civilrightsmuseum.org — AI Site Grade

Analysis

The site's llms.txt is a 205KB auto-generated dump — the opposite of what an llms.txt should be

The National Civil Rights Museum has an unusually strong technical AI-visibility posture that is undermined by a single self-inflicted wound: its llms.txt file exists but is a 205,862-byte auto-generated dump from Rank Math SEO containing every blog post title and URL, rather than a curated, concise AI briefing. This is the worst possible outcome — the file exists, so AI crawlers will read it, but it provides no structured summary, no key facts, no FAQ, no pricing, no hours, no exhibition descriptions. It is noise.

Crawler Access

All major AI crawlers — GPTBot, ClaudeBot, PerplexityBot, Google-Extended, OAI-SearchBot, ChatGPT-User, Bytespider, Applebot-Extended, anthropic-ai — receive a 200 with full HTML content (617KB) from Cloudflare. No UA-based blocking exists. The robots.txt is a bare Yoast-generated file with a single User-agent: * Disallow: rule and a Crawl-delay: 10. No AI bot is explicitly named or restricted. The site runs on Kinsta (WordPress) behind Cloudflare, with Incapsula WAF signals in DNS. Pages render server-side HTML with no JS-shell risk — all bots see the same content a browser does.

Cold-Knowledge Gap

The LLM knows the museum as "housed in the former Lorraine Motel where MLK was assassinated, opened 1991, $27.5M renovation in 2014, Smithsonian Affiliate, UNESCO World Heritage site." The site itself never mentions UNESCO anywhere in its visible text or schema — a major positioning gap. The LLM also recalls a 2023 Juneteenth controversy (declining then reversing a local event) that the site does not address. The site's own narrative focuses on the Legacy Experience reopening May 16, 2026 — a major capital project the LLM knows nothing about cold. The gap between what the model knows (1991 opening, 2014 renovation, UNESCO) and what the site is actively promoting (2026 Legacy Building reopening, contemporary civil rights commentary) is wide.

Schema Posture

The site has rich JSON-LD schema on every page: LocalBusiness + Organization (with legalName: Lorraine Civil Rights Museum Foundation), Place with GeoCoordinates, WebSite with SearchAction, BreadcrumbList, and WebPage with datePublished/dateModified. The schema includes priceRange: $, openingHours, telephone, email, address, sameAs (Facebook only — missing Twitter/X, Instagram, YouTube, TikTok from the sameAs array despite those links existing in the page footer). No Event schema on the events page. No FAQPage schema despite FAQ content on the visit page. No TouristAttraction or Museum subtype — the schema uses LocalBusiness rather than the more specific Museum type.

Content & Structure

The homepage is a 376-word, JS-free, text-rich page with clear H1/H2 hierarchy. The blog is active and contemporary (posts from 2025-2026 on voting rights, education, LGBTQIA+ rights, federal oversight of Black history museums). The exhibitions page is thin — 163 words with a JS-powered load-more grid that may not render for all crawlers. The permanent galleries page is even thinner at 102 words. The events page lists upcoming dates (Freedom Award Oct 2025, King Day Jan 2026, Remembering MLK Apr 2026, Juneteenth Jun 2026, Ruby Bridges Reading Festival Aug 2026) but has no Event schema markup.

External Signals

The museum's DNS reveals integration with Blackbaud (ticketing/CRM), Mandrill (email), and Oracle Eloqua (marketing automation). The sameAs schema only lists Facebook, but the footer links to Twitter/X, Instagram, YouTube, and TikTok. The LLM's cold knowledge includes a UNESCO World Heritage designation (as part of "Civil Rights Movement Sites") that the site never claims — a significant missed trust signal for AI-generated answers about the museum's stature.

Findings

1. llms.txt is a 205KB auto-generated dump from Rank Math SEO High

   The llms.txt file exists but is an uncurated dump of every blog post title and URL, providing no structured summary, key facts, or exhibition descriptions. AI crawlers will read it but gain no useful briefing.

   What to change: Replace the auto-generated llms.txt with a curated, concise AI briefing that includes museum name, location, hours, ticket prices, exhibition highlights, and key facts.

2. UNESCO World Heritage designation not mentioned on the site Medium

   The LLM knows the museum is part of a UNESCO World Heritage site, but the site never mentions this designation in visible text or schema, missing a major trust signal for AI-generated answers.

   What to change: Add a mention of the UNESCO designation on the homepage and in the site's schema markup.

3. No Event schema on events page Medium

   The events page lists upcoming events (Freedom Award, King Day, etc.) but lacks Event schema markup, reducing visibility in AI-generated event listings.

   What to change: Add Event schema markup to each event listing on the events page.

4. Schema uses LocalBusiness instead of Museum type Medium

   The site's JSON-LD schema uses LocalBusiness rather than the more specific Museum subtype, which may reduce relevance in AI-generated answers about museums.

   What to change: Change the schema type from LocalBusiness to Museum (a subtype of LocalBusiness) to improve semantic specificity.

5. sameAs schema only includes Facebook, missing other social profiles Low

   The schema's sameAs array lists only Facebook, but the footer links to Twitter/X, Instagram, YouTube, and TikTok. This limits the site's social signal to AI crawlers.

   What to change: Add the museum's Twitter/X, Instagram, YouTube, and TikTok URLs to the sameAs array in the schema.

6. Exhibition pages are thin on content Medium

   The exhibitions page has only 163 words and the permanent galleries page has 102 words, with a JS-powered load-more grid that may not render for all crawlers.

   What to change: Expand the exhibition pages with detailed descriptions, images with alt text, and ensure all content is server-side rendered.

7. LLM lacks knowledge of the Legacy Experience reopening Medium

   The site heavily promotes the Legacy Experience reopening on May 16, 2026, but the LLM's cold knowledge does not include this major capital project, creating a gap in AI-generated answers.

   What to change: Ensure the Legacy Experience page is well-indexed and consider adding it to the llms.txt curated summary.

8. No FAQPage schema despite FAQ content on visit page Low

   The visit page contains FAQ content (hours, tickets, etc.) but lacks FAQPage schema markup, reducing the chance of appearing in AI-generated FAQ snippets.

   What to change: Add FAQPage schema markup to the visit page for the Q&A content.

What's working

• All major AI crawlers receive full HTML content — GPTBot, ClaudeBot, PerplexityBot, and others receive a 200 with full HTML content from Cloudflare. No UA-based blocking exists, and robots.txt allows all crawlers.
• Rich JSON-LD schema on every page — Every page includes LocalBusiness, Organization, Place with GeoCoordinates, WebSite with SearchAction, BreadcrumbList, and WebPage schema with dates.
• Pages render server-side HTML with no JS-shell risk — All pages serve full HTML content to crawlers, with no JavaScript dependency for core text. Bots see the same content as browsers.
• Active blog with contemporary civil rights topics — The blog publishes recent posts (2025-2026) on voting rights, education, LGBTQIA+ rights, and federal oversight, keeping the site relevant for AI training data.
• Homepage has clear H1/H2 hierarchy and 376 words of text — The homepage is text-rich with a clear heading structure, providing a strong signal for AI crawlers about the site's topic and authority.
• Sitemap with 80 URLs and index enabled — The sitemap is accessible and contains 80 URLs, helping crawlers discover all pages.