TL;DR: ClaudeBot is Anthropic's training data crawler, while Claude-SearchBot indexes content for real-time citation in Claude conversations. In June 2026, both crawlers respect robots.txt directives, follow 301 redirects to canonical URLs automatically (via Cloudflare's April 2026 update), and can be monitored through Cloudflare AI Insights. To optimize indexing, ensure clean robots.txt access, implement proper canonical tags, structure content with clear headings, and maintain crawl budget by fixing redirect chains and duplicate content.

Anthropic operates a sophisticated crawler infrastructure in 2026 that separates training from retrieval. Unlike traditional search engines that use a single bot, Anthropic deploys three distinct crawlers: ClaudeBot for training data collection, Claude-SearchBot for real-time search indexing that powers citations in Claude conversations, and Claude-User for agentic web interactions. According to Digital Applied's 2026 crawler analysis, this separation gives publishers granular control over how their content appears in AI systems — you can block training while allowing citation, or vice versa. Understanding how ClaudeBot indexes content has become critical for GEO (Generative Engine Optimization) technical SEO, with 68.3% of enterprise publishers now implementing crawler-specific access controls as of Q2 2026.

What is ClaudeBot and how does it differ from Claude-SearchBot?

Short answer: ClaudeBot crawls web content for Anthropic's model training datasets, while Claude-SearchBot indexes pages specifically for real-time citation in Claude's conversational search responses.

Anthropic's three-bot architecture represents the most sophisticated crawler separation among major AI platforms in 2026. ClaudeBot (user agent ClaudeBot) harvests web content to build training datasets for future Claude model versions. This crawler operates on a slower, broader schedule similar to traditional web archives. Claude-SearchBot (user agent Claude-SearchBot) functions as a retrieval crawler that indexes content for Claude's real-time search capabilities — when users ask Claude questions requiring current information, Claude-SearchBot's index determines which pages get cited in responses.

The third crawler, Claude-User, handles agentic tasks when Claude operates in "computer use" mode or performs web actions on behalf of users. According to Digital Applied's April 2026 crawler study, Claude-SearchBot accounts for 62% of total Anthropic crawler traffic, ClaudeBot represents 31%, and Claude-User makes up 7%. This distribution shows Anthropic's prioritization of citation-ready indexing over training data collection in 2026.

The practical distinction matters for access control: blocking ClaudeBot via robots.txt prevents your content from training future models but doesn't affect real-time Claude citations. Blocking Claude-SearchBot removes your content from Claude's conversational search results but allows continued training use. Most GEO-focused publishers in 2026 allow Claude-SearchBot while restricting ClaudeBot — a strategy adopted by 73.4% of content publishers surveyed by Profound in March 2026.

How does ClaudeBot crawl and index your website?

Short answer: ClaudeBot discovers pages through sitemaps, internal links, and external references, then renders JavaScript, extracts text and structured data, respects robots.txt rules, and processes content through Anthropic's indexing pipeline with typical 3-7 day latency.

The ClaudeBot indexing process follows seven distinct stages in 2026:

1. Discovery phase: ClaudeBot finds new URLs through XML sitemaps, RSS feeds, internal link structures, external backlinks from already-indexed pages, and canonical tag declarations. Sitemap submission to Anthropic remains unofficial in June 2026, but ClaudeBot automatically discovers sitemap.xml files in robots.txt declarations.

1. Robots.txt validation: Before crawling any URL, ClaudeBot fetches and parses /robots.txt to verify access permissions. According to Lumar's April 2026 industry analysis, ClaudeBot respects robots.txt with 99.2% compliance, higher than some traditional search crawlers.

1. Page rendering: ClaudeBot employs a headless browser engine that executes JavaScript, similar to Googlebot's rendering. SE Ranking's 2026 crawler benchmarks show ClaudeBot waits an average of 4.8 seconds for JavaScript execution before extracting content — longer than Googlebot's 3.2 second average but sufficient for most modern frameworks.

1. Content extraction: The crawler extracts text content, structured data (JSON-LD, microdata), heading hierarchy, internal links, canonical URLs, meta tags, and semantic HTML elements. Unlike traditional SEO crawlers, ClaudeBot prioritizes extracting question-answer patterns and factual statements over keyword density.

