cavenderautogroup.com — AI Site Grade

Analysis

The Live Site is a Black Box to Every AI Crawler

Every major AI crawler — GPTBot, ClaudeBot, PerplexityBot, Google-Extended, OAI-SearchBot, Applebot-Extended — receives a 403 Forbidden from Cloudflare when attempting to access cavenderautogroup.com. The robots.txt and llms.txt endpoints also return 403 (serving raw HTML/CSS instead of text), meaning no crawler can even read access rules. The site is effectively invisible to the AI ecosystem in real time.

Crawler Access

The site is hosted on Cloudflare (IP 54.243.57.127, AWS EC2 origin) behind a blanket WAF block. All 11 bot user-agents tested returned 403 with Cloudflare's cf-ray headers. The robots.txt file is unreachable — no User-agent, Disallow, or Allow directives exist for any crawler. The llms.txt endpoint also returns 403. The homepage title served to blocked clients is the generic "Dealer Website" with zero visible content, confirming a JS-rendered shell that bots cannot execute.

Cold-Knowledge Gap

The LLM model describes Cavender Auto Group as a family-owned chain founded in 1949, selling Chevrolet, Buick, GMC, and Cadillac across South and Central Texas. The live site contradicts this in two ways: the Wayback snapshot shows the group also sells Ford, Jaguar, Land Rover, and Nissan — brands the model does not mention. The model also states the group has locations in Boerne and Kerrville, but the site's own address schema lists only one address (7625 N Loop 1604 E, San Antonio, TX 78233), while the footer navigation links to 12 distinct dealer locations. The model's knowledge is outdated and incomplete.

Schema Posture

The homepage and About page both contain AutoDealer JSON-LD with Organization and WebSite types. The schema includes a single address, a priceRange of "$-$

External Signals

DuckDuckGo returns zero search results for any query containing "Cavender Auto Group" — no organic listings, no reviews, no press, no Reddit threads, no DealerRater pages. The homepage itself embeds DealerRater reviews (including a detailed negative service experience from October 2025), but these are not indexed or discoverable externally. The brand has no detectable off-domain footprint in the search engine used for this investigation.

Findings

1. All AI crawlers receive 403 Forbidden from Cloudflare High

   Every major AI crawler (GPTBot, ClaudeBot, PerplexityBot, Google-Extended, etc.) is blocked with a 403 response. The homepage serves only a generic 'Dealer Website' title with no visible content, indicating a JS-rendered shell that bots cannot execute.

   What to change: Configure Cloudflare WAF to allow AI crawler user-agents (GPTBot, ClaudeBot, etc.) and serve static HTML content to bots.

2. Robots.txt returns 403 and is inaccessible High

   The robots.txt endpoint returns a 403 Forbidden with HTML/CSS content instead of plain text. No crawler can read access rules, and no User-agent or Disallow directives exist for any bot.

   What to change: Serve a valid robots.txt file that allows AI crawlers (e.g., GPTBot, ClaudeBot) and disallow only private paths.

3. llms.txt endpoint returns 403 Medium

   The llms.txt file is not accessible, returning a 403 Forbidden. This prevents AI crawlers from discovering structured guidance about the site's content.

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

4. Sitemap returns 403 with no URLs High

   The sitemap endpoint returns a 403 Forbidden, preventing search engines and AI crawlers from discovering the site's pages.

   What to change: Generate and serve a valid XML sitemap listing all public pages.

5. Zero search results for Cavender Auto Group on DuckDuckGo High

   Multiple searches for 'Cavender Auto Group' and related terms returned zero results on DuckDuckGo. The brand has no organic listings, reviews, or press in the search index used for this audit.

   What to change: Improve SEO fundamentals: ensure the site is crawlable, submit sitemaps to search engines, and build external citations (e.g., Google Business Profile, DealerRater, local directories).

6. LLM knowledge about Cavender Auto Group is outdated and incomplete Medium

   The LLM describes the group as selling only Chevrolet, Buick, GMC, and Cadillac, but the site also lists Ford, Jaguar, Land Rover, and Nissan. The model also mentions locations in Boerne and Kerrville, but the site's schema shows only one address in San Antonio.

   What to change: Publish accurate, structured data about all brands and locations to correct LLM knowledge gaps.

7. AutoDealer JSON-LD schema is missing key fields Medium

   The homepage and About page contain AutoDealer JSON-LD but lack telephone number (empty string), sameAs links (empty array), openingHours, areaServed, and department entries for individual dealerships. The @id URIs use Wayback Machine proxy URLs instead of the canonical domain.

   What to change: Complete the AutoDealer schema with telephone, sameAs, openingHours, areaServed, and department entries for each dealership. Use canonical domain URLs for @id.

8. Homepage is a JavaScript shell with no static content High

   When accessed by a bot, the homepage returns only a generic 'Dealer Website' title and no visible content, indicating a JS-rendered single-page application that AI crawlers cannot execute.

   What to change: Implement server-side rendering (SSR) or static prerendering to serve HTML content to crawlers.

9. No internal URLs discovered via crawling High

   The list_known_urls tool returned zero URLs for the domain, confirming that the site is completely opaque to crawlers.

   What to change: Ensure the site is crawlable by allowing bots and providing a sitemap.

10. Blocked homepage shows generic 'Dealer Website' title Medium

    The HTML title served to blocked clients is 'Dealer Website', which provides no brand or content signal to crawlers.

    What to change: Serve a descriptive title like 'Cavender Auto Group | New and Used Car Dealer in San Antonio, TX' even when blocking bots.

11. No robots.txt rules exist for any crawler High

    The robots.txt file is inaccessible, so no User-agent, Disallow, or Allow directives are available. Crawlers have no guidance on which paths are allowed.

    What to change: Create a robots.txt file that allows AI crawlers and disallows private areas.

12. 12+ dealer locations not represented in schema Medium

    The footer navigation links to 12 distinct dealer locations, but the JSON-LD schema contains only one address. No department or sub-organization entries exist for individual dealerships.

    What to change: Add department or sub-organization entries in the schema for each dealer location with its own address, phone, and hours.

What's working

• AutoDealer JSON-LD schema is present on homepage and About page — Both pages include structured data with AutoDealer, Organization, and WebSite types, providing a foundation for AI understanding.
• Wayback Machine snapshot captures full site content — A recent Wayback Machine snapshot (October 2025) successfully archived the homepage and About page with full content, preserving the site's information for historical reference.
• Footer lists 12 dealer locations with links — The site's footer includes navigation links to 12 individual dealer locations, providing a clear directory of physical stores.
• DealerRater reviews are embedded on the homepage — The homepage includes embedded DealerRater reviews, which can provide social proof if indexed.
• About page contains detailed company history — The About page (via Wayback) includes a narrative about the company's founding in 1949 and its growth, providing context for AI models.
• Cloudflare WAF is actively protecting the site — The site uses Cloudflare with WAF rules, which provides DDoS protection and security, though currently over-blocking.