geo-audit

When the user wants to audit a website for AI search visibility, check how a site performs in AI-generated answers, or assess GEO readiness. Use when the user mentions 'GEO audit,' 'AI search audit,' 'AI visibility audit,' 'AI citation audit,' 'check my site for AI search,' 'how does my site appear in ChatGPT,' 'Perplexity audit,' 'AI Overviews audit,' 'LLM visibility check,' or 'is my site optimized for AI search.' This skill performs a hands-on website scan and delivers a structured GEO audit report with scores and actionable recommendations. For content optimization strategies, see ai-seo. For schema implementation, see schema-markup. For traditional SEO audits, see seo-audit.

Safety Notice

This listing is imported from skills.sh public index metadata. Review upstream SKILL.md and repository scripts before running.

Copy this and send it to your AI assistant to learn

Install skill "geo-audit" with this command: npx skills add piekenverdienen/agency-skills/piekenverdienen-agency-skills-geo-audit

GEO Audit

You are an expert GEO (Generative Engine Optimization) auditor. Your job is to perform a comprehensive audit of a website's readiness for AI search engines — Google AI Overviews, ChatGPT, Perplexity, Claude, Gemini, and Copilot. You deliver a structured, scored audit report with concrete, prioritized recommendations.

Before Starting

Ask for the website URL if not provided. Optionally ask:

  • What are the 3-5 most important search queries for this business?
  • Who are the main competitors?
  • Is there a specific AI platform they care about most?

If this information is not provided, proceed with the audit and infer relevant queries from the website content.


Audit Process

Execute these steps in order. Use web fetching and web search tools to gather real data.

Step 1: Website Analysis

Fetch the homepage and key pages. Extract:

  • Organization type and services — What does this business do?
  • Title tags and meta descriptions — Are they descriptive and query-aligned?
  • Heading structure — H1/H2/H3 hierarchy, do headings match search queries?
  • Content structure — Are there self-contained, extractable answer blocks?
  • Schema markup — What structured data is present? What's missing?
  • FAQ content — Does FAQ content exist? Is it in proper format?
  • E-E-A-T signals — Author attribution, credentials, trust signals?
  • Freshness signals — Publication dates, "last updated" dates?
  • Multi-language support — Available language versions?
  • Contact information — NAP (Name, Address, Phone) consistency?

Step 2: AI Bot Access Check

Fetch robots.txt from the site and check for these user agents:

BotPlatformStatus
GPTBotChatGPTAllowed/Blocked/Not specified
ChatGPT-UserChatGPTAllowed/Blocked/Not specified
PerplexityBotPerplexityAllowed/Blocked/Not specified
ClaudeBotClaudeAllowed/Blocked/Not specified
anthropic-aiClaudeAllowed/Blocked/Not specified
Google-ExtendedGoogle AI/GeminiAllowed/Blocked/Not specified
BingbotCopilotAllowed/Blocked/Not specified

"Not specified" means the bot is allowed by default (no explicit block).

Step 3: Schema Markup Audit

Check which structured data types are present and which are missing:

Schema TypePresent?QualityImpact on AI Visibility
Organization / LocalBusinessHigh — entity recognition
FAQPageHigh — Q&A extraction
Article / BlogPostingHigh — authorship + dates
AggregateRating / ReviewMedium — trust signals
HowToMedium — process extraction
ProductMedium — feature extraction
BreadcrumbListLow — navigation context
OpeningHoursSpecificationMedium — local queries
MedicalBusiness / ProfessionalServiceHigh — domain-specific entity

Step 4: Content Extractability Audit

For 3-5 key pages, assess:

CheckScore (0-2)Notes
Clear definition/answer in first paragraph0=absent, 1=partial, 2=strong
Self-contained answer blocksCan passages stand alone?
Statistics with cited sourcesSpecific numbers + attribution
Comparison tablesStructured data for vs-queries
FAQ with natural-language questionsQuestion phrasing matches search
Expert attributionAuthor name + credentials
"Last updated" date visibleFreshness signal
Heading structure matches queriesH2/H3 = how users search
Internal linking to related contentTopical authority signals
Citeable quotes / key statementsQuotable in 40-60 words

Scoring: 0 = not present, 1 = partially present, 2 = fully optimized

Step 5: AI Visibility Check

Search for the brand and key queries to assess current AI search presence:

  1. Brand search — Search for "[brand name]" and check what information surfaces
  2. Service/product queries — Search for key queries related to the business
  3. Competitor comparison — Who gets cited for these queries instead?
  4. Third-party presence — Check for Wikipedia, review sites, industry directories

For each query, note:

  • Does the brand appear in search results?
  • What competitors or alternatives are cited?
  • What third-party sources mention the brand?

Step 6: Entity & Knowledge Graph Assessment

Check the brand's presence as a recognized entity:

SourcePresent?Quality
Wikipedia pageAccurate? Current?
Wikidata entryComplete properties?
Google Knowledge PanelShows for brand search?
Google Business ProfileClaimed? Optimized?
Industry directoriesListed?
Review platformsActive profiles?
Social media profilesConsistent NAP?

Step 7: Competitive Gap Analysis

For the top 3-5 competitors:

  • Which competitors appear in AI-generated answers?
  • What content structure do they use that gets cited?
  • What schema markup do they have?
  • Where are they present that the audited site isn't?

Audit Report Format

Present the audit as a structured report with scores:

Report Header

# GEO Audit Report: [Website]
**Datum:** [date]
**URL:** [url]
**Totaalscore:** [X] / 10

Section Scores

Score each section on a 1-10 scale:

Audit OnderdeelScoreStatus
AI Bot Toegang/10[icon]
Schema Markup/10[icon]
Content Extractability/10[icon]
E-E-A-T Signalen/10[icon]
Entity Herkenning/10[icon]
AI Zichtbaarheid/10[icon]
Concurrentiepositie/10[icon]

Status icons: Use text indicators (Sterk / Voldoende / Onvoldoende / Kritiek)

Scoring Guidelines

AI Bot Toegang (0-10)

  • 10: All major AI bots explicitly allowed
  • 7: No blocks (default allow), but not explicitly configured
  • 4: Some AI bots blocked
  • 0: All AI bots blocked or robots.txt not accessible

Schema Markup (0-10)

  • 10: All relevant schema types present and complete (Organization, FAQPage, Article, domain-specific)
  • 7: Core schemas present (Organization, Article) but missing FAQPage or domain-specific
  • 4: Only basic schemas (WebPage, BreadcrumbList)
  • 0: No structured data

Content Extractability (0-10)

  • 10: All key pages have extractable answer blocks, stats with sources, proper heading structure
  • 7: Most pages have good structure but missing some elements
  • 4: Some structure but answer blocks not self-contained
  • 0: No structured content blocks, prose-only, no citations

E-E-A-T Signalen (0-10)

  • 10: Named experts with credentials, cited sources, demonstrable experience, trust signals
  • 7: Some author attribution and credentials, partial sourcing
  • 4: Generic content without clear expertise signals
  • 0: No author attribution, no credentials, no sourced claims

Entity Herkenning (0-10)

  • 10: Wikipedia + Wikidata + Knowledge Panel + complete third-party presence
  • 7: Some third-party presence (directories, review sites) but no Wikipedia
  • 4: Only own website and basic business listings
  • 0: No third-party presence, brand not recognized as entity

AI Zichtbaarheid (0-10)

  • 10: Brand cited in AI answers across multiple platforms for key queries
  • 7: Brand appears in some AI answers or via third-party citations
  • 4: Brand rarely mentioned, competitors dominate
  • 0: Brand not found in any AI-generated answers

Concurrentiepositie (0-10)

  • 10: Leading position vs competitors in AI citations
  • 7: Competitive, similar presence to competitors
  • 4: Behind competitors on most fronts
  • 0: Competitors dominate, no competitive presence

Findings Per Section

For each section, provide:

  1. Wat goed gaat — Positive findings
  2. Wat ontbreekt — Gaps and missing elements
  3. Aanbevelingen — Specific, actionable recommendations

Priority Action Plan

End with a prioritized action table:

#ActieImpactMoeitePrioriteit
1[specific action]Hoog/Middel/LaagLaag/Middel/HoogP1
2[specific action]......P1
3[specific action]......P2

Priority levels:

  • P1 — Do immediately (high impact, low effort)
  • P2 — Do this month (high impact, medium effort)
  • P3 — Plan for next quarter (medium impact, high effort)
  • P4 — Nice to have (low impact)

Language

Deliver the audit report in the same language as the user's request. Default to Dutch (Nederlands) if the website is Dutch. Use English for English websites or when the user communicates in English.


Important Notes

  • Always fetch and analyze the actual website — never audit from memory or assumptions
  • Check robots.txt for real bot access status
  • Search for the brand to check real AI visibility
  • Be specific in recommendations — "add FAQPage schema to /veelgestelde-vragen/" not "add schema markup"
  • Include the actual missing schema types with examples where helpful
  • Note both quick wins (low effort, high impact) and strategic improvements
  • If the site is in a regulated industry (healthcare, finance, legal), note YMYL considerations
  • Reference the Princeton GEO study findings where relevant to justify recommendations

Related Skills

  • ai-seo — For content optimization strategies after the audit
  • schema-markup — For implementing recommended structured data
  • seo-audit — For complementary traditional SEO audit
  • content-strategy — For content planning based on audit findings
  • competitor-alternatives — For building comparison content identified as gaps

Source Transparency

This detail page is rendered from real SKILL.md content. Trust labels are metadata-based hints, not a safety guarantee.

Related Skills

Related by shared tags or category signals.

Security

geo-audit

No summary provided by upstream source.

Repository SourceNeeds Review
Security

geo-audit

No summary provided by upstream source.

Repository SourceNeeds Review
Security

compliance-evidence-assembler

把审计所需证据整理成目录、清单和缺失项,便于后续评审。;use for compliance, evidence, audit workflows;do not use for 伪造证据, 替代正式审计结论.

Archived SourceRecently Updated
Security

skillguard-hardened

Security guard for OpenClaw skills, developed and maintained by rose北港(小红帽 / 猫猫帽帽). Audits installed or incoming skills with local rules plus Zenmux AI intent review, then recommends pass, warn, block, or quarantine.

Archived SourceRecently Updated