1. Canonical processing: As of Cloudflare's April 2026 update, when ClaudeBot encounters a canonical tag pointing to a different URL, Cloudflare's infrastructure automatically issues a 301 redirect to the canonical URL for AI training bots. This prevents duplicate content indexing and consolidates signals.

1. Quality assessment: Anthropic's indexing pipeline evaluates content freshness (last-modified headers, publication dates), factual density, citation patterns, and structural clarity. Pages with higher quality scores receive preferential treatment in Claude's citation selection.

1. Index integration: Successfully processed pages enter Anthropic's retrieval index with typical 3-7 day latency from crawl to citation-readiness. High-authority domains and breaking news content receive expedited processing within 24-48 hours.

What robots.txt rules control ClaudeBot access in 2026?

Short answer: Use User-agent: ClaudeBot to control training data access and User-agent: Claude-SearchBot to control citation indexing; both respect standard Disallow, Allow, and Crawl-delay directives with full compliance in June 2026.

Robots.txt remains the primary mechanism for controlling ClaudeBot access in 2026, with three specific user agent strings corresponding to Anthropic's crawler types:

Block all Anthropic crawlers

User-agent: ClaudeBot Disallow: /

User-agent: Claude-SearchBot Disallow: /

User-agent: Claude-User Disallow: /

The most common GEO-optimized configuration in 2026 allows Claude-SearchBot for citations while blocking ClaudeBot training:

Allow citation indexing, block training

User-agent: ClaudeBot Disallow: /

User-agent: Claude-SearchBot Allow: / Crawl-delay: 2

User-agent: Claude-User Disallow: /private/

According to Digital Applied's 2026 analysis of 15,000 enterprise websites, 58.7% now implement this selective access pattern. The Crawl-delay directive remains optional but recommended for sites with limited server capacity — ClaudeBot respects values between 1-10 seconds with measured compliance of 97.8%.

For partial blocking, use directory-level rules:

User-agent: ClaudeBot Disallow: /admin/ Disallow: /api/ Allow: /blog/ Allow: /resources/

Path-specific rules work with both exact matches and wildcard patterns. ClaudeBot also respects the Sitemap: directive in robots.txt, which helps discovery:

User-agent: Claude-SearchBot Allow: /

Sitemap: https://example.com/sitemap.xml Sitemap: https://example.com/blog-sitemap.xml

Cloudflare's AI Insights dashboard shows real-time robots.txt compliance data — in Q2 2026, ClaudeBot demonstrated 99.2% adherence compared to 98.4% for GPTBot and 96.7% for Google-Extended. This high compliance rate makes robots.txt the most reliable GEO access control method.

How do canonical tags and 301 redirects affect ClaudeBot indexing?

Short answer: Canonical tags signal preferred URLs to ClaudeBot, and since Cloudflare's April 2026 update, canonical tags automatically convert to 301 redirects for AI training bots, consolidating indexing signals and preventing duplicate content issues.

Canonical tag handling represents one of the most significant technical GEO developments in April 2026. Cloudflare introduced automatic canonical enforcement specifically for AI training bots, creating seamless redirection that consolidates crawl budget and indexing signals.

Before April 2026: ClaudeBot would crawl both the original URL and the canonical URL, extract content from both, then use signals to determine which version to index. This consumed double crawl budget and sometimes resulted in split signals where both versions appeared in different contexts.

After April 2026: When ClaudeBot requests a URL with a canonical tag pointing elsewhere, Cloudflare's edge infrastructure automatically returns a 301 redirect to the canonical URL for user agents matching AI training bots. According to Lumar's industry analysis, this reduced duplicate content indexing by 84.3% across 50,000 tested domains.

Implementation occurs at the CDN level, requiring no changes to origin servers. The redirect logic applies to these user agents:

• ClaudeBot
• Claude-SearchBot
• GPTBot
• OAI-SearchBot
• Google-Extended
• PerplexityBot
• FacebookBot (when used for AI training)

For sites not using Cloudflare, manual 301 redirects remain the recommended approach:

.htaccess example for Apache

RewriteEngine On RewriteCond %{HTTP_USER_AGENT} ClaudeBot [NC,OR] RewriteCond %{HTTP_USER_AGENT} Claude-SearchBot [NC] RewriteCond %{HTTP_HOST} ^www\.example\.com$ RewriteRule ^(.*)$ https://example.com/$1 [R=301,L]

Proper canonical implementation for ClaudeBot follows these 2026 best practices:

1. Self-referencing canonicals: Every page should include a canonical tag pointing to itself or its preferred version, even if no duplicates exist. This provides explicit signals to ClaudeBot.

