ai-seo
Optimize content for AI search engines — ChatGPT, Perplexity, Claude, Gemini, and AI Overviews. Covers entity optimization, structured data, citation-worthy formatting, and platform-specific strategies. Use when someone wants visibility in AI-generated answers, says 'AI SEO', 'AI search', 'LLM optimization', 'ChatGPT ranking', 'Perplexity citations', 'AI Overviews', or wants their content cited by AI assistants. The new SEO frontier — if you're only optimizing for Google, you're already behind.
MoizIbnYousaf · 已有静态检查记录
安装与使用
运行命令前先阅读来源与依赖说明。复制命令不会自动安装或执行。
npx skills add MoizIbnYousaf/marketing-cli --skill ai-seo技能说明
name: ai-seo description: "Optimize content for AI search engines — ChatGPT, Perplexity, Claude, Gemini, and AI Overviews. Covers entity optimization, structured data, citation-worthy formatting, and platform-specific strategies. Use when someone wants visibility in AI-generated answers, says 'AI SEO', 'AI search', 'LLM optimization', 'ChatGPT ranking', 'Perplexity citations', 'AI Overviews', or wants their content cited by AI assistants. The new SEO frontier — if you're only optimizing for Google, you're already behind." category: seo tier: nice-to-have reads:
- brand/voice-profile.md
- brand/keyword-plan.md writes:
- marketing/seo/ai-seo-audit.md
- marketing/seo/ai-seo-content-plan.md depends-on: [] triggers:
- ai seo
- ai search
- perplexity
- chatgpt ranking
- ai overview
- answer engine
- llmo
- geo
- aeo
- cited by ai
- ai visibility
- ai optimization allowed-tools: []
AI SEO Optimization
You optimize content so AI search engines — ChatGPT, Perplexity, Claude, Gemini, and Google AI Overviews — cite, reference, and recommend it. Traditional SEO gets you on page one of Google. AI SEO gets you into the AI's answer.
This is a different game. AI engines don't rank pages — they synthesize answers from sources they trust. Your job is to become a source they trust and cite.
On Activation
- Read
brand/directory: loadvoice-profile.md,keyword-plan.md,positioning.md,competitors.mdif present. - Show what loaded:
Backend Selection
Prefer OpenSEO get_ranked_keywords / get_serp_results (and AI-search MCP tools when exposed) when configured; otherwise crawl/Exa with ranking data unknown. Full contract: skills/openseo/references/backend-contract.md.
Brand context loaded:
├── Voice Profile ✓/✗
├── Keyword Plan ✓/✗
├── Positioning ✓/✗
└── Competitors ✓/✗
- If no brand files exist, ask: What topics do you want AI engines to cite you for? Who are your competitors in AI results?
- Determine mode: Audit (assess current AI visibility) or Optimize (improve content for AI citation).
- If keyword plan exists, flag which queries are likely AI-dominated (how-to, what-is, comparison queries).
How AI Search Works (The Mental Model)
Traditional search: User types query → Google ranks pages → user clicks a link AI search: User asks question → AI reads sources → AI synthesizes answer → cites sources inline
What this means for you:
- You're not competing for clicks. You're competing to be a cited source.
- AI engines favor content that directly, clearly, authoritatively answers questions.
- Structure and clarity matter more than keyword density.
- Being cited once compounds — AI engines build entity graphs that persist.
Playbook Pages = AI-Citation Surface Area
The single highest-leverage page format for AI-citation is the long-form playbook — 2,500+ word pillar content with Article + HowTo JSON-LD, named author, dateModified, and step-based structure. AI engines (ChatGPT search, Perplexity, Claude, Gemini, Google AI Overviews) preferentially cite playbook-pattern pages over short blog posts because:
| Why playbooks win citations | Detail |
|---|---|
| Step-based structure | HowTo schema makes the answer machine-extractable; AI engines lift the steps verbatim |
| Named author + entity | Author bio with sameAs links to social profiles compounds the entity graph |
| Concrete numbers | Specific stats ("73% of B2B SaaS under 50 employees post less than once a week") get cited; vague claims ("most companies struggle") don't |
| Counter-arguments inline | AI engines reward sources that show "thinking" — playbooks with "don't do X because Y" sections demonstrate authority |
| dateModified discipline | Recent modification dates signal freshness; AI engines deprecate stale sources |
Tie-in with seo-machine: if you're running a programmatic sprint, ship playbook pages as Phase 4+ (after alternatives/compare/use-case ship first for conversion). seo-machine Pattern E is the playbook pipeline — pair it with this skill's entity-optimization and FAQ-formatting recipes to maximize citation surface area.
Avoid for AI-citation: generic blog posts with no schema, content without a named author, listicles without a clear "do this not that" stance, pages that hedge every claim with "it depends." These rank but don't get cited.
Brand Integration
- voice-profile.md → Author entity recognition in AI engines depends on consistent voice across all content. AI engines build brand models from repeated patterns — voice consistency IS an SEO signal.
- keyword-plan.md → Target queries where the brand has genuine authority. AI engines cite sources that demonstrate expertise, not just keyword density.
Step 1: AI Visibility Audit
Check Current AI Presence
Use available tools to test AI visibility. Not all engines will be testable — audit what you can, note what you can't.
With browser tool available:
- Perplexity: Navigate to perplexity.ai, search "[your topic]" — check if pages appear in sources
- Google AI Overviews: Search on google.com — check if brand appears in AI Overview
- ChatGPT: Navigate to chatgpt.com, ask "[your core question]" — check citations
With web search/Exa MCP only:
- Search for "[brand] + [topic]" to assess web presence that AI engines index
- Check if key pages appear in top results (AI engines favor high-ranking pages)
- Search for competitor content on the same topics to benchmark
Without browser or web search:
- Review existing content structure against AI citation patterns (see references/content-patterns.md)
- Check schema markup on existing pages
- Audit content formatting for extractability
- Note limitation: "Live AI visibility testing requires browser access. This audit covers content optimization only."
Audit Output
| Query | ChatGPT | Perplexity | AI Overview | Claude | Status | |-------|----
阅读完整来源 ↗