript prefetches, images from pictures.dealer.com). No JSON-LD structured data of any kind was found in the page source. The meta description on inventory pages lists "Acura, Volkswagen, Subaru, Audi, Porsche vehicles." The /about/index.htm page returns a 404 — there is no about page. The /contact/index.htm exists but is also a JS shell. No FAQ, comparison tables, or answer-format signals were detected.

Cold-Knowledge Gap

LLM cold knowledge describes McDaniels Auto Group as a family-owned South Carolina dealership network selling Ford, Chevrolet, Honda, Hyundai, and Kia — with a "No Haggle" policy and over 30 years in business. The actual site sells Acura, Audi, Porsche, Subaru, and Volkswagen — none of the brands the LLM knows about. The LLM mentions Columbia and Charleston locations; the site adds Newnan, GA. The LLM recalls community sponsorships and mixed reviews; no external review sites or press mentions were found in web searches. The gap between what AI models "know" and what the site actually sells is nearly total — wrong brands, missing locations, no about page to correct the record.

External Signals

No external search results returned for any query about McDaniels Auto Group — no reviews, no Reddit threads, no press, no dealer listings. The Wayback Machine has zero snapshots of the domain. The site has effectively no discoverable off-domain footprint, meaning AI models have no secondary sources to triangulate against. The only source an AI could use is the site itself, which is invisible to every crawler except Anthropic's.

Findings

1. Akamai WAF blocks all AI crawlers except anthropic-ai High

   The site returns 403 Access Denied to GPTBot, Google-Extended, PerplexityBot, ClaudeBot, and standard browsers. Only the anthropic-ai user-agent receives a 200 response, bypassing the WAF.

   What to change: Remove the Akamai WAF rule that blocks non-anthropic-ai user-agents, or configure it to allow all legitimate AI crawlers.

2. robots.txt returns 403 instead of serving directives High

   The robots.txt file at /robots.txt returns a 403 Forbidden response, meaning no crawler can read it. This prevents any bot from understanding crawl permissions.

   What to change: Serve a valid robots.txt that allows all crawlers, or at least specifies allowed paths.

3. Sitemap returns 403 to browsers but 200 to anthropic-ai High

   The sitemap at /sitemap.xml is blocked by the WAF for most user-agents but returns 200 with 211KB of XML to anthropic-ai, revealing hundreds of inventory URLs. Other crawlers cannot discover these pages.

   What to change: Make the sitemap publicly accessible to all crawlers by adjusting the WAF rules.

4. No JSON-LD structured data on any page High

   The homepage and inventory pages contain no JSON-LD structured data. This means AI crawlers cannot extract entity information like business name, address, inventory, or reviews in a machine-readable format.

   What to change: Add JSON-LD structured data for LocalBusiness, VehicleForSale, and other relevant schemas to all pages.

5. Pages are heavy JavaScript shells with no static content High

   The homepage is 1.1MB of HTML with dozens of bundled JavaScript prefetches and images from pictures.dealer.com. The actual content is rendered client-side, which many AI crawlers cannot execute, resulting in empty or minimal content extraction.

   What to change: Implement server-side rendering or static HTML fallback for critical content to ensure AI crawlers can read it.

6. About page returns 404 Medium

   The /about/index.htm page returns a 404 error, meaning there is no about page. This is a key page for AI models to learn about the business, its history, and locations.

   What to change: Create an about page with accurate business information, including history, locations, and brands sold.

7. LLM cold knowledge lists wrong brands and locations High

   LLM knowledge describes McDaniels Auto Group as selling Ford, Chevrolet, Honda, Hyundai, and Kia, but the actual site sells Acura, Audi, Porsche, Subaru, and Volkswagen. The LLM also misses the Newnan, GA location. This mismatch means AI models will provide incorrect information to users.

   What to change: Publish accurate brand and location information on the site, and ensure it is crawlable and structured with schema markup.

8. No external search results or off-domain footprint Medium

   Web searches for the domain and brand name return zero results. No reviews, press mentions, or dealer listings exist. The Wayback Machine has no snapshots. This means AI models have no secondary sources to verify or supplement site content.

   What to change: Build an off-domain presence through Google Business Profile, dealer directories, and social media to provide external signals.

9. llms.txt returns 403 Medium

   The llms.txt file, intended to provide AI-friendly documentation, returns a 403 Forbidden response. This prevents AI crawlers from accessing any curated AI guidance.

   What to change: Create and serve a publicly accessible llms.txt file with a summary of the business and key pages.

10. No FAQ, comparison, or answer-format content Medium

    The site lacks FAQ pages, comparison tables, or other structured answer-format content that AI models often use to extract factual information. This limits the site's ability to appear in AI-generated answers.

    What to change: Add FAQ schema and comparison pages for vehicle models to provide structured answer content.

11. No robots.txt rules for AI crawlers Medium

    Since robots.txt returns 403, no crawler can read any directives. Even if the WAF were fixed, there are no rules to guide AI crawlers on allowed paths.

    What to change: Create a robots.txt that allows all crawlers and includes sitemap location.

12. No Wayback Machine snapshots exist Low

    The Wayback Machine has zero snapshots of the domain, indicating the site has never been archived. This means historical content is unavailable for AI training or verification.

    What to change: Ensure the site is publicly accessible to allow archiving by the Wayback Machine.

What's working

• Anthropic crawler receives full 1.1MB page content — The anthropic-ai user-agent successfully retrieves the full homepage and inventory pages with 200 status, including JavaScript bundles and images. This ensures Claude can index the site's content.
• Sitemap accessible to Anthropic with 211KB of URLs — The sitemap.xml returns 200 to anthropic-ai, containing hundreds of inventory page URLs. This allows Claude to discover the full site structure.
• Multi-franchise inventory pages for Acura, Audi, Porsche, Subaru, Volkswagen — The site has dedicated inventory pages for multiple brands, providing detailed vehicle listings that are crawlable by anthropic-ai.
• Contact page exists and is crawlable by Anthropic — The /contact/index.htm page returns 200 to anthropic-ai, providing a way for Claude to extract contact information.
• Value Your Trade page available to Anthropic — The /value-your-trade.htm page returns 200 to anthropic-ai, offering a service page that can be indexed.
• Hot Used Car Specials page crawlable by Anthropic — The /hot-used-car-specials.htm page returns 200 to anthropic-ai, providing promotional content.
• Event page crawlable by Anthropic — The /mcdaniels-auto-group-event.htm page returns 200 to anthropic-ai, allowing event information to be indexed.
• Dealer.com platform provides structured inventory data — The site uses Dealer.com (DDC) platform, which typically provides structured inventory data and vehicle details that can be extracted by crawlers that can execute JavaScript.