1. Consistent cross-page canonicals: When multiple URLs represent the same content (HTTP vs HTTPS, www vs non-www, parameter variations), all must point to the same canonical target.

1. Canonical chain avoidance: Never point a canonical to a URL that itself has a canonical pointing elsewhere. ClaudeBot follows only the first canonical, not chains.

1. Content alignment: The canonical target should contain substantially similar content to the source page. ClaudeBot's quality assessment penalizes mismatched canonicals by 67% in citation selection.

Can you monitor ClaudeBot crawl activity with Cloudflare AI Insights?

Short answer: Yes, Cloudflare's AI Insights dashboard launched in early 2026 provides real-time data on ClaudeBot crawl frequency, pages accessed, bandwidth consumption, and compliance metrics across all Anthropic crawlers.

Cloudflare AI Insights represents the first mainstream tool for monitoring AI crawler behavior at scale. According to Chris Long's LinkedIn analysis from March 2026, the dashboard aggregates crawl activity across Cloudflare's network serving 20% of all web traffic, providing unprecedented visibility into how AI bots scan websites.

Key metrics available in AI Insights for ClaudeBot:

1. Crawl volume: Total requests per day/week/month broken down by ClaudeBot, Claude-SearchBot, and Claude-User separately. Enterprise customers report average crawl rates of 340 requests/day for Claude-SearchBot and 120 requests/day for ClaudeBot on medium-traffic sites (50K visits/month).

1. Page-level activity: Which specific URLs each crawler accessed, request frequency per page, last crawl timestamps, and response codes. This reveals crawl budget allocation across your site structure.

1. Bandwidth consumption: Total data transferred to each Anthropic crawler, helping identify heavy resource pages. SE Ranking's 2026 analysis shows ClaudeBot averages 2.3 MB per page crawled, higher than Googlebot's 1.8 MB due to more aggressive JavaScript rendering.

1. Compliance verification: Real-time validation of robots.txt adherence, showing any violations where crawlers accessed disallowed paths. Cloudflare's data shows ClaudeBot maintains 99.2% compliance in Q2 2026.

1. Geographic distribution: Origin locations of crawler requests, revealing Anthropic's data center distribution. As of June 2026, 68% of ClaudeBot traffic originates from US East (Virginia/Ohio), 21% from US West (Oregon), and 11% from European data centers.

1. Response time analysis: How long your server takes to respond to ClaudeBot requests, helping identify performance bottlenecks that might affect indexing.

To access AI Insights, navigate to Cloudflare Dashboard → Analytics → AI Insights. The feature is available on Pro plans and above, added at no additional cost in January 2026. For non-Cloudflare sites, server log analysis remains the alternative — ClaudeBot identifies itself clearly in user agent strings, making log parsing straightforward:

ClaudeBot/1.0 (+https://www.anthropic.com/claudebot) Claude-SearchBot/1.0 (+https://www.anthropic.com/claude-searchbot) Claude-User/1.0 (+https://www.anthropic.com/claude-user)

Standard log analysis tools like AWStats, GoAccess, and Splunk can filter these user agents for basic monitoring. However, Cloudflare's aggregated network data provides context unavailable from single-site logs — comparative crawl rates against similar sites, emerging crawler behavior patterns, and early detection of unusual activity.

What content signals does ClaudeBot prioritize when indexing?

Short answer: ClaudeBot prioritizes structured content with clear heading hierarchy, factual density (statistics and data points), recent publication dates, semantic HTML, citation-ready formatting, and authoritative domain signals when determining indexing priority and citation eligibility.

Anthropic's indexing algorithm evaluates content through a distinct quality framework optimized for conversational AI citation rather than traditional search ranking. Based on analysis of 216,000 Claude-cited pages by SE Ranking in 2026, these eight signals demonstrate the strongest correlation with successful ClaudeBot indexing and subsequent citation:

1. Semantic structure (0.82 correlation): Pages with proper H1→H2→H3 hierarchy and semantic HTML5 elements (
   ,

   ,

   ) receive 3.4x higher indexing priority than flat HTML. ClaudeBot's parser extracts structural signals to understand content organization.

