ai-discovery-expert（AI 发现专家）

AI Discovery Expert Skill

Expert-level optimization for AI-powered search engines and answer engines. Covers Perplexity, ChatGPT Search, Gemini, Claude, and other LLM-based discovery platforms.

Auto-Detection

This skill activates when:

• Discussing AI search engines (Perplexity, ChatGPT Search, Gemini)
• Optimizing content for LLM citation and discovery
• Implementing AI-friendly content structures
• Working on answer engine optimization (AEO)
• Discussing generative search and AI crawlers

1. AI Discovery vs Traditional SEO

comparison[4]{aspect,traditional_seo,ai_discovery}:
  Goal,Rank on SERP position 1-10,Get cited in AI answers
  Format,Keywords + backlinks,Clear answers + structured data
  Content,Keyword density,Comprehensive + authoritative
  Measurement,Rankings + clicks,Citations + brand mentions

Key Insight

AI search engines synthesize answers from multiple sources. Your goal is to be cited as an authoritative source, not just ranked.

2. Content Structure for AI Citation

Answer-First Writing

<!-- ✅ GOOD: Direct answer first -->
# What is Server-Side Rendering (SSR)?

Server-Side Rendering (SSR) is a technique where web pages are rendered
on the server and sent as fully-formed HTML to the browser. This improves
initial page load time and SEO compared to client-side rendering.

## How SSR Works
1. User requests a page
2. Server renders the HTML
3. Browser receives complete HTML
4. JavaScript hydrates for interactivity

<!-- ❌ BAD: Buried answer -->
# Understanding Modern Web Development

In the evolving landscape of web development, there are many techniques...
[500 words later]
...which brings us to Server-Side Rendering (SSR).

Structured Answer Patterns

answer_patterns[5]{pattern,usage,example}:
  Definition first,Concepts/terms,"X is a technique that..."
  Step-by-step,How-to guides,"1. First... 2. Then... 3. Finally..."
  Comparison tables,Alternatives,Feature comparison grids
  Pro/con lists,Decision making,"Pros: ... Cons: ..."
  FAQ format,Common questions,"Q: What is? A: It is..."

3. AI-Optimized Content Formats

Comprehensive FAQ Sections

<!-- Structured for both SEO and AI -->
<section itemscope itemtype="https://schema.org/FAQPage">
  <h2>Frequently Asked Questions</h2>

  <div itemscope itemprop="mainEntity" itemtype="https://schema.org/Question">
    <h3 itemprop="name">What is the best framework for SSR?</h3>
    <div itemscope itemprop="acceptedAnswer" itemtype="https://schema.org/Answer">
      <p itemprop="text">
        Next.js is widely considered the best framework for SSR due to its
        built-in support for server components, automatic code splitting,
        and excellent developer experience. Alternatives include Nuxt.js
        for Vue and SvelteKit for Svelte.
      </p>
    </div>
  </div>
</section>

Authoritative Data Tables

<!-- Tables are easily parsed by AI -->
<table>
  <caption>SSR Framework Comparison 2025</caption>
  <thead>
    <tr>
      <th>Framework</th>
      <th>Language</th>
      <th>Performance</th>
      <th>Learning Curve</th>
    </tr>
  </thead>
  <tbody>
    <tr>
      <td>Next.js</td>
      <td>React/TypeScript</td>
      <td>Excellent</td>
      <td>Medium</td>
    </tr>
    <tr>
      <td>Nuxt.js</td>
      <td>Vue/TypeScript</td>
      <td>Excellent</td>
      <td>Easy</td>
    </tr>
  </tbody>
</table>

4. Entity Optimization (Critical for AI)

Build Topic Authority

entity_signals[5]{signal,implementation}:
  Author entities,Author pages with credentials + schema
  Organization entities,About page with company schema
  Topic clusters,Hub pages linking to detailed subtopics
  External mentions,Citations from authoritative sources
  Consistent naming,Same entity names across all content

Author Schema for E-E-A-T

{
  "@context": "https://schema.org",
  "@type": "Person",
  "@id": "https://example.com/authors/john-doe#person",
  "name": "John Doe",
  "jobTitle": "Senior Software Engineer",
  "description": "10+ years experience in web development, contributor to React and Next.js",
  "url": "https://example.com/authors/john-doe",
  "sameAs": [
    "https://twitter.com/johndoe",
    "https://github.com/johndoe",
    "https://linkedin.com/in/johndoe"
  ],
  "knowsAbout": [
    "React",
    "Next.js",
    "Server-Side Rendering",
    "Web Performance"
  ],
  "alumniOf": {
    "@type": "Organization",
    "name": "MIT"
  },
  "worksFor": {
    "@type": "Organization",
    "name": "Company Name"
  }
}

5. Semantic HTML for AI Parsing

Clear Content Hierarchy

<!-- ✅ GOOD: Clear semantic structure -->
<article>
  <header>
    <h1>Complete Guide to Server-Side Rendering</h1>
    <p class="summary">
      Learn how SSR improves performance and SEO for modern web apps.
    </p>
    <time datetime="2025-01-15">January 15, 2025</time>
  </header>

  <nav aria-label="Table of contents">
    <h2>Contents</h2>
    <ol>
      <li><a href="#what-is-ssr">What is SSR?</a></li>
      <li><a href="#benefits">Benefits of SSR</a></li>
      <li><a href="#implementation">Implementation Guide</a></li>
    </ol>
  </nav>

