Optimize for search engine visibility and ranking. Use when asked to "improve SEO", "optimize for search", "fix meta tags", "add structured data", "sitemap optimization", or "search engine optimizatio
Install
The skill file stays in the publisher's repository.
Paste the copied command in a terminal. It installs this SKILL.md into that agent.
npx skills add addyosmani/web-quality-skills --skill seo -a cursor -y
Optimize for search engine visibility and ranking. Use when asked to "improve SEO", "optimize for search", "fix meta tags", "add structured data", "sitemap optimization", or "search engine optimization".
Questions
How do I install seo?
Download skills/seo/SKILL.md from the publisher, or copy the path and open the repository. PlugMCP lists the skill and does not host the file.
Where does seo live?
The publisher keeps it at https://github.com/addyosmani/web-quality-skills.
Which agents can use seo?
Claude, Cursor, and Codex can read a SKILL.md you add from the publisher's repository.
Optimize for search engine visibility and ranking. Use when asked to "improve SEO", "optimize for search", "fix meta tags", "add structured data", "sitemap optimization", or "search engine optimization".
name: seo
description: Optimize for search engine visibility and ranking. Use when asked to "improve SEO", "optimize for search", "fix meta tags", "add structured data", "sitemap optimization", or "search engine optimization".
license: MIT
metadata:
author: web-quality-skills
version: "2.0"
SEO optimization
Search engine optimization based on Lighthouse SEO audits and Google Search guidelines. Focus on technical SEO, on-page optimization, and structured data.
Evidence-led audit workflow
When a rendered page is available:
Run live Lighthouse SEO and Agentic Browsing checks when that capability is available; with Chrome DevTools MCP, use lighthouse_audit. Use the results to localize rendered-page failures.
Inspect signals Lighthouse cannot establish on its own: response headers, redirects, robots.txt, sitemap coverage, canonical consistency across page templates, structured-data eligibility, and Search Console evidence when the user provides access.
Separate technical crawl/index findings from content quality and authority. Do not invent ranking-factor weights or promise ranking changes.
Fix the source and re-run the same checks. For indexation or ranking outcomes, report that search-engine validation remains pending.
If live tools are unavailable, use category-specific Lighthouse CLI output plus direct source and HTTP inspection. A Lighthouse SEO score covers a useful subset of technical checks; it is not a prediction of rankings.
Area
What this skill can verify
Crawl and index controls
Technical configuration and consistency
Rendered metadata and semantics
Presence, validity, and page-template issues
Structured data
Syntax and eligibility signals, not guaranteed rich results
Core Web Vitals
Link to measured field/lab evidence from the Core Web Vitals skill
Content usefulness and authority
Review quality, but do not assign synthetic ranking percentages
<title>Page</title>
<title>Home</title>
<title>Blue Widgets for Sale | Premium Quality | Example Store</title>
Title tag guidelines:
Use 50–60 characters only as a rough linting proxy, not a pass/fail limit. Google truncates title links to fit the rendered device width, so preview width when tooling supports it.
Describe the page topic naturally near the beginning
Unique for every page
Add the brand when it helps users distinguish the result
Action-oriented when appropriate
Treat title-link rewriting separately from truncation. Google may build a different title link from the visible page title, headings, anchor text, and other sources even when the <title> is short; investigate accuracy and consistency rather than shortening it automatically. See Google's title-link guidance.
Meta descriptions
<meta name="description" content="">
<meta name="description" content="Shop premium blue widgets with free shipping. 30-day returns. Rated 4.9/5 by 10,000+ customers. Order today and save 20%.">
Meta description guidelines:
Use roughly 150–160 characters only as a linting proxy. Snippets are query- and device-dependent, and Google may select page content instead of the meta description.
Make the primary page heading