Skip to content
cccskills
FreePublished on GitHub

content-seo

Copy + discoverability requirements every delivered app must meet — real-brand extraction, anti-slop voice, per-page SEO/JSON-LD, pSEO, GEO/AI-search, truth/citations, competitor floor, i18n, trust surfaces.

The English summary is being prepared. Showing the original description.

See how to install

Files included (1)

  • SKILL.md3.0 KB

SKILL.md (original)

Review the instructions the agent will receive before you install.

Content & SEO — every app's copy + discoverability must satisfy

Brand extraction (rebuilds / clients)

  • Extract the REAL brand (logo, colors, fonts, tone; Wayback if the source is down). NEVER invent colors, never use Emdash defaults for clients, never infer from category.
  • Colors derive from the logo (hue-distance verified). Logo luminance + source-site theme decide light/dark. Every logo render contrasts its container ≥4.5:1. Exactly ONE logo asset per container.

Anti-AI-slop (mandatory scan)

  • Banned words: delve, leverage, unleash, revolutionize, seamless, robust, synergy, elevate, empower, transformative, best-in-class, cutting-edge, discover, innovative.
  • Banned patterns: "Welcome to", "Discover [product]", vague aspirational headlines, hedging ("may help"), generic superlatives. Test: "Would the founder actually say this?" — no → rewrite.

Copy

  • Headlines benefit-first, specific, ≤8 words. Subheads 15-25 words. Body one idea/paragraph, active, concrete. Max sentence 25 words. Flesch ≥60 (microcopy ≥70).
  • CTAs: action-verb-first, ≤3 words preferred; never "Click here" / "Submit" / "Learn more". Above fold + page end.
  • Microcopy: errors = [what happened] + [what to do]; empty states suggest the first action; toasts past-tense success / present in-progress.

SEO — every page

  • Title 50-60 chars keyphrase-first · meta description 120-156 · canonical · OG image 1200×630 · exactly one H1 in prerendered HTML.
  • JSON-LD: WebPage floor; Org/BreadcrumbList/FAQPage/Person/Product/Service ONLY when the entity is real — never pad (FAQPage only with real FAQs).
  • sitemap.xml (+lastmod) · robots.txt · llms.txt · humans.txt · security.txt.

pSEO (when used)

  • 5 page types: integration · comparison · use-case · template · location. Each: unique H1 + meta + 800+ unique words + ≥1 unique image + ≥3 internal links + ≥1 outbound citation. Never templatize verbatim.

GEO / AI search

  • Quotable answer blocks (40-60 words). FAQPage schema. Every JSON-LD fact also appears as visible HTML. Lead paragraph answers the query in <40 words. EEAT: author bio + Person schema + sameAs + dated revisions + ownership.

Truth & citations

  • Every factual claim APA-cited to a real source. Never fabricate people, stats, timelines, or testimonials. Testimonials carry verifiable attribution.

Reach

  • Competitor floor: outscore the top 5-10 peer sites on every rubric dimension by ≥15% before "done".
  • i18n by demographics: any community ≥10% local share (ACS B16001) gets a /{locale}/* mirror + hreflang.

Required pages

  • /privacy · /terms · /accessibility · About/Team with real bios.

Reviews

No reviews yet. Tell others what you thought.

Comments

No login needed. Share how it worked for you, ask a question, or note where it did not run.

No comments yet. Be the first to write one.

Comments are published right away. Do not include personal information such as email addresses or phone numbers. Up to 2 URLs.

0 / 1,000 characters

This form is protected by reCAPTCHA, and the Google Privacy Policy and Terms of Service apply.

More skills from this repository

Every delivered app — the non-inferable business requirements + exact stack/brand choices every build must satisfy. Loads on every app/site build.

The English summary is being prepared. Showing the original description.

heymegabyte/agent-skills24Updated Oct 11, 2026

Cloudflare-first platform selection. Decision trees for Workers, D1, R2, KV, DO, Queues, Vectorize, Containers, Sandboxes, Flagship, Agent Memory, Workflows v2. Default stack, override conditions, auth, data patterns, reliability.

The English summary is being prepared. Showing the original description.

heymegabyte/agent-skills24Updated Oct 11, 2026

Extract real brands (Wayback for rebuilds). Copy system, headline/CTA rules, trust surfaces, legal pages, SEO+structured data, anti-AI-slop, microcopy, DESIGN.md, W3C DTCG tokens, pSEO 5 types, GEO/AI search.

The English summary is being prepared. Showing the original description.

heymegabyte/agent-skills24Updated Oct 11, 2026

Implements features in vertical slices, always starting with homepage. Enforces anti-placeholder rules — no lorem ipsum, no TODO stubs, no gray boxes. Real content, real images, real interactions. TypeScript strict mode, Zod validation, and structured file organization.

The English summary is being prepared. Showing the original description.

heymegabyte/agent-skills24Updated Oct 11, 2026

The consolidated, world-class Cloudflare + integrations reference — CF platform/data/AI-edge patterns, media pipeline, observability/growth, payments, and the site-generation pipeline. The non-inferable choices that build the greatest web apps.

The English summary is being prepared. Showing the original description.

heymegabyte/agent-skills24Updated Oct 11, 2026

One-line-prompt → cinematic, gorgeous, functional, well-tested, deployed website. Pre-hydrated SPA + full PWA kit + JSON-LD rich snippets + third-party integrations. React 19+Vite default (Angular optional). 100 concrete improvements grouped into 10 categories that EVERY single-prompt site build must satisfy before being marked done. Trigger whenever the user says 'make a website for X', 'build a site for X', 'rebuild X.com', or any equivalent one-liner.

The English summary is being prepared. Showing the original description.

heymegabyte/agent-skills24Updated Oct 11, 2026

See all skills by heymegabyte

Report a problem with this skill