  <section id="what-is-ssr">
    <h2>What is SSR?</h2>
    <p>Server-Side Rendering is...</p>
  </section>

  <section id="benefits">
    <h2>Benefits of SSR</h2>
    <ul>
      <li>Faster initial page load</li>
      <li>Better SEO performance</li>
      <li>Improved accessibility</li>
    </ul>
  </section>

  <aside>
    <h3>Related Articles</h3>
    <ul>
      <li><a href="/static-generation">Static Site Generation</a></li>
      <li><a href="/client-rendering">Client-Side Rendering</a></li>
    </ul>
  </aside>
</article>

6. AI Crawler Considerations

Perplexity Bot

# robots.txt - Allow Perplexity
User-agent: PerplexityBot
Allow: /
Crawl-delay: 1

Common AI Crawlers

ai_crawlers[6]{bot,engine,recommendation}:
  GPTBot,OpenAI/ChatGPT,Allow for AI search visibility
  Google-Extended,Google Gemini,Allow for Gemini citations
  PerplexityBot,Perplexity,Allow - major AI search engine
  ClaudeBot,Anthropic Claude,Allow for Claude citations
  Applebot-Extended,Apple Intelligence,Allow for Apple AI features
  CCBot,Common Crawl,Allow - trains many AI models

Recommended robots.txt

# Allow major AI crawlers for discovery
User-agent: GPTBot
Allow: /

User-agent: Google-Extended
Allow: /

User-agent: PerplexityBot
Allow: /

User-agent: ClaudeBot
Allow: /

User-agent: Applebot-Extended
Allow: /

# Block training-only crawlers if desired
User-agent: CCBot
Disallow: /

# Standard search engines
User-agent: Googlebot
Allow: /

User-agent: Bingbot
Allow: /

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

7. Content Freshness Signals

Keep-Current Patterns

// Show last updated date prominently
<article>
  <header>
    <h1>Next.js SEO Guide</h1>
    <div className="meta">
      <time dateTime="2025-01-15">
        Last updated: January 15, 2025
      </time>
      <span>Originally published: March 2024</span>
    </div>
  </header>
</article>

// Add update log for major changes
<section>
  <h2>Update History</h2>
  <ul>
    <li><strong>Jan 2025:</strong> Added Next.js 15 features</li>
    <li><strong>Oct 2024:</strong> Updated for App Router changes</li>
    <li><strong>Mar 2024:</strong> Initial publication</li>
  </ul>
</section>

Schema with Dates

{
  "@type": "Article",
  "datePublished": "2024-03-15T08:00:00+00:00",
  "dateModified": "2025-01-15T10:30:00+00:00"
}

8. Citation-Worthy Content Patterns

What Gets Cited

citation_factors[6]{factor,why}:
  Original research,Unique data AI can't find elsewhere
  Expert opinions,Authoritative perspectives
  Concrete numbers,Statistics and benchmarks
  Step-by-step guides,Actionable how-to content
  Comparison tables,Structured decision-making data
  Definition sections,Clear explanations of concepts

Content Quality Signals

quality_signals[5]{signal,implementation}:
  Expertise,Author credentials + experience displayed
  Sources,Citations to authoritative references
  Comprehensiveness,Cover topic fully with depth
  Recency,Regular updates with visible dates
  Uniqueness,Original insights not found elsewhere

9. LLM.txt Standard (Emerging)

What is llm.txt?

A proposed standard (like robots.txt) specifically for AI/LLM consumption.

# llm.txt - AI-specific instructions
# Place at site root: https://example.com/llm.txt

# Site identity
name: Example Tech Blog
description: Expert guides on web development and JavaScript frameworks
author: Example Team
contact:  [email protected] 

# Content focus
topics: React, Next.js, TypeScript, Web Performance, SEO

# Citation preferences
citation-name: Example Tech Blog
citation-url: https://example.com

# Content freshness
update-frequency: weekly
last-major-update: 2025-01-15

# Preferred citation format
cite-as: "According to Example Tech Blog..."

# Key resources for AI understanding
resources:
  - /about - Company and author information
  - /glossary - Technical term definitions
  - /guides - Comprehensive tutorials

10. Monitoring AI Citations

Track Brand Mentions

monitoring_tools[4]{tool,purpose}:
  Perplexity,Search your brand - see if cited
  ChatGPT,Ask about your topic - check citations
  Google Alerts,Monitor brand mentions across web
  BrandMentions,Track citations in AI responses

Manual Testing Prompts

Test queries to check AI citations:

1. "What is [your topic]?" - Check if your definition is cited
2. "Best [your product category]" - Check if you appear in recommendations
3. "How to [task you teach]" - Check if your guide is referenced
4. "[Your brand name] review" - Check brand awareness

Quick Reference Checklist

ai_discovery_checklist[12]{check,priority}:
  Answer-first content structure,Critical
  Clear semantic HTML hierarchy,Critical
  Comprehensive FAQ sections,High
  Author/organization schema,High
  Allow AI crawlers in robots.txt,High
  Visible update dates,High
  Original research/data,High
  Structured comparison tables,Medium
  Topic cluster architecture,Medium
  External authority signals,Medium
  llm.txt file (emerging),Low
  Citation monitoring setup,Low

Related Skills

• seo-expert - Traditional search engine optimization
• web-expert - Frontend implementation
• nextjs-expert - Next.js metadata implementation

Version: 1.0.0