Best Claude Code Skills for SEO: 5 Libraries Compared (2026)

TL;DR: Claude Code skills are reusable SKILL.md instruction folders that turn one-off SEO prompts into repeatable workflows, and in November 2026 claude-seo leads the field with about 26,500 GitHub stars, 26 sub-skills and 19 agents.

Pick claude-seo for full audits, geo-seo-claude for AI search, and a content-focused library for writing.

We run claude-seo and the SearchFit SEO plugin in our own Claude Code setup.

The catch: these libraries change fast, so star counts and commands here will date within months.

Claude Code skills are reusable instruction folders that turn one-off SEO prompts into standardised, repeatable workflows.

Several open-source SEO skill libraries now exist, each covering different parts of technical audits, content, schema and GEO/AEO readiness.

This guide compares the five most useful ones as of November 2026, with install commands, what each does well, and which suits each SEO role.

The ecosystem has grown fast.

The most popular library, claude-seo by AgriciDaniel, had about 26,500 GitHub stars and 3,400 forks in November 2026, up from a few thousand at the start of the year.

What Are Claude Code Skills and Why Do They Matter for SEO?

A Claude Code skill is a folder containing a SKILL.md file, plus optional scripts and reference files, that defines how to do a specific task.

The SKILL.md starts with a short frontmatter block: a name and a description that tells Claude when to use it.

Claude loads the skill when your request matches that description, or you run it directly with /skill-name.

The format follows the open Agent Skills standard, so many skills also work in Codex, Cursor and other agents.

For SEO, skills are the difference between different results every time you prompt and the same method every time.

Without a skill, you type: "Do an SEO audit of this page." Claude improvises.

Sometimes it checks schema, sometimes it doesn't, and the output format changes each time.

With a skill, you type /seo audit https://example.com and Claude follows a defined method: specialist agents check technical SEO, content, schema, performance and more in parallel, then the findings are combined into one scored, prioritised report.

Same method, same format, easy to compare and share.

Where skills live: ~/.claude/skills/<name>/ for all your projects, or .claude/skills/<name>/ inside a project to share it with your team through the repository.

Most SEO libraries now also ship as plugins, which bundle skills, agents, hooks and MCP servers and install through /plugin.

Plugin skills are namespaced, so they don't clash with your own.

Skills can call other skills, use MCP servers for live data, and write reports as Markdown, PDF or spreadsheets.

The 5 Claude Code SEO Skill Libraries Compared

Here are the main Claude Code SEO skill libraries as of November 2026.

Star counts are from GitHub on 4 November 2026 and move quickly.

Library 1: claude-seo by AgriciDaniel (The Full-Stack Auditor)

This is the most complete and most popular SEO skill library.

Version 3 is built from 26 sub-skills and 19 specialist agents, with up to 17 agents working on a site at once.

It covers technical SEO, content and E-E-A-T, schema, AI search readiness, local SEO, links, ecommerce and reporting, through about 30 commands (claude-seo on GitHub).

Installation (plugin, recommended):

/plugin marketplace add AgriciDaniel/claude-seo
/plugin install claude-seo@agricidaniel-claude-seo
/seo setup

The /seo setup step prepares the browser tooling the audits use.

A script install is also available:

git clone --depth 1 https://github.com/AgriciDaniel/claude-seo.git
bash claude-seo/install.sh

Key commands:

/seo audit [url] runs the full multi-agent audit with a prioritised plan.

/seo page [url] analyses one page in depth.

/seo geo [url] checks readiness for AI search.

/seo schema [url] checks existing structured data and generates corrected JSON-LD.

/seo local [url] covers local SEO and Google Business Profile.

Each recommendation explains why it matters and how you will know it worked.

Optional extensions connect live data sources such as DataForSEO for SERP and keyword data.

What to watch: it needs Python 3.11+ for its scripts, it is designed mainly for Claude Code, and a full audit with many parallel agents uses a lot of your plan's usage.

Best for: technical SEOs and agencies that need full site audits with parallel agents and client-ready reports.

Library 2: geo-seo-claude by zubair-trabzada (The GEO Specialist)

This library is built for Generative Engine Optimization first.

