How to use the GEO Audit Orchestrator agent skill (geo-audit)

What is an agent skill and what does geo-audit do?

An agent skill is a folder containing a SKILL.md file that tells an AI agent how to carry out a specific task. When the skill is installed, the agent can follow its workflow when a request matches the skill's description.

Given a website URL, geo-audit has the agent fetch the homepage and sitemap, sample up to 50 pages, and collect data such as titles, meta descriptions, headings, word counts, schema.org types, alt text and Open Graph tags. The skill then describes delegating analysis to five sub-tasks: AI citability, platform readiness, technical GEO infrastructure, content E-E-A-T quality, and schema and structured data.

• Composite GEO Score from 0 to 100, a weighted average of six categories.
• Score bands from Excellent (90-100) to Critical (0-39).
• Issues sorted by severity: Critical, High, Medium and Low, each with a suggested fix timeframe.
• Weights: AI Citability 25%, Brand Authority 20%, Content E-E-A-T 20%, Technical GEO 15%, Schema 10%, Platform Optimization 10%.

How do you install the geo-audit skill?

Download the ZIP of the skill from this page and unzip it. The skill's file is SKILL.md, so the unzipped folder should contain that file.

For a personal install, place the unzipped folder in ~/.claude/skills/. To limit the skill to one project, place it in that project's .claude/skills/ folder instead.

You can also upload the skill in the Claude apps under Settings → Capabilities → Skills.

• Personal: ~/.claude/skills/
• Single project: the project's .claude/skills/ folder
• Claude apps: Settings → Capabilities → Skills, then upload

How do you trigger the audit with a natural request?

You do not need a special command. Describe the job in plain language and include the website address. The agent matches your request against the skill's description, which covers GEO, AI visibility and full site audits.

Requests that mention a URL work best, for example asking for a GEO audit of a named domain, or asking how visible a site is to AI search engines.

• "Run a GEO audit on https://example.com and give me a prioritized fix list."
• "How well is my site set up to be cited by ChatGPT and Perplexity? Audit example.com."
• "Check example.com for AI crawler access and schema problems, and score it out of 100."

What does the audit output look like?

The skill asks for a composite GEO score and a prioritized action plan. Each issue is grouped by severity, so you can see what to fix first. The source describes the score as a weighted combination of category scores, and the severity list gives example issues for each level.

Examples of Critical issues include all AI crawlers blocked in robots.txt and a domain-level noindex directive. High issues include a missing llms.txt file and missing Organization or LocalBusiness schema. Medium and Low issues cover items such as partial crawler blocking, missing FAQ schema and images without alt text.

What does the skill need and what are the safety notes?

The skill needs an agent with file read and write tools and web fetch, specifically Read, Grep, Glob, Bash, WebFetch and Write as listed in its allowed-tools. It runs scripts and uses the network, so review what it will do before running it on a site.

The skill fetches pages across the domain. Its instructions say to respect robots.txt and to limit crawling to 50 pages with a 30-second timeout per fetch. Only audit sites you own or have permission to crawl.

The skill checks security headers only as a side item. It is not a penetration test or vulnerability scan.

• Only audit sites you own or are permitted to crawl.
• Scores are based on the agent's judgement; the file does not show how each sub-score is computed.
• Treat the output as a structured first pass, not a verified measurement.

What tips will get better results?

Give the agent the exact homepage URL, including the protocol. Tell it the business type if you know it, because the skill classifies sites as SaaS, Local Business, E-commerce, Publisher, Agency or Hybrid and uses those signals during discovery.

Read the critical and high items first, then check the medium and low items against your own priorities. Re-run the audit after changes to see how the findings move.

• Use the full homepage URL.
• State the business type if you know it.
• Start with Critical and High severity items.