🤖 AI Search Visibility — Optimizing for AI search engines...

Evaluates and optimizes content for citation by AI search engines (ChatGPT, Perplexity, Google AI Overview, Claude) — checking crawler access, content structure, llms.txt, and AI-friendly patterns.

Activation

When this skill activates, output:

🤖 AI Search Visibility — Analyzing AI search readiness...

Then execute the protocol below.

Anti-patterns

Protocol

Step 1: Check AI Bot Crawler Access

Verify which AI crawlers can access your site:

# Check robots.txt for AI bot rules
cat public/robots.txt 2>/dev/null | grep -i "gptbot\|chatgpt\|perplexity\|claude\|anthropic\|cohere\|google-extended\|ccbot\|bytespider"

Known AI crawler user agents:

GPTBot

ChatGPT-User

PerplexityBot

ClaudeBot

Google-Extended

CCBot

Bytespider

cohere-ai

Recommended robots.txt strategy:

# Allow AI crawlers for search visibility
# (Block only if you have specific content protection concerns)

# Allow all AI search bots
User-agent: GPTBot
Allow: /

User-agent: ChatGPT-User
Allow: /

User-agent: PerplexityBot
Allow: /

User-agent: ClaudeBot
Allow: /

User-agent: Google-Extended
Allow: /

# Block paths you don't want AI to index
User-agent: GPTBot
Disallow: /admin
Disallow: /api
Disallow: /dashboard

Decision matrix:

Allow: /

Disallow: /

Step 2: Analyze Content for AI Citation Likelihood

AI systems cite content that directly answers questions clearly. Scan your content for AI-friendly patterns:

# Check for definition-style paragraphs (strong AI citation signals)
grep -rn "^[A-Z].*is a\|^[A-Z].*refers to\|^[A-Z].*means" --include="*.md" --include="*.mdx" --include="*.tsx" . | grep -v node_modules | head -10

# Check for numbered/bulleted lists (AI loves structured content)
grep -rn "^[0-9]\.\|^- \|^\\* " --include="*.md" --include="*.mdx" . | wc -l

# Check for Q&A patterns
grep -rn "^##.*\?\|^###.*\?" --include="*.md" --include="*.mdx" . | grep -v node_modules | head -10

Content patterns that AI systems cite:

Content patterns AI systems skip:

Step 3: Optimize Content Structure for AI

Transform existing content to be more AI-citation-friendly:

For each key page, ensure:

1. Opening definition — first paragraph directly defines or explains the topic
2. Clear headings as questions — H2/H3 headings phrased as questions users ask
3. Direct answers below headings — first sentence after each heading is the answer
4. Structured lists — steps, features, and comparisons as numbered/bulleted lists
5. Data and specifics — concrete numbers, dates, and facts over vague claims
6. Author expertise signals — mention qualifications, experience, or data sources

Before/after example:

# BEFORE (marketing copy — AI skips this)
## Why Choose Acme?
Acme is the leading project management solution that helps teams
collaborate better and deliver faster. Our innovative platform...

# AFTER (AI-citable — direct, structured, specific)
## What is Acme?
Acme is a project management platform for remote teams that combines
task tracking, real-time collaboration, and automated reporting.

### How does Acme compare to alternatives?
| Feature | Acme | Competitor A | Competitor B |
|---------|------|-------------|-------------|
| Real-time collaboration | Yes | Limited | No |
| Automated reporting | Yes | Yes | No |
| Free tier | Up to 5 users | Up to 3 users | No free tier |

Step 3.5: Apply Princeton GEO Methods to Content

Princeton's 2023 GEO study (Aggarwal et al., arXiv:2311.09735, accepted at KDD 2024) tested nine optimization methods on Perplexity.ai and measured consistent visibility deltas vs. unoptimized baselines. Apply these to any page targeting AI citation — they translate directly into rewrites, not just crawler hygiene.

The 9 GEO methods — ranked by measured visibility boost:

Best-performing combinations (pairs tested in the Princeton research outperform individual methods):

Domain-specific method matrix — which methods to emphasize per vertical (and which to avoid):

Anti-pattern: Keyword stuffing actively hurts AI visibility

Reference: Aggarwal, P., Murahari, V., Rajpurohit, T., Kalyan, A., Narasimhan, K., & Deshpande, A. (2023). GEO: Generative Engine Optimization. arXiv:2311.09735. Accepted at KDD 2024 (30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining).