It has 13 sub-skills and 5 parallel agents, with about 11,500 GitHub stars in November 2026 (geo-seo-claude on GitHub).

Its skills cover AI citation readiness, AI crawler access in robots.txt, llms.txt, brand mention analysis across platforms such as YouTube, Reddit and Wikipedia, schema for AI extraction, and platform-specific advice for Google AI Overviews, ChatGPT, Perplexity and Copilot.

Installation:

curl -fsSL https://raw.githubusercontent.com/zubair-trabzada/geo-seo-claude/main/install.sh | bash

On Windows, run it in Git Bash rather than PowerShell.

Run /geo audit <url> for a full GEO and SEO audit with parallel agents, which produces a 0 to 100 score and client-ready PDF reports.

Its brand mention focus matches independent research: Ahrefs' study of about 75,000 brands found branded web mentions correlated more strongly with AI visibility than backlinks did.

What to watch: traditional SEO coverage is lighter than claude-seo, so most teams pair the two.

Best for: GEO/AEO specialists and agencies building AI visibility services.

Use our GEO/AEO score checker alongside it for a second opinion on AI readiness.

Library 3: Agentic-SEO-Skill by Bhanunamikaze (The Cross-Platform Option)

This library began as a restructured fork of claude-seo and is now packaged for many agents: Claude Code, Codex, Antigravity, Cursor, Windsurf, Continue, Copilot and Cline (Agentic-SEO-Skill on GitHub).

It has 16 sub-skills, 10 specialist agents and 89 scripts that collect evidence and automate workflows.

Installation:

curl -fsSL https://raw.githubusercontent.com/Bhanunamikaze/Agentic-SEO-Skill/main/install.sh | bash -s -- --online --target claude

Change --target to codex, global or all for other agents.

Its standout feature is confidence labelling: every finding is marked Confirmed (directly verified), Likely (strong evidence) or Hypothesis (a pattern worth checking).

That stops uncertain findings being presented with the same confidence as verified ones.

It also includes GitHub SEO workflows that write GITHUB-SEO-REPORT.md and GITHUB-ACTION-PLAN.md for repository optimisation.

Best for: teams that use several coding agents and want one SEO skill set across them.

Library 4: aaron-marketing SEO/GEO skills by aaron-he-zhu (The Quality Frameworks)

This library takes a modular approach built around two scoring frameworks: CORE-EEAT for content quality and CITE for domain authority.

Its original repository, seo-geo-claude-skills, is frozen at version 9.9.12 with 20 skills; development moved in 2026 to the aaron-marketing-skills bundle, where 16 SEO/GEO skills sit alongside other marketing disciplines.

Existing installs keep working, but new installs come from the bundle.

Installation:

# Claude Code
/plugin marketplace add aaron-he-zhu/aaron-marketing-skills
/plugin install aaron-marketing@aaron

# Codex, Cursor, Gemini CLI and other agents
npx skills add aaron-he-zhu/aaron-marketing-skills

The skills cover research, building, optimisation and monitoring, and the entry command is /aaron-marketing:seo-geo.

The CORE-EEAT benchmark scores content on clarity, originality, rigour and E-E-A-T signals such as first-hand experience and transparent methodology, which makes it useful as an editing checklist even if you write by hand.

What to watch: the move to a larger bundle means old tutorials and commands for seo-geo-claude-skills are out of date.

Best for: content teams that want a rigorous quality gate before publishing.

Library 5: SearchFit SEO (Claude Code Plugin)

SearchFit SEO is a Claude Code plugin that installs from a plugin marketplace.

It adds 11 skills, including SEO audit, technical SEO, on-page SEO, schema markup, internal linking, broken links, keyword clustering, content briefs, content strategy and AI visibility, plus 3 agents: an SEO auditor, a content strategist and a competitor analyser.

Commands such as /seo-check, /generate-schema, /keyword-cluster and /create-content run specific jobs.

We use it in our own Claude Code setup alongside claude-seo.

The differentiator is auto-activation. You don't need to remember commands: ask naturally, such as "audit example.com for SEO issues," and the matching skill loads.

That makes it the most beginner-friendly option.

What to watch: its audits are lighter than claude-seo's multi-agent audits, and it has fewer schema types.

Best for: SEOs who want everyday helpers without learning a command set.

How to Choose: Decision Framework

The choice depends on three things: how technical you are, your main workflow (audits, content or GEO), and whether you use agents other than Claude Code.

Technical SEO running audits: install claude-seo.

Add geo-seo-claude if you also do GEO work.

Content SEO writing and optimising articles: install the aaron-marketing SEO/GEO skills for the CORE-EEAT quality gate.

Agency selling GEO services: install geo-seo-claude for GEO scoring and client PDFs, plus claude-seo for traditional audit deliverables.

Using Codex or Cursor as well as Claude Code: start with Agentic-SEO-Skill or the aaron-marketing bundle, which install across many agents.

Want the gentlest start: install SearchFit SEO and ask for what you need in plain language.

The power move: claude-seo for audits plus a content library for writing.

Plugin skills are namespaced, so they don't conflict.

Connect Ahrefs or Semrush through MCP and you have live data, audit methods and content checks in one terminal.

Building Your Own SEO Skills

All the open-source libraries above use permissive licences, so you can fork them or use them as templates.

The smallest working skill is one SKILL.md file in ~/.claude/skills/your-skill-name/:

---
name: your-seo-skill
description: What this skill does and when Claude should use it.
---

# Skill Name

## Workflow
1. What to check first
2. What to analyse
3. How to format the output

## Quality gates
- Checks that must pass before reporting
- What to do when data is missing

## Output format
- Report structure
- File naming convention

Write the description carefully, because Claude uses it to decide when to load the skill.

Keep SKILL.md under about 500 lines and move long reference material into separate files in the same folder.

Add disable-model-invocation: true to the frontmatter for skills that should only run when you call them, such as anything that publishes or deletes.

At SEOShouts we use custom skills for article production checks, alongside our own internal link checker, on-page analyser and schema generator.

The lesson from using these libraries: the method in a skill matters more than its code.

A clear SKILL.md with specific checklists, quality gates and a defined output format beats a complex skill with vague instructions.

Start simple, test the results, and iterate.

Once you have a few skills, build your own SEO agent that uses these skill libraries.

Frequently Asked Questions

What are Claude Code skills for SEO?

Claude Code skills are folders with a SKILL.md file that define a repeatable method for an SEO task: what to check, how to format the output and what quality standard to apply.

The biggest open-source libraries are claude-seo (about 26,500 stars, 26 sub-skills) and geo-seo-claude (about 11,500 stars, 13 GEO skills).

Which Claude Code SEO skill library is the best?

For full technical audits: claude-seo.

For GEO and AI search: geo-seo-claude.

For content quality scoring: the aaron-marketing SEO/GEO skills with CORE-EEAT.

Most professionals install two that complement each other.

How do I install Claude Code SEO skills?

Most libraries install as plugins: for claude-seo, run /plugin marketplace add AgriciDaniel/claude-seo, then /plugin install claude-seo@agricidaniel-claude-seo and /seo setup.

Standalone skills go in ~/.claude/skills/ or a project's .claude/skills/ folder.

Claude Code picks up new skills without a restart.

Do Claude Code SEO skills work with MCP servers?

Yes.

Skills can use any MCP server you have connected, and some libraries ship extensions for specific data sources such as DataForSEO.

See the MCP servers for SEO guide for setup.

Can I use these skills with Cursor or Codex?

Often, yes.

Skills follow the open Agent Skills format, so plain SKILL.md skills are portable.

Agentic-SEO-Skill and the aaron-marketing bundle ship installers for Codex, Cursor and other agents; claude-seo and geo-seo-claude are built mainly for Claude Code.

Are Claude Code SEO skills free?

The libraries in this guide are free and open source.

You need a paid Claude plan to run Claude Code, and optional data extensions may need paid subscriptions such as Ahrefs, Semrush or DataForSEO.

How do these skills differ from paid SEO tools?

Paid tools provide proprietary data such as keyword databases and backlink indexes.

Skills provide method: what to check, how to interpret it and how to prioritise.

They complement each other: connect a data source through MCP and use skills for the analysis.