How to use the geo-content agent skill for E-E-A-T review

What the geo-content skill makes the agent do

E-E-A-T stands for Experience, Expertise, Authoritativeness and Trustworthiness. The skill scores each of the four dimensions at 25 points, for 100 in total. It looks for signals such as author credentials, cited sources, contact details, privacy policy and terms, affiliate or sponsorship disclosures, and first-hand evidence like first-person accounts or original data.

The skill also looks at content structure. It assesses word count against page-type benchmarks, a Flesch Reading Ease target of 60 to 70, and whether each paragraph can stand alone so AI platforms can extract it. The output is a written report named GEO-CONTENT-ANALYSIS.md with scores and suggested weaknesses to fix.

• Experience: first-person accounts, original research, case studies, evidence of direct use
• Expertise: author credentials, technical depth, methodology, data-backed claims
• Authoritativeness: inbound citations, press mentions, speaker credentials, topical coverage
• Trustworthiness: contact information, privacy policy, HTTPS, corrections policy, disclosures, accurate claims

Install the geo-content skill

Download the ZIP for the skill from the directory page and unzip it. For a personal setup, place the unzipped skill folder inside ~/.claude/skills/. For a single project, place it inside that project's .claude/skills/ folder instead.

If you use the Claude apps, you can instead upload the skill under Settings → Capabilities → Skills. The skill's only file is SKILL.md, so check that it sits inside the skill folder after unzipping.

• Requires an agent that supports SKILL.md skills
• Requires the ability to fetch web pages (WebFetch) and write files (Write)

How to trigger a review with natural requests

You do not need a special command. Describe the page you want reviewed and ask for an E-E-A-T or AI citability assessment. Give a full URL where you can, and name the page type (homepage, blog post, service page) so the agent applies the right word count benchmark.

The skill's own workflow is to fetch the target pages, score E-E-A-T across the four dimensions, assess content quality and AI content signals, evaluate topical authority across the site, and then write GEO-CONTENT-ANALYSIS.md.

• Review https://example.com/blog/post-name for E-E-A-T and AI citability
• Audit our homepage and service pages for trust signals and write the report
• Check whether this article has the author credentials and sources AI search tools would need to cite it

Safety, network and data notes

The skill uses the network. It fetches the pages you name, and those pages are sent through the agent's model context. Do not use it on pages you cannot share with an AI tool.

The directory lists this skill as not running scripts. The skill's allowed-tools list does include Bash, so review the permissions your agent grants before running it on sensitive material.

The skill checks signals of trust, not whether facts are true. A well-presented page with wrong facts can still score well. Legal, medical or financial compliance review needs a qualified human reviewer.

Tips for getting useful results

Review the page types separately. Word count benchmarks differ between homepages, blog posts, product pages and pillar guides, so a short homepage is not penalised for being short the way a long-form guide would be.

Treat the score as a to-do list. The most useful part of the report is the list of weak signals, such as missing author bios or undisclosed affiliate links, which you can fix directly.

• Ask for the report on a specific page first, then on the wider site for topical authority
• Check any claimed statistic or medical, legal or financial point against primary sources yourself