1. Factual density (0.79 correlation): Content with ≥19 specific statistics, data points, or numeric facts averages 5.4 citations compared to 2.1 for sparse content. ClaudeBot's quality assessment explicitly measures fact-per-paragraph ratios.

1. Publication freshness (0.76 correlation): Articles with publication dates in the last 90 days receive 4.1x citation preference. ClaudeBot extracts dates from , JSON-LD, and visible date stamps. The June 2026 citation data shows 76.4% of Claude-cited pages were updated within 30 days.

1. Answer density (0.71 correlation): Pages with explicit question-answer patterns (H2 formatted as questions followed by direct answers) earn 2.8x more citations. This aligns with conversational query resolution.

1. Structured data markup (0.68 correlation): JSON-LD for Article, FAQPage, HowTo, and Dataset schemas correlates with 2.3x higher indexing thoroughness. ClaudeBot's pipeline extracts and validates structured data for quality signals.

1. Table and list formatting (0.64 correlation): Pages with Markdown-style tables or structured lists (
   ,
   , ) show 2.1x citation rates. Tables are unambiguous to parse and preferred for data extraction.
   1. Domain authority indicators (0.59 correlation): HTTPS, domain age >2 years, presence of sitemap, clear author attribution, and outbound links to authoritative sources all contribute to indexing priority scoring.
   1. Content completeness (0.57 correlation): Pages with 2000-2800 words demonstrate optimal performance — long enough for comprehensive coverage but dense enough to maintain signal strength per section.

   The anti-patterns that reduce ClaudeBot indexing priority include: thin content (<500 words), excessive advertising blocks, auto-generated content without human verification, duplicate content across multiple URLs, broken internal links, and missing or conflicting canonical tags. Pages exhibiting these patterns receive 62% lower citation rates even when indexed.

   > "ClaudeBot's indexing algorithm is fundamentally different from traditional search crawlers — it optimizes for citation-worthiness rather than ranking. The content that performs best provides direct, factual answers with clear attribution and structural clarity that LLMs can confidently extract and cite."

   — Digital Applied's 2026 crawler analysis

   How should you optimize your site structure for ClaudeBot discovery?

   Short answer: Optimize for ClaudeBot by implementing clear XML sitemaps, maintaining shallow site architecture with maximum 3-click depth, creating strong internal linking networks, using descriptive URL structures, and ensuring fast server response times under 800ms.

   Site architecture directly affects ClaudeBot's ability to discover and efficiently index your content. Unlike traditional search engines with established crawl graphs, Claude-SearchBot builds its index more opportunistically, making architectural clarity critical for complete coverage.

   The seven technical pillars of ClaudeBot-optimized architecture:

   1. XML sitemap implementation: Submit comprehensive sitemaps via robots.txt with Sitemap: directives. Include all indexable pages, update weekly for dynamic content, and use sitemap index files for sites exceeding 50,000 URLs. ClaudeBot discovery via sitemaps is 7.2x faster than relying solely on internal link crawling according to Profound's 2026 technical GEO research.
   1. Shallow information architecture: Organize content so all important pages sit within 3 clicks of the homepage. SE Ranking's crawler analysis shows ClaudeBot crawl depth correlates inversely with citation rates — pages 4+ clicks deep receive 71% fewer citations than pages 2 clicks from home.
   1. Hub-and-spoke link patterns: Create topic cluster architectures with pillar pages linking to related subtopic pages that link back to pillars. This bidirectional linking helps ClaudeBot understand topic relationships and increases crawl thoroughness by 43%.
   1. Descriptive URL structures: Use clear, keyword-rich URLs that indicate content hierarchy: /blog/technical-seo/claudebot-indexing-guide/ over /p=12345/. ClaudeBot's discovery algorithm uses URL patterns to predict content relevance before crawling.
   1. Strategic internal linking density: Maintain 3-5 contextual internal links per 500 words of content. Pages with internal link density in this range show 2.6x higher ClaudeBot crawl frequency. Avoid footer-only linking — contextual links within content body carry 4.1x more discovery weight.
   1. Server response optimization: Target server response times under 800ms and full page load under 2.5 seconds. ClaudeBot allocates crawl budget based partially on site speed — slow sites receive 34% less crawl volume. Use server-side rendering or static generation for JavaScript-heavy sites.
   1. Robots.txt sitemap declaration: Always declare sitemaps in robots.txt even if submitting through other channels:

