Index of every page on this site: llms.txt
All skills
factive1 avatar

/claude-code-seo-skill

@043cd04
Run it yourself

Write or edit a piece of content that ranks in Google, gets cited by ChatGPT, Perplexity, and AI Overviews, and reads like a person wrote it. Use when the user asks for a blog post, article, guide, comparison, listicle, landing page, or FAQ meant to rank in search or get cited by AI, or asks to humanize, de-slop, review, or refresh a specific draft or page. For auditing many pieces at once — a library, resource section, blog, or docs set — use content-library-audit instead.

  • 11 files
  • 115.2 KB
  • MIT
  • Updated 2 months ago
  • GitHub
Use this Skill: https://skilld.dev/gh/factive1/claude-code-seo-skill/claude-code-seo-skill

Nothing lands on disk. Nothing to clean up.

Fork this Skill

Edit a local copy. It keeps the author and licence.

SKILL.md

≈126 tokens for metadata: the name and description. ≈670 when used: this file. ≈25k more on demand in 8 files.

Description uses 6.2% of example budget

Before choosing a Skill, your Agent reads its name and description. All available Skills share that space.

  • A shorter description leaves more room for other Skills. This entry exceeds our 1% size suggestion.

In our Claude Code example, all Skill names and descriptions share 8,000 characters. This Skill uses ≈496 characters, or 6.2%.

The 1% threshold is a size suggestion. Longer descriptions can still fit.

Your model, settings, and other Skills decide how much text your Agent can read.

Example settings and source

The example uses a 200k-token context and default Claude Code settings. The count includes the name, description, separators, and when_to_use when present. Codex also counts local file paths.

Skit's source and limits: Codex 0.160.1, Claude Code 2.1.292.

SEO + GEO content

Read seo-geo-content-guide.md in this directory in full before writing or editing anything. It is the source of truth. This file only says how to apply it.

Before drafting

  1. Ask for or find the author's voice samples. A sample overrides every pattern rule in the guide, including the dash rule.
  2. Know the job: who the piece is for, where it will be published, and what the reader should do after reading it. Ask if the brief doesn't say.
  3. Run the research in the guide's SEO and GEO sections. Collect every statistic, quote, and source you plan to use, with its URL, before you write a sentence.

While drafting

Follow the four rules at the top of the guide. Never invent a fact. Match the voice samples. Every sentence adds something. Every sentence passes the portability test.

Put the keyword in the title, the H1, and one or two H2s. Open with what the reader doesn't already know from the query. Lead sections with a quotable statement. Use only sourced numbers, names, dates, and quotes, and leave [SOURCE NEEDED: ...] where you don't have one.

After drafting

Run the guide's rewrite pass before the checklist: mark every tell strongest first, rewrite around the point, search for the five survivors (not-X-but-Y, one-line closer, dash, triad, bold label), check the ending, verify no fact was added, dropped, or smoothed into a generic claim, then read it aloud. Then run the final gate in the implementation checklist. If any item fails, fix it and run the checks again.

Editing or auditing existing content

When the user asks whether a draft sounds like AI, name each pattern from the guide that appears, quote the line, and give the fix in a few words. Do not score the draft or guess whether AI wrote it. When the user asks for an edit, make the minimum effective edit: fix the tells, keep strong human sentences as they are, and list what changed. The writer should recognize the result as their own.

When it's a whole library, not a piece

This skill works on one piece at a time, and the standard holds when a person is applying it to one piece. Across a corpus it fails in ways a single-piece review can't see: writers certify their own work as clean, repetition is invisible from inside one file, and fabrication recurs during the rounds meant to fix it.

If the user points at a directory, a resource section, a blog, or more than a handful of pieces, use content-library-audit. It runs this guide as its standard and adds the verification the scale requires.

What to return

The full draft, every [SOURCE NEEDED] marker listed at the end, and a short note on what voice samples or sources were used.

Source: SKILL.md on GitHub

skilld matched fixed text patterns in SKILL.md and file names. Patterns miss obfuscated code.

skilld run checks every file with the same patterns. It asks for approval before it loads a Skill with a behavior marked Needs approval.

No third-party reports yet.

Signed by skilld at 043cd04. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub last month.

Steadyupdated 2 months ago
metadata
{
  "version": "2.1.0"
}

README badge

README badge for factive1/claude-code-seo-skill