What does this skill do, and when should you use it?
SERP Markup Builder handles the SEO/GEO assets that live in a page’s document head. It produces title and meta description options, Open Graph and Twitter Card tags, and Schema.org JSON-LD for FAQ, HowTo, Article, Product, and LocalBusiness candidates. It has separate meta and schema modes and can run both for a complete SERP package. It does not write body copy, diagnose crawl or indexing issues, or issue a publish-readiness verdict.
Reads the page URL, page type and intent, target keywords, audience, CTA, value proposition, current tags or markup, entity inputs, and quality constraints. It can use WebFetch for page information and a bundled Python pre-flight for server-side JSON-LD inspection. Meta mode returns three title options, three description options, an OG/Twitter/canonical/robots package, and CTR analysis; schema mode returns JSON-LD, placement guidance, validation steps, rich-result eligibility notes, and visible-content alignment notes. It can prepare a reusable handoff for memory/content/ and promote missing evidence or publish blockers to shared memory.
- An SEO specialist needs three title and description variants for an article and a paste-ready social metadata block.
- A content team needs FAQ JSON-LD with implementation and validation guidance.
- An ecommerce team is preparing Product structured data for a product page.
- A local business needs LocalBusiness markup based on facts visible on its location page.
- A website team wants to run both metadata and schema work as one SERP package.
How do you install this skill?
- Do not treat “validated JSON-LD” as execution evidence; verify the rendered DOM, Schema.org Validator, and Google Rich Results Test before publishing.
- The FAQ/HowTo rich-result guidance is internally inconsistent and date-sensitive; re-check current search-engine policy at generation time.
- Do not publish placeholders, invisible properties, fabricated prices/ratings/dates/authors, or CTR percentages as measured results.
- Assess data egress and mainland-China reachability before relying on WebFetch or external validation services.
- Shell / CLI
- Network access
- Local filesystem
Python 3
In Claude Code, run /plugin marketplace add aaron-he-zhu/aaron-marketing-skills, then /plugin install aaron-marketing@aaron. On compatible Agent Skills hosts, run npx skills add aaron-he-zhu/aaron-marketing-skills, or clone the repository with git clone https://github.com/aaron-he-zhu/aaron-marketing-skills. The skill file is at seo-geo/implement/serp-markup-builder/SKILL.md.
tmp="$(mktemp -d)"
git clone --depth 1 https://github.com/aaron-he-zhu/aaron-marketing-skills.git "$tmp"
mkdir -p ~/.claude/skills
cp -R "$tmp/seo-geo/implement/serp-markup-builder" ~/.claude/skills/
rm -rf "$tmp"Generated from the source repository and skill path; it copies only this skill's folder. If the author's install steps above differ, follow those first. To scope it to one project, replace ~/.claude/skills with that project's .claude/skills.
How do you use this skill?
Use a request containing the target page or topic, such as [meta] Optimize meta tags for a page about [topic] targeting [keyword] or [schema] Generate schema markup for this [content type]: [content/URL]. If no mode is stated, it infers meta from title, description, or OG language and schema from JSON-LD or rich-result language. If both are requested, it runs meta first and schema second. It asks for clarification when no page or inferable topic is available.
What are this skill's strengths and limitations?
- Covers titles, meta descriptions, Open Graph, Twitter Cards, canonical and robots tags, and JSON-LD.
- Provides two explicit modes with ready-to-paste deliverables.
- Labels metrics as Measured, User-provided, or Estimated and prohibits invented performance or entity facts.
- Documents FAQ and HowTo rich-result limitations and the risk of false negatives for client-injected JSON-LD.
- Uses the Apache-2.0 license and is intended for Claude Code and compatible Agent Skills hosts.
- Reliable output depends on supplied page content, keywords, entity facts, and other inputs.
- WebFetch and the JSON-LD pre-flight inspect server HTML and may miss markup injected client-side.
- It does not write body copy, diagnose technical crawl or indexing issues, or perform the publish-readiness audit.
- The source does not provide independent test coverage, platform validation results, or performance evidence for this individual skill.
- FAQ and HowTo markup may retain semantic or AEO value, but should not generally be treated as a Google rich-result promise.
How does this skill compare with similar options?
Side by side with related skills; every score comes from the same FSRS standard.
| Skill | FS score | Stars | Last updated | License |
|---|---|---|---|---|
| SERP Markup Builder this page | 58 · Recommended | ★ 3k | 12d ago | Apache-2.0 |
| Blog SEO Validator | 44 · Not recommended | ★ 2.4k | 13d ago | MIT |
| SEOBuild Onpage (SEO-AGI) | 50 · Use with care | ★ 257 | 10d ago | — |
| Schema Markup Assistant | 52 · Use with care | ★ 19k | 22d ago | MIT |
| Blog Schema: JSON-LD Markup Generator | 46 · Use with care | ★ 2.4k | 13d ago | MIT |
How did FollowSkills review this skill?
The evidence limits the declared tool to WebFetch, treats fetched content as untrusted data, forbids invention, and saves results only after user confirmation; no malware, credential theft, or destructive default is evident. Points are deducted because external fetching, writes to memory/, and potentially publish-affecting canonical/robots output lack detailed per-step confirmation, rollback, and permission-boundary controls.
Mode selection, inputs, outputs, handoff, stop conditions, and missing-data behavior are reasonably specified, with a JSON-LD checklist. Points are deducted because the skill claims to produce validated JSON-LD without static test evidence; it references a local schema_lint.py while declaring only WebFetch; and its FAQ/HowTo rich-result guidance is inconsistent across the supplied files.
Triggers, meta/schema modes, required inputs, non-fit boundaries, and some Chinese localization guidance are explicit. Points are deducted because localization is mostly metadata and character-count guidance while templates and examples are primarily English; the validation workflow relies on Google and Schema.org services whose mainland-China reachability is not established.
The skill has clear progressive structure, Quick Start prompts, a contract, reference materials, version and Apache-2.0 license metadata, and handoff guidance. Points are deducted because the supplied skill evidence does not clearly identify a maintainer, update path, or changelog, and it depends on multiple referenced files and paths that are not included here.
The intended deliverables are concrete: title and description options, social tags, JSON-LD, placement, and validation steps; the templates are generally usable starting points. Static calibration limits this to 7; points are deducted because validation was not executed, placeholders and visible-content alignment still require review, and CTR-lift claims are not independently supported by the supplied material.
The files provide auditable workflows, templates, field rules, named validation tools, and troubleshooting cases, supporting limited static verification. Points are deducted because there is no committed test suite, CI coverage for this skill, third-party execution evidence, or corroborating sources; static reading cannot confirm validation, character counts, or rich-result eligibility.
Open a dimension to read why it scored that way
Evidence confidence：Low — Mostly static review, author material or a limited demo; useful for discovery, not high-risk decisions.
See the full review method →