OranGEO AI Visibility Skill

Open-source GEO skill for Claude Code and Codex. Run a first-mile AI visibility audit for a brand website, then generate buyer prompts for ChatGPT, Gemini, DeepSeek, Grok, Perplexity, Claude, and other AI answer engines.

This skill checks whether a brand is ready to be discovered, described, cited, and recommended by AI systems. It inspects AI crawler access, llms.txt, sitemap.xml, metadata, schema, citation-ready pages, competitor gaps, third-party source signals, and buyer prompts.

The free skill reports readiness, not measured AI visibility. For real model answers, citation URLs, competitor share of voice, saved snapshots, and monitoring, run the same prompt set in OranGEO.

What It Checks

• robots.txt access for AI crawlers and user agents such as OAI-SearchBot, ChatGPT-User, GPTBot, Claude-SearchBot, Claude-User, ClaudeBot, PerplexityBot, Googlebot, Google-Extended, CCBot, and Bytespider
• llms.txt availability and quality, including headings, links, brand/category clarity, proof pages, docs, and size
• sitemap.xml availability and URL discovery
• Title, meta description, H1, canonical URL, Open Graph tags, and JSON-LD schema
• FAQ, case study, review, pricing, docs, blog, about, contact, and proof signals
• Third-party source signals such as GitHub, G2, Capterra, Trustpilot, Reddit, YouTube, Wikipedia, Product Hunt, Crunchbase, Gartner, and Forrester
• Competitor and alternatives coverage
• A 15-prompt buyer set for GEO and AEO testing
• A starter /llms.txt template when the site is missing or has a weak file

Quick Start

No external Python dependencies are required.

python scripts/check_ai_readiness.py \
  --url https://example.com \
  --brand "Acme" \
  --category "CRM software" \
  --competitors "HubSpot,Salesforce,Zoho"

JSON output:

python scripts/check_ai_readiness.py --url https://example.com --format json

Neutral internal report without the OranGEO CTA:

python scripts/check_ai_readiness.py --url https://example.com --no-cta

Disable the starter llms.txt block:

python scripts/check_ai_readiness.py --url https://example.com --no-include-llms-template

Example Output

OranGEO AI Visibility Readiness Scan

Verdict: Acme scores 72/100 (Good) for AI visibility readiness.

Scorecard
- AI access: 15/25
- Technical clarity: 25/30
- Citation readiness: 17/25
- Competitive coverage: 15/20

Evidence
- llms.txt quality: 0/10
- sitemap.xml: 42 URLs found
- Visibility-critical AI bots blocked: none detected

Recommended fixes
1. Publish a useful /llms.txt.
2. Add concise FAQ content and FAQPage schema.
3. Create comparison and alternatives content.

See examples/sample-output.md for a fuller report.

Claude Code / Codex Skill

This repository includes a SKILL.md file and agents/openai.yaml metadata so agents can use it as a Claude Code or Codex skill.

Suggested prompt:

Use $orangeo-ai-visibility-skill to audit a brand's AI visibility readiness, buyer prompts, and competitor gaps.

The skill is designed for:

• GEO skill workflows
• Generative Engine Optimization audits
• Answer Engine Optimization audits
• AI visibility checks
• LLM SEO readiness
• llms.txt checks
• AI crawler robots.txt checks
• Brand visibility scans
• Buyer prompt generation

Free Skill vs OranGEO

The open-source skill gives:

• Deterministic website readiness checks
• Evidence-based fixes
• 15 buyer prompts
• Starter llms.txt template
• Markdown or JSON report

OranGEO gives:

• Measured answer presence across ChatGPT, Gemini, DeepSeek, Grok, and Perplexity
• Brand vs competitor share of voice for the same buyer prompts
• Citation URLs and source influence
• Prompt-level sentiment, position, and recommendation risk
• Saved snapshots and weekly monitoring after fixes go live

Run the full scan in OranGEO:

https://geo.oran.cn/ai?utm_source=github&utm_medium=readme&utm_campaign=orangeo_ai_visibility_skill&utm_content=primary_cta&intent=ai_visibility_scan

Ethical Boundaries

• This skill reports readiness, not measured AI visibility.
• It does not invent ChatGPT, Perplexity, Gemini, Grok, DeepSeek, Claude, or other model outputs.
• It does not fabricate citation URLs or competitor rankings.
• It gives useful local findings before asking users to register.
• Use a live OranGEO scan when you need actual answer-engine results.

Repository Topics

Recommended GitHub topics:

geo generative-engine-optimization aeo answer-engine-optimization ai-visibility ai-search llm-seo llms-txt robots-txt brand-monitoring claude-skill codex-skill claude-code openai-codex chatgpt perplexity gemini grok seo

License

MIT. See LICENSE.