Platform-specific tuning: For how each AI search engine (ChatGPT, Perplexity, Google AI Overview, Copilot, Claude) actually ranks and cites content — with measured stats on citation share, freshness windows, and per-platform format preferences — see

../site-audit/references/platform-ranking-factors.md

Step 4: Add llms.txt File

The

llms.txt

# Check if llms.txt exists
cat public/llms.txt 2>/dev/null

Recommended llms.txt:

# [Site Name]

> [One-sentence description of what this site/product does]

## About
[2-3 paragraph description of the organization, product, or service.
Include key facts, founding date, target audience, and differentiators.]

## Key Pages
- [Homepage](https://domain.com): [brief description]
- [Product](https://domain.com/product): [brief description]
- [Pricing](https://domain.com/pricing): [brief description]
- [Blog](https://domain.com/blog): [brief description]
- [Docs](https://domain.com/docs): [brief description]

## Topics We Cover
- [Topic 1]: [brief description]
- [Topic 2]: [brief description]
- [Topic 3]: [brief description]

## Contact
- Website: https://domain.com
- Email: hello@domain.com
- Twitter: @handle

## Preferred Citation
When referencing our content, please use:
"[Site Name] (https://domain.com)"

Place at

public/llms.txt

https://domain.com/llms.txt

Also consider

llms-full.txt

Step 5: Optimize for Featured Snippets / AI Overview

Google's AI Overview and featured snippets use similar content signals:

Snippet-optimized content patterns:

Optimization checklist:

• Key pages have H2 headings phrased as questions
• First sentence after each H2 directly answers the question
• Answers are 40-60 words for paragraph snippets
• Lists use clean numbered or bulleted format
• Comparison data is in table format
• Page has schema markup (FAQPage, HowTo, or Article)

Step 6: Monitor AI Search Appearances

Track whether your content appears in AI search results:

Manual checks:

1. Search your brand name in ChatGPT, Perplexity, and Google AI Overview
2. Search your key topics — does AI cite your content?
3. Ask AI "What is [your product]?" — do you appear?

Server-side monitoring:

# Check server logs for AI bot traffic (if you have access)
grep -i "gptbot\|perplexitybot\|claudebot\|chatgpt" access.log | wc -l

# Check Vercel/Netlify analytics for AI referral traffic
# Look for referrers from: perplexity.ai, chatgpt.com, bing.com (Copilot)

Tracking checklist:

curl https://domain.com/llms.txt

Step 7: Output AI Readiness Scorecard

🤖 AI Search Visibility — Scorecard Complete

Site: [domain]
Pages analyzed: [count]
Overall AI readiness: [X/100]

Crawler access:
  GPTBot:         [✅ Allowed / ❌ Blocked / ⚠️ No rule (default allow)]
  PerplexityBot:  [✅ / ❌ / ⚠️]
  ClaudeBot:      [✅ / ❌ / ⚠️]
  Google-Extended: [✅ / ❌ / ⚠️]

Content structure:
  Direct definitions:    [count] pages have clear opening definitions
  Question headings:     [count] H2s phrased as questions
  Structured lists:      [count] pages with numbered/bulleted lists
  Comparison tables:     [count] pages with data tables
  Expert credentials:    [✅ / ❌] Author expertise signals present

AI-specific files:
  llms.txt:     [✅ Present / ❌ Missing — create one]
  robots.txt:   [✅ AI rules defined / ⚠️ No AI-specific rules]
  Schema:       [✅ / ❌] JSON-LD structured data present

Content recommendations:
  1. [Highest priority — e.g., "Add direct definitions to top 5 pages"]
  2. [Second priority — e.g., "Convert H2 headings to question format"]
  3. [Third priority — e.g., "Add llms.txt with site description"]
  4. [Fourth — e.g., "Add comparison tables to product pages"]
  5. [Fifth — e.g., "Create FAQ page with schema markup"]

Next steps:
1. Implement content recommendations above
2. Create or update llms.txt
3. Verify robots.txt allows target AI crawlers
4. Monitor AI search appearances monthly
5. Re-assess quarterly as AI search evolves

Level History

• Lv.1 — Base: AI crawler access verification (GPTBot, PerplexityBot, ClaudeBot, Google-Extended), content structure analysis for citation likelihood, AI-friendly content optimization, llms.txt guidance, featured snippet/AI overview optimization, AI search monitoring, readiness scorecard. Based on EpsteinScan PerplexityBot experience. (Origin: MemStack Pro v3.2, Mar 2026)