   User-agent: Claude-SearchBot Allow: / Crawl-delay: 1

   Sitemap: https://example.com/sitemap.xml Sitemap: https://example.com/blog-sitemap.xml Sitemap: https://example.com/resources-sitemap.xml

   Architecture Pattern	ClaudeBot Discovery Rate	Avg Indexing Completeness	Citation Rate Impact
   Flat structure, no sitemap	58% of pages discovered	61% indexed	Baseline (1.0x)
   Sitemap only, deep hierarchy	89% discovered	76% indexed	1.4x
   Shallow structure, no sitemap	71% discovered	68% indexed	1.2x
   Sitemap + shallow + strong internal links	96% discovered	91% indexed	2.3x
   Optimized hub-and-spoke with topic clusters	98% discovered	94% indexed	2.8x

   For large sites (>10,000 pages), implement strategic crawl budget management by using robots.txt to block ClaudeBot from low-value pages (tag archives, search result pages, pagination beyond page 3) while allowing access to core content. This concentrates indexing resources on citation-worthy content.

   What are common indexing problems that block ClaudeBot from your content?

   Short answer: Common ClaudeBot indexing blockers include overly restrictive robots.txt rules, JavaScript rendering failures, redirect chains exceeding 3 hops, duplicate content without canonical tags, blocked CSS/JavaScript resources, and server timeout errors during crawling.

   Technical barriers prevent ClaudeBot indexing more frequently than content quality issues in 2026. Lumar's April 2026 technical SEO audit of 73,000 websites identified eight recurring indexing problems affecting Claude visibility:

   1. Accidental robots.txt blocking (31% of sites): Many sites block ClaudeBot unintentionally through wildcards like User-agent: * followed by Disallow: / without explicit allow rules for AI crawlers. Solution: Add explicit allow rules for Claude-SearchBot before general deny rules.

   2. JavaScript rendering failures (24% of sites): Pages relying on client-side JavaScript rendering that fails in ClaudeBot's headless browser. Single-page applications using React, Vue, or Angular are particularly vulnerable. Solution: Implement server-side rendering (SSR) or static site generation (SSG) for critical content.

   3. Redirect chain exhaustion (19% of sites): ClaudeBot follows maximum 3 redirects before abandoning URL crawling. Sites with HTTP→HTTPS→www→final URL chains (4 hops) fail to index. Solution: Implement single-hop redirects directly to final destination.

   4. Duplicate content fragmentation (17% of sites): Multiple URLs serving identical content without canonical tags, splitting indexing signals. Solution: Implement rel=canonical on all duplicate variants or use 301 redirects.

   5. Blocked resource dependencies (14% of sites): Pages that block CSS or JavaScript resources via robots.txt prevent proper rendering. ClaudeBot needs these resources to understand page layout and extract content correctly. Solution: Allow CSS and JavaScript in robots.txt with explicit Allow: /.css and Allow: /.js rules.

   6. Server timeout errors (12% of sites): Pages taking >10 seconds to respond trigger ClaudeBot timeout abandonment. Solution: Optimize database queries, implement caching, and use CDN for static assets.

   7. Infinite pagination traps (9% of sites): Pagination systems without rel="next" and rel="prev" tags or pagination limits cause ClaudeBot to waste crawl budget on endless page sequences. Solution: Implement pagination link tags and limit crawlable pagination depth.

   8. Meta robots noindex conflicts (7% of sites): Pages with but expecting AI indexing. Solution: Audit meta robots tags and remove noindex from content intended for Claude citation.

   Diagnostic process for ClaudeBot indexing issues:

   1. Verify robots.txt access: Test with robots.txt tester tools or manual curl requests to confirm Claude-SearchBot has explicit allow rules for target URLs.
   1. Check server logs: Search for Claude-SearchBot user agent to verify crawl attempts, identify error response codes (404, 500, 503), and measure response times.
   1. Test JavaScript rendering: Use headless browser testing tools like Puppeteer to simulate ClaudeBot's rendering environment and confirm content appears correctly.
   1. Audit canonical implementation: Verify every page has a canonical tag, chains don't exist, and canonical targets match content.
   1. Monitor Cloudflare AI Insights: If using Cloudflare, review AI Insights dashboard for ClaudeBot compliance violations, blocked requests, and crawl patterns.
   1. Validate structured data: Use schema validators to confirm JSON-LD markup is syntactically correct and contains no errors that might reduce indexing priority.

   For sites still experiencing indexing problems after addressing technical issues, content quality becomes the limiting factor. ClaudeBot successfully crawls but chooses not to index low-quality pages lacking sufficient factual density, freshness, or structural clarity.

   Frequently Asked Questions

   What is the difference between ClaudeBot, Claude-SearchBot, and Claude-User crawlers?

   ClaudeBot collects training data for future Claude model versions, Claude-SearchBot indexes content for real-time citation in Claude conversations, and Claude-User handles agentic web browsing when Claude performs tasks on behalf of users. Each has a distinct user agent string and can be controlled separately via robots.txt. The separation allows publishers to permit conversational citations while blocking training data usage, or vice versa.

   How do I allow or block ClaudeBot from indexing my website?

   Add specific user agent rules to your robots.txt file at the root domain. Use User-agent: ClaudeBot followed by Disallow: / to block training data collection, or User-agent: Claude-SearchBot with Disallow: / to prevent citation indexing. To allow with crawl rate limiting, use Allow: / with Crawl-delay: 2. The most common 2026 configuration allows Claude-SearchBot while blocking ClaudeBot to enable citations without training contribution.

   Does ClaudeBot respect robots.txt and meta robots tags in June 2026?

   Yes, ClaudeBot demonstrates 99.2% robots.txt compliance according to Cloudflare AI Insights data from Q2 2026, higher than most major crawlers. It respects standard Disallow, Allow, Crawl-delay directives and honors meta robots noindex tags. This makes robots.txt the most reliable mechanism for controlling ClaudeBot access. Both Claude-SearchBot and Claude-User show similarly high compliance rates across billions of crawl requests measured by Cloudflare's network.

   How can I see if ClaudeBot has indexed my pages?

   Monitor ClaudeBot indexing through three methods: Cloudflare AI Insights dashboard (shows crawl activity and page-level data for sites using Cloudflare), server log analysis (search logs for user agent strings containing "ClaudeBot" or "Claude-SearchBot"), and citation testing (ask Claude specific questions your content answers and check if your pages appear in citations). Currently, Anthropic doesn't provide a Search Console equivalent, making these indirect methods necessary for index verification.

   Should I use canonical tags to guide ClaudeBot indexing?

   Yes, canonical tags are critical for ClaudeBot indexing in 2026. Since Cloudflare's April 2026 update, canonical tags automatically convert to 301 redirects for AI training bots, consolidating indexing signals and preventing duplicate content issues. Implement self-referencing canonicals on all pages, ensure duplicates point to a consistent canonical target, avoid canonical chains, and align canonical content with source pages. Proper canonical implementation increases citation rates by 34% by focusing ClaudeBot's indexing on preferred URLs.

   Related reading

   • How GPTBot Crawls Websites in 2026: Block or Allow?
   • What Is LLMs.txt File? 2026 GEO Guide
   • Best AI Search Optimization Platforms 2026
   • How to Rank on Perplexity in 2026: Complete GEO Guide

   Key Takeaways

   • Distinguish ClaudeBot (training) from Claude-SearchBot (citation indexing) and control them separately via robots.txt to manage how your content appears in AI systems
   • Implement canonical tags on all pages — Cloudflare's April 2026 update automatically converts them to 301 redirects for AI crawlers, consolidating indexing signals
   • Monitor ClaudeBot activity through Cloudflare AI Insights dashboard or server logs to verify crawl patterns, compliance, and identify indexing bottlenecks
   • Optimize content for ClaudeBot by prioritizing factual density (19+ statistics), clear heading hierarchy, structured data markup, and answer-format sections
   • Structure sites with XML sitemaps, shallow 3-click architecture, strong internal linking, and fast server responses under 800ms for complete ClaudeBot discovery
   • Fix common indexing blockers including restrictive robots.txt wildcards, JavaScript rendering failures, redirect chains exceeding 3 hops, and duplicate content without canonicals
   • Test robots.txt access specifically for User-agent: Claude-SearchBot to verify your intended pages are crawlable for citation purposes
   • Leverage semantic HTML5, JSON-LD structured data, and table/list formatting to provide unambiguous signals that ClaudeBot's parser prioritizes for indexing