last update
Apr 15, 2026
license
MITv1.0.0
format
SKILL.md
- Install
// Downloads SKILL.md into ~/.claude/skills/seo-content-optimizer/
$ curl -fsSL --create-dirs -o ~/.claude/skills/seo-content-optimizer/SKILL.md https://claudeskil.com/skills/seo-content-optimizer/SKILL.md - Invoke
// Inside Claude Code, type the skill name as a slash command
> /seo-content-optimizer help me ship this - Or just ask
// Claude loads the skill by itself when a request matches its description
> help me ship this
- name:
- SEO Content Optimizer
- slug:
- seo-content-optimizer
- version:
- v1.0.0
- license:
- MIT
- author:
- @chiragthummar16
- categories:
- tags:
- #seo#schema#jsonld#content
- description:
Audit pages for on-page SEO, write schema JSON-LD, and generate meta tags that actually rank.
README.md
SEO Content Optimizer
A skill for shipping pages that pass an on-page SEO review before they go live. It runs a structured audit against a single URL or a local HTML/Markdown file, rewrites titles and meta descriptions to fit pixel/character limits, emits valid schema.org JSON-LD, and hands back a prioritized internal linking plan. Built for content engineers, technical writers, and marketers who edit pages in a repo rather than a CMS dashboard.
When to use it
- You drafted a blog post or landing page and want a pre-publish on-page check.
- A page ranks on page 2 and you need concrete fixes, not a vague "improve content" note.
- You're adding Article, FAQPage, or HowTo rich results and want JSON-LD that validates.
- You're auditing an existing site section and need an internal linking map across related URLs.
What it does
The audit walks the page and reports on each lever: title tag length and keyword placement, meta description, a single H1, heading hierarchy (no skipped levels), image alt coverage, canonical tag, and Open Graph / Twitter Card presence. Readability is scored with Flesch Reading Ease plus average sentence and paragraph length, flagging passages above grade target.
For structured data it generates JSON-LD blocks for Article (headline, author, datePublished, image), FAQPage (question/answer pairs pulled from the content), and HowTo (ordered steps). Output is plain <script type="application/ld+json"> you paste into <head>, shaped to pass Google's Rich Results Test.
The linking plan groups your supplied pages by topical cluster and suggests anchor text plus source→target pairs, surfacing orphan pages with no inbound links.
Usage
curl -fsSL --create-dirs -o ~/.claude/skills/seo-content-optimizer/SKILL.md https://claudeskil.com/skills/seo-content-optimizer/SKILL.md
# audit a live page against a target query
> audit https://example.com/blog/cold-brew for "cold brew ratio"
# generate FAQ + Article JSON-LD from a local file
> write Article and FAQ JSON-LD for ./content/cold-brew.md
# build a linking plan across a folder
> internal linking plan for the URLs in ./sitemap-blog.txt
The skill returns the rewritten title/meta inline, JSON-LD ready to paste, and a table of link suggestions.
Tips & gotchas
- Titles are scored by pixel width (~580px desktop), not just character count — a 58-char title with wide caps can still truncate.
- FAQ JSON-LD must mirror visible on-page Q&A. Don't inject answers that aren't in the rendered HTML, or you risk a manual action.
- Provide the target keyword explicitly; without it the audit scores structure but can't judge intent match or keyword placement.
- For SPA pages, give the rendered HTML or a prerendered URL — the audit reads markup, not post-hydration DOM.
FAQ
Does it crawl my whole site? No. It works one page at a time; the linking plan operates over a list of URLs you supply.
Will the JSON-LD validate? It's structured to pass Rich Results Test, but always paste the final block in and confirm there, since required fields vary by content.
Can it write new content? It rewrites titles, metas, and schema, and flags thin or hard-to-read sections — it audits and edits rather than ghostwriting whole articles.
$ cat reviews/