# seo-geo > Generative Engine Optimization analysis for AI Overviews, ChatGPT web search, and Perplexity — covering brand mention signals, AI crawler accessibility, llms.txt compliance, passage-level citability scoring, and platform-specific optimization. **Use case**: Make content citable by AI search engines, not just Google **Canonical URL**: https://agentcookbooks.com/skills/seo-geo/ **Topics**: claude-code, skills, marketing, seo **Trigger phrases**: "AI Overviews", "GEO", "AI search", "Perplexity optimization", "AI citations" **Source**: [AgriciDaniel](https://github.com/AgriciDaniel/claude-seo/tree/main/skills/seo-geo) **License**: MIT --- ## What it does `seo-geo` is a Claude Code skill from AgriciDaniel's [claude-seo repo](https://github.com/AgriciDaniel/claude-seo). It scores content for Generative Engine Optimization — the emerging discipline of optimizing for AI-generated answers rather than traditional blue-link rankings. The skill covers five scoring dimensions: citability (25%), structural readability (20%), multi-modal content (15%), authority and brand signals (20%), and technical accessibility for AI crawlers (20%). Key findings the skill surfaces: whether AI crawlers (GPTBot, ClaudeBot, PerplexityBot) are allowed in robots.txt; whether an `/llms.txt` file exists and is structured correctly; whether content contains self-contained 134–167 word answer blocks that AI systems can extract and cite; and whether the brand has presence on the platforms AI systems source from (YouTube 0.737 correlation, Reddit, Wikipedia). The skill notes that only 11% of domains are cited by both ChatGPT and Google AI Overviews for the same query, making platform-specific optimization meaningful. ## When to use it Reach for it when: - You want to understand whether your content is structured to appear in AI Overviews or get cited by ChatGPT web search - You are adding an `/llms.txt` file and want to know what else to fix for AI crawler accessibility - Traditional rankings are strong but AI-referred sessions are low When *not* to reach for it: - The site is brand new with no rankings or brand presence; AI systems preferentially cite sources with existing topical authority and social signals - You need live AI citation data — use `seo-dataforseo ai-mentions` for that; this skill does static analysis of the page structure ## Install Copy the [`seo-geo` SKILL.md](https://github.com/AgriciDaniel/claude-seo/tree/main/skills/seo-geo) into `.claude/skills/seo-geo/`. Trigger phrases: "AI Overviews", "SGE", "GEO", "AI search", "LLM optimization", "Perplexity", "AI citations", "ChatGPT search", "AI visibility". Invoke with `/seo geo ` for a full GEO analysis. DataForSEO integration is optional — if available, the `ai_optimization_chat_gpt_scraper` and `ai_opt_llm_ment_search` tools provide real citation data rather than inferred scores. ## What a session looks like A typical session has three phases: 1. **AI crawler and access audit.** The skill checks `robots.txt` for the nine known AI crawlers, flags which are blocked vs allowed, checks for `/llms.txt` presence and structure, and checks for RSL 1.0 (Really Simple Licensing) implementation. 2. **Citability and structure scoring.** The page is analyzed for self-contained answer blocks at optimal length, question-based headings, direct definition patterns ("X is..."), specific statistics with attribution, and structured data presence — all signals that correlate with AI citation behavior. 3. **Brand mention analysis and platform breakdown.** Scores are broken down by platform: Google AI Overviews (sources from top-10 organic), ChatGPT (Wikipedia 47.9%, Reddit 11.3%), Perplexity (Reddit 46.7%), and Bing Copilot. Top 5 highest-impact changes are ranked, with quick wins (add definitions in first 60 words, add question-based H2s) separated from high-impact work (build Wikipedia presence, build YouTube channel). ## Receipts **Works well:** The AI crawler audit is genuinely useful even if you never touch the citability scoring — finding out that GPTBot or PerplexityBot is blocked in your robots.txt is an actionable fix that can be made in minutes, and it's something most site owners haven't deliberately configured. **Backfires:** The brand mention analysis (YouTube presence, Reddit presence, Wikipedia) identifies the highest-correlation signals for AI visibility, but building them takes months of work. The quick-win items in the skill are faster, but the high-impact ones require sustained effort that no tool can accelerate. **Pattern that works:** Pair this skill with `seo-content` — E-E-A-T improvements and GEO improvements overlap substantially. Answer-first formatting, specific statistics with attribution, and clear authorship help both traditional quality raters and AI citation systems. ## Source and attribution Originally written by [AgriciDaniel](https://github.com/AgriciDaniel). The canonical SKILL.md and supporting files live in the [`seo-geo` folder](https://github.com/AgriciDaniel/claude-seo/tree/main/skills/seo-geo) of the [claude-seo repository](https://github.com/AgriciDaniel/claude-seo). License: MIT. Install, adapt, and redistribute with attribution preserved. This page documents the skill from a practitioner's perspective. For the formal spec and updates, defer to the source repo.