Skip to content
SkillModel-led3,032 tokens in use

seo-content

Score a page's content for E-E-A-T, readability, thinness and AI citation readiness, and clean drafts of AI-typical phrasing and invisible watermark characters.

31,135 stars on the repository MIT LicenseAgriciDaniel/claude-seo · skills/seo-content
/plugin marketplace add AgriciDaniel/claude-seo /plugin install claude-seo@agricidaniel-claude-seo
View source

Install command as written in the repository's README

Skill label

measured from its files

What installing seo-content costs

53tokens idle
3,032tokens in use

Idle is what sits in context while the skill is installed. In use is its instructions, loaded when it runs. Counts are approximate, made offline with cl100k_base.

Model dependence Model-led

Script-ledMixedModel-led

Mostly instructions, so how well it works depends on the model that follows them.

This describes the skill's structure, how much of it is scripts. It does not say how well it works.

Touches

shelldetectedwrites filesdetected

Declared tags come from the skill's frontmatter; detected ones from its scripts, its code blocks and the commands it names in inline code. This is detected, not audited.

How it scores

Quality

How well the skill is written, counted from its own files: its frontmatter, its steps, examples, stated output and constraints, its size and whether every file it ships is one it names.How we score it →
Good
  • Frontmatter11/15
  • Structure15/25
  • Examples12/15
  • Output15/15
  • Size10/15
  • Resources10/15

Security

Starts at 100 and loses points for each thing in the skill's files that a reader should look at first.How we score it →
Excellent
  • Runs shell commands−5

    does this on the machine, and the frontmatter does not say so

    SKILL.md, line 146
  • Writes files−8

    does this on the machine, and the frontmatter does not say so

    SKILL.md, line 214

Repository health

Pre-computed score based on development activity, maintenance, community, maturity, and trend momentum.How we score it →
Excellent
  • Development activity96
  • Maintenance100
  • Community72
  • Maturity32
  • Momentum40

seo-content evaluates page content only, leaving technical checks to other skills. It starts with Google’s Who, How and Why test, then scores E-E-A-T across experience, expertise, authoritativeness and trust with the weights of its own model, and reviews word count floors, readability, keyword use, structure, media, links and freshness.

It also assesses AI-generated content patterns and AI citation readiness, checks for hidden text, and runs a cleanup script on your own drafts to strip zero-width characters and swap AI-typical phrases.

What You Get

  • A Content Quality Score with an E-E-A-T breakdown across four factors
  • An AI Citation Readiness score with quotability and structure findings
  • Page-type word count floors presented as coverage guides, not ranking targets
  • A hidden-text spam signal from the HTML parser
  • A cleaned Markdown draft with a JSON log of removed characters and swapped phrases

Common Use Cases

  • Review a blog post for experience and trust signals before publishing
  • Find thin or gated pages that need more depth
  • Strip zero-width characters and stock AI phrasing from your own draft
  • Check a traffic drop against the list of confirmed Google updates

What Makes It Unique

It states its own limits: word counts are coverage floors, readability scores are not ranking factors, its E-E-A-T weights are an internal model that puts trust first, and its scores are heuristics. The cleanup script is limited to editing your own drafts and declines to strip provenance from content you do not own.

Under The Hood

The skill reads shared E-E-A-T reference files from the plugin and runs shell commands that start Python scripts through the claude-seo launcher. One parses a page fetched over the network and reports hidden text, one lists Google updates from a bundled ledger, and one reads a draft from a file or standard input and writes a cleaned copy to disk. Cleanup is two fixed passes, invisible characters and a table of phrase swaps, with nothing paraphrased. It can use DataForSEO tools when present. Fetched content is treated as untrusted data, and the frontmatter does not declare the shell, network or file access.

Join founders building with open source

Opinionated takes, migration guides, cost-saving tips, and insights from the open source ecosystem.

Subscribe on SubstackJoin 750+ subscribers