mcgeecars.com — AI Site Grade

Analysis

I have a comprehensive picture now. Let me compile the audit.

The Akamai WAF grants anthropic-ai exclusive full-site access while returning 403 to every other AI crawler and all browser traffic — a configuration that makes the site invisible to every AI engine except Claude.

Crawler Access

The site sits behind Akamai (AkamaiGHost) with a WAF rule that blocks every User-Agent except anthropic-ai. compare_bot_access across 11 UAs showed GPTBot, OAI-SearchBot, ChatGPT-User, ClaudeBot, PerplexityBot, Google-Extended, Bytespider, Applebot-Extended, and even a standard browser all receiving 403 Access Denied from Akamai. Only anthropic-ai gets a 200 with 648KB of full HTML served from nginx. The robots.txt is also 403'd to all non-anthropic UAs — meaning GPTBot, ClaudeBot, and Google-Extended cannot even read the crawl rules. The robots.txt (visible only to anthropic-ai) contains permissive rules for AI bots: GPTBot, OAI-SearchBot, ChatGPT-User, Claude-User, Claude-SearchBot, and PerplexityBot are only blocked from /api/, /apis/, /pixall/, and static asset extensions — the core site is allowed. But the Akamai WAF overrides this entirely.

Content & Schema

The homepage (via anthropic-ai) titles itself "McGee Auto Family – Auto Dealerships in CT, NY, MA, VT, & NH" and serves as a multi-franchise dealer hub for Toyota, Jeep, Ford, Hyundai, Mazda, Audi, BMW, Chevrolet, Chrysler, Dodge, Ram, Volkswagen, and Lexus. The page contains an FAQ section, vehicle inventory tables, and lifestyle filters. However, zero JSON-LD schema was detected on the homepage — no AutoDealer, Organization, LocalBusiness, or Product structured data. The site runs on the Dealer.com (DDC) platform, a common auto-dealer CMS. An llms.txt exists (862KB, one of the largest seen) and contains a comprehensive link directory of inventory pages — a strong signal that the platform intentionally supports AI discovery, but only for the one bot that can reach it.

Cold-Knowledge Gap

The LLM's cold knowledge describes McGee as a New England dealer group serving "Saugus, Wakefield, and Manchester, NH" with "Toyota, Honda, Subaru, Ford, and Hyundai." The actual site reveals a much larger footprint: locations in CT, NY, MA, VT, and NH with brands including Audi, BMW, Lexus, Chrysler, Dodge, Ram, Jeep, Mazda, and Volkswagen — none of which appear in the cold knowledge. The model also missed the "over 60 years" claim and the "McGee Automotive Family" branding (it used "McGee Automotive Family" correctly but listed wrong locations and missing brands). The gap is substantial: the model knows a smaller, older version of the business.

External Signals

Web searches returned zero results for "mcgeecars.com", "McGee Automotive Family reviews", or "McGee Toyota Claremont" — the site has no indexed presence in standard search engines. This is consistent with the Akamai WAF blocking all non-anthropic traffic, including Googlebot. The Wayback Machine shows 250 captures since 2001, confirming the domain is long-established. DNS records show Barracuda email security and Network Solutions nameservers. The site is effectively invisible to Google, Bing, and every AI crawler except Claude — a WAF misconfiguration that creates a single-bot monoculture.

Findings

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

   The Akamai WAF returns 403 to GPTBot, ClaudeBot, Google-Extended, PerplexityBot, and all other tested AI crawlers, as well as standard browsers. Only anthropic-ai receives a 200 response with full HTML. This prevents all other AI engines from indexing the site.

   What to change: Reconfigure the Akamai WAF to allow access for all major AI crawlers (GPTBot, Google-Extended, ClaudeBot, PerplexityBot, etc.) and standard browsers, or at minimum remove the blanket 403 for non-anthropic UAs.

2. robots.txt inaccessible to most AI crawlers High

   The robots.txt file returns 403 to all User-Agents except anthropic-ai. GPTBot, ClaudeBot, and Google-Extended cannot read crawl rules, leaving them without guidance on allowed paths.

   What to change: Ensure robots.txt is publicly accessible and returns 200 for all legitimate crawlers, not just anthropic-ai.

3. No JSON-LD structured data on homepage High

   The homepage contains no JSON-LD schema for AutoDealer, Organization, LocalBusiness, or Product. This limits the site's ability to appear in rich results and AI knowledge panels.

   What to change: Add JSON-LD structured data for AutoDealer, Organization, LocalBusiness, and individual vehicle listings.

4. No indexed presence in search engines High

   Web searches for the domain, brand name, and dealership locations returned zero results. The site is invisible to Google, Bing, and other search engines due to the WAF blocking Googlebot.

   What to change: Allow Googlebot and other search engine crawlers through the WAF to enable indexing.

5. LLM cold knowledge omits most brands and locations Medium

   The LLM's prior knowledge lists only Toyota, Honda, Subaru, Ford, and Hyundai, and locations in Saugus, Wakefield, and Manchester, NH. The actual site includes Audi, BMW, Lexus, Chrysler, Dodge, Ram, Jeep, Mazda, Volkswagen, and locations across CT, NY, MA, VT, and NH.

   What to change: Publish a comprehensive 'about' page listing all brands and locations, and ensure it is accessible to all crawlers.

6. llms.txt only accessible to anthropic-ai Medium

   The site hosts a large llms.txt (862KB) with a comprehensive inventory link directory, but it is only served to anthropic-ai due to the WAF. Other AI crawlers cannot access it.

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

7. No external reviews or mentions found Medium

   Web searches for reviews and Reddit mentions returned zero results. The site has no visible external signals or backlinks from review sites or social media.

   What to change: Encourage customer reviews on Google, Yelp, and DealerRater, and ensure those pages are crawlable.

8. Standard browser traffic blocked by WAF High

   A standard browser User-Agent receives a 403 Access Denied from Akamai, meaning human visitors cannot access the site directly. This is a critical usability issue.

   What to change: Allow standard browser User-Agents through the WAF to enable human access.

What's working

• Full site access granted to anthropic-ai — The Akamai WAF allows anthropic-ai to crawl the entire site, returning 200 with full HTML. This enables Claude to index all content.
• Comprehensive llms.txt with inventory links — The site hosts an 862KB llms.txt containing a detailed directory of inventory pages, providing a strong signal for AI discovery.
• Permissive robots.txt for AI crawlers — The robots.txt (visible to anthropic-ai) allows AI crawlers to access the core site, only blocking /api/, /apis/, /pixall/, and static assets.
• Large sitemap.xml available — The sitemap.xml (593KB) is accessible to anthropic-ai and likely contains a comprehensive list of URLs for indexing.
• Long-established domain with history — The domain has been active since at least 2001 with 250 Wayback Machine captures, indicating a stable online presence.