tasca.com — AI Site Grade

Analysis

Tasca.com — AI-Visibility Audit

The live tasca.com returns HTTP 403 to every entity — every AI crawler, every browser, every request — behind a Cloudflare wall that serves zero content to anything. The site is a ghost to the entire internet, yet the cold LLM knowledge about Tasca is rich, detailed, and historically grounded, creating a massive gap between what AI models *could* say and what they can actually *verify* from the live domain.

Crawler Access

All eleven tested user-agents — GPTBot, ClaudeBot, PerplexityBot, Google-Extended, OAI-SearchBot, ChatGPT-User, Bytespider, Applebot-Extended, anthropic-ai, Perplexity-User, and a standard browser — receive identical 403 responses from Cloudflare. The robots.txt (last captured October 2024) contains no AI-bot directives whatsoever — only a generic User-agent: * with Crawl-delay: 1 and standard WordPress admin-path disallows. No llms.txt exists (returns the Cloudflare 403 shell). The DNS TXT records include an anthropic-domain-verification token, indicating Tasca intentionally registered for Anthropic's crawler verification, yet ClaudeBot and anthropic-ai are blocked identically to everything else — the verification is configured but non-functional on the live site.

Cold-Knowledge Gap

The frontier model queried cold knows Tasca as a family-owned Rhode Island dealership group operating since the 1940s, famous for pioneering the Ford Mustang Cobra Jet in 1968, running 40+ franchises across Ford, Chevrolet, Hyundai, Kia, Audi, and others in New England, New York, Florida, Illinois, Pennsylvania, and Missouri. It knows the "Tasca Promise" warranty, the multi-state footprint, and even the mixed customer-review sentiment. None of this can be corroborated from the live domain — the site's actual content (visible only via Wayback Machine) confirms the model's priors are accurate, but AI engines performing live retrieval will find a blank 403 wall and must rely entirely on stale third-party citations.

Schema Posture

The Wayback snapshot reveals the site deploys AutoDealer and Organization JSON-LD schema with full address (1300 Pontiac Ave, Cranston, RI 02920), telephone, price range, logo, and @id references. The schema is structurally correct but lives behind the Cloudflare block — AI crawlers never see it. The homepage headings (H1: Welcome to Tasca Automotive Group) and 3,700+ words of visible text confirm the site is content-rich when accessible, but the live domain returns zero words.

External Signals

DuckDuckGo returns zero search results for any query containing "Tasca Automotive Group," "Tasca Ford Cranston," or "tasca.com" — the domain has no organic search presence on that engine. The Wayback Machine shows 2,876 captures of the robots.txt URL since 1999, confirming a long-standing web presence that has recently been locked down. The site runs on a DealerInspire CMS platform (visible in Wayback image URLs) behind Cloudflare, with an Azure WAF on the subdomain group.tascatrucks.com. The contradiction between a well-structured, schema-rich dealership site and its total invisibility to all crawlers is the single defining AI-visibility problem.

Findings

1. Cloudflare 403 blocks all AI crawlers from accessing any content High

   Every tested user-agent (GPTBot, ClaudeBot, PerplexityBot, Google-Extended, OAI-SearchBot, ChatGPT-User, Bytespider, Applebot-Extended, anthropic-ai, Perplexity-User, and standard browser) receives an identical 403 response from Cloudflare. The live site returns zero words of content to any crawler.

   What to change: Configure Cloudflare to allow AI crawlers (GPTBot, ClaudeBot, etc.) through the WAF, or serve a static HTML version to verified bot IPs.

2. Robots.txt lacks any AI-bot directives Medium

   The robots.txt (last captured October 2024) contains only a generic User-agent: * with Crawl-delay: 1 and standard WordPress admin disallows. No AI-specific user-agents are mentioned, leaving crawlers without guidance.

   What to change: Add explicit allow/disallow rules for AI crawlers (e.g., GPTBot, ClaudeBot) in robots.txt.

3. No llms.txt file available Medium

   The llms.txt endpoint returns a Cloudflare 403 shell, indicating no file exists. This prevents AI models from discovering a curated set of URLs for training or retrieval.

   What to change: Create an llms.txt file listing key pages (homepage, inventory, about) for AI crawlers.

4. Anthropic domain verification token present but non-functional High

   DNS TXT records include an anthropic-domain-verification token, indicating Tasca registered for Anthropic's crawler verification. However, ClaudeBot and anthropic-ai are blocked identically to all other crawlers, making the verification ineffective.

   What to change: Ensure Cloudflare allows Anthropic crawlers through, or remove the unused verification token to avoid confusion.

5. Zero organic search results on DuckDuckGo Medium

   Multiple queries for 'Tasca Automotive Group', 'Tasca Ford Cranston', and 'tasca.com' return zero results on DuckDuckGo, indicating the site has no organic search presence on that engine.

   What to change: Improve SEO fundamentals and ensure the site is indexable by search engines to build organic visibility.

6. AutoDealer and Organization schema live behind Cloudflare block High

   The Wayback snapshot reveals valid AutoDealer and Organization JSON-LD schema with full address, telephone, price range, and logo. However, AI crawlers never see this schema because the live domain returns 403.

   What to change: Allow AI crawlers to access the homepage so schema can be retrieved and used for knowledge graph enrichment.

7. Content-rich homepage with 3,700+ words invisible to crawlers High

   The Wayback snapshot shows the homepage contains over 3,700 words of visible text, H1 headings, and structured content. The live site returns zero words to any crawler, wasting valuable content.

   What to change: Remove the blanket 403 block for AI crawlers to expose this content.

8. Subdomain group.tascatrucks.com blocked by Azure WAF Medium

   The subdomain group.tascatrucks.com returns a 403 from Azure WAF with only 5 words of content, further limiting AI visibility.

   What to change: Configure Azure WAF to allow AI crawlers or serve a static version.

9. Sitemap returns 403, no URLs discoverable High

   The sitemap endpoint returns a 403 error, preventing crawlers from discovering the site's URL structure. No URLs were discovered via known URL listing.

   What to change: Ensure the sitemap is publicly accessible and lists all important pages.

10. Rich cold knowledge about Tasca cannot be verified from live site High

    Frontier models have detailed knowledge about Tasca's history (Cobra Jet, 40+ franchises, multi-state footprint) but none of this can be corroborated from the live domain, forcing AI to rely on stale third-party sources.

    What to change: Allow AI crawlers to access the site so that accurate, up-to-date information can be retrieved and cited.

What's working

• Valid AutoDealer and Organization JSON-LD schema on homepage — The Wayback snapshot shows correctly structured schema with address, phone, price range, and logo, which would be valuable for AI knowledge graphs if accessible.
• Content-rich homepage with over 3,700 words of text — The homepage contains substantial descriptive text, H1 headings, and structured content that would be valuable for AI training and retrieval if exposed.
• Anthropic domain verification token present in DNS — The DNS TXT records include an anthropic-domain-verification token, showing intent to allow Anthropic crawlers, though currently non-functional.
• Long-standing web presence with 2,876 Wayback captures — The domain has been captured by the Wayback Machine 2,876 times since 1999, indicating a well-established online presence.
• Rich cold knowledge about Tasca's history and operations — Frontier models have detailed, accurate knowledge about Tasca's founding, Cobra Jet heritage, franchise footprint, and customer sentiment, providing a strong foundation for AI visibility if the live site were accessible.