Documentation Index
Fetch the complete documentation index at: https://mintlify.com/AgricIDaniel/claude-seo/llms.txt
Use this file to discover all available pages before exploring further.
/seo geo analyzes any page through the lens of Generative Engine Optimization — the practice of making content more likely to be cited by Google AI Overviews, Google AI Mode, ChatGPT, Perplexity, and Bing Copilot. It produces a GEO Readiness Score from 0 to 100 across five weighted dimensions, checks which AI crawlers are allowed or blocked in robots.txt, audits llms.txt presence, and surfaces the specific passage-level changes that will have the highest impact on citation rates. For a full list of content quality improvements, pair /seo geo with /seo content.
Primary-source alignment. Claude SEO’s GEO analysis is grounded in Google’s AI Optimization Guide, published under Search Central docs. Google’s stated position — that AEO and GEO are rebranded labels for SEO, and that AI Overviews and AI Mode are grounded in the same ranking and quality systems as classic Search — is the canonical reference for every recommendation this command makes. When community advice contradicts Google’s primary source, Claude SEO defers to Google and notes the contradiction.
Syntax
What Claude SEO Scores
/seo geo evaluates five dimensions, each contributing to the 0–100 GEO Readiness Score:
| Dimension | Weight | What It Measures |
|---|---|---|
| Citability | 25% | Self-contained answer blocks, passage length, specific facts with attribution, definition patterns |
| Structural Readability | 20% | H1→H2→H3 hierarchy, question-based headings, paragraph length, tables and lists |
| Multi-Modal Content | 15% | Text + images, embedded video, infographics, interactive tools |
| Authority & Brand Signals | 20% | Author byline with credentials, publication/update dates, citations to primary sources, entity presence on Wikipedia/Reddit/YouTube/LinkedIn |
| Technical Accessibility | 20% | Server-side rendering vs client-only content, AI crawler access in robots.txt, llms.txt presence, RSL 1.0 licensing |
Passage Citability
Optimal passage length for AI citation is 134–167 words. Research shows approximately 44% of AI citations come from the first 30% of a page — front-load your most citable, self-contained answers rather than burying them below the fold. Strong citability signals/seo geo rewards:
- Self-contained answer blocks that can be extracted without surrounding context
- Direct answer in the first 40–60 words of each section
- Claims attributed to specific sources (study name, publication, date)
- Definitions following “X is…” or “X refers to…” patterns
- Unique data points not available on competing pages
AI Crawler Access
/seo geo checks robots.txt for each AI crawler’s token and reports the current access status:
| Crawler | Owner | robots.txt Token | Recommendation |
|---|---|---|---|
| GPTBot | OpenAI | GPTBot | Allow for AI search visibility |
| OAI-SearchBot | OpenAI | OAI-SearchBot | Allow for AI search visibility |
| ChatGPT-User | OpenAI | ChatGPT-User | Allow for live browsing citation |
| ClaudeBot | Anthropic | ClaudeBot | Allow for AI search visibility |
| PerplexityBot | Perplexity | PerplexityBot | Allow for AI search visibility |
| CCBot | Common Crawl | CCBot | Block if you want to prevent training use |
| anthropic-ai | Anthropic | anthropic-ai | Optional block (training only) |
| Bytespider | ByteDance | Bytespider | Optional block (training only) |
Google-Extended prevents Gemini model training but does not affect Google Search, AI Overviews, or AI Mode — those use Googlebot.
Platform-Specific Scores
Claude SEO scores your content against each major AI surface separately, because citation overlap is lower than most practitioners assume:| Platform | Citation Characteristics | Optimization Focus |
|---|---|---|
| Google AI Overviews | 92% of citations from top-10 ranking pages; 47% from below position 5 | Classic SEO + passage optimization |
| Google AI Mode | 1B+ monthly users; Gemini 3.5 Flash; shares only 13.7% of cited URLs with AI Overviews | Freshness, entity authority, citable passages beyond position 5 |
| ChatGPT | Wikipedia 47.9%, Reddit 11.3% of citations | Entity presence, authoritative primary sources |
| Perplexity | Reddit 46.7%, Wikipedia | Community validation, discussion presence |
| Bing Copilot | Bing index, authoritative sites | Bing SEO, IndexNow protocol |
Three Evidence-Based Myth Rebuttals
Myth 1: llms.txt is a citation lever for AI search
Myth 1: llms.txt is a citation lever for AI search
Verdict: Not currently true for any major AI search system.The evidence is direct from primary sources:
What Claude SEO does:
| Source | Date | Statement |
|---|---|---|
| John Mueller (Google) — Reddit + Bluesky | 2025 | ”No AI system currently uses llms.txt.” Compared it to deprecated meta keywords. |
| Gary Illyes (Google) — Search Central Live | July 2025 | Google has no plans to support llms.txt. |
| SE Ranking — 300,000-domain study | November 2025 | Among the 50 most AI-cited domains, only one had an /llms.txt. |
| OtterlyAI — server-log audit | 2025 | Only 0.1% of AI-bot traffic targets /llms.txt (84 of 62,100 requests). |
/seo geo reports whether llms.txt is present and whether it is well-formed, but assigns no citation-ranking weight to it. If you ask Claude SEO to generate one, it produces a valid template with a clear banner: “no major LLM provider has confirmed consumption as of May 2026; ship for optionality, not for citation.”The one real use case: llms.txt is increasingly consumed by AI coding agents (Cursor, Continue, Cline, Claude Code) when loading per-library documentation. For a developer-tooling site, publishing llms.txt is a net win for agent documentation accuracy.Myth 2: Chunking content into small pieces improves AI citation
Myth 2: Chunking content into small pieces improves AI citation
Verdict: Google explicitly rejects this.Google’s AI Optimization Guide states directly that you do not need to “chunk your content into small pieces for AI.” The guide’s myth-busting section lists this alongside
llms.txt and AI-specific keyword rewriting as tactics that lack a basis in how AI Overviews or AI Mode actually work.Why the myth persists: early GEO research conflated RAG (Retrieval-Augmented Generation) document chunking — a technique used when building private AI systems — with public web content optimization. These are fundamentally different contexts. Public search AI (Google, Bing, ChatGPT) retrieves full pages from an index, not chunks from a vector store you control.What actually works: self-contained answer blocks of 134–167 words. This is not “chunking” — it is writing each section so that it stands alone as a complete answer to one question, which aids both human readers and AI citation engines simultaneously.Myth 3: AI-specific keyword rewriting improves AI citation
Myth 3: AI-specific keyword rewriting improves AI citation
Verdict: Google explicitly rejects this too.From Google’s AI Optimization Guide: you do not need to “rewrite content for AI with specific phrasings or long-tail keyword variations.” Synonym understanding is sufficient — modern search AI systems understand semantic equivalents without requiring exact phrasing.Deeper reason: AI Overviews and AI Mode are grounded in the same ranking systems as classic Search. A page that is not indexed, or not eligible for snippet display in Google Search, will not appear in any AI feature. The eligibility floor is normal SEO.What
/seo geo recommends instead: question-based headings (H2/H3 phrased as the question your target reader is asking), which simultaneously serves classic query matching and passage-extraction citability. This is one technique that genuinely does double duty.Brand Mentions and Entity Presence
Brand mentions correlate 3× more strongly with AI visibility than backlinks, per an Ahrefs study of 75,000 brands (December 2025).| Entity Signal | Correlation with AI Citations |
|---|---|
| YouTube mentions | ~0.737 (strongest signal measured) |
| Reddit mentions | High |
| Wikipedia entity presence | High |
| LinkedIn presence | Moderate |
| Domain Rating (backlinks) | ~0.266 (weak) |
/seo geo checks entity presence across Wikipedia, Reddit, YouTube, and LinkedIn and surfaces which platforms have the highest gap relative to your topic’s authority floor.
IPTC TrainedAlgorithmicMedia
For e-commerce sites using AI-generated product images, Google Merchant Center requires IPTCDigitalSourceType: TrainedAlgorithmicMedia metadata on every AI-generated image. /seo geo flags compliance gaps and references python3 scripts/iptc_ai_label.py for the audit and injection workflow.
Parasite-SEO Risk Scanner (v2 Phase E)
Version 2 added a parasite-SEO risk scanner that checks whether third-party content published on your domain may trigger Google’s November 2024 site reputation abuse policy. This policy targets publishers who host low-quality content from external contributors (sponsored posts, syndicated articles, UGC sections) that exploits the host domain’s authority. The scanner detects high-risk patterns and classifies exposure level.GEO vs /seo content: How They Relate
/seo geo and /seo content are complementary, not overlapping:
| Layer | Command | Focus |
|---|---|---|
| AI-citability layer | /seo geo | Passage structure, crawler access, entity presence, AI surface scores, platform-specific optimization |
| E-E-A-T layer | /seo content | Experience/Expertise/Authoritativeness/Trustworthiness signals, QRG alignment, content quality, AI-pattern detection |
Output
/seo geo writes GEO-ANALYSIS.md with:
- GEO Readiness Score (0–100) with dimension breakdown
- Platform scores — Google AIO, ChatGPT, Perplexity, Bing Copilot
- AI Crawler Access Status — per-crawler allow/block status with
robots.txtdirectives to add llms.txtStatus — present, missing, or malformed; ready-to-use template if absent- Brand Mention Analysis — Wikipedia, Reddit, YouTube, LinkedIn presence gaps
- Passage-Level Citability — identified 134–167 word blocks; specific passages to restructure
- Server-Side Rendering Check — JavaScript dependency analysis for AI crawler accessibility
- Top 5 Highest-Impact Changes with effort estimates
- Schema Recommendations — structured data gaps that affect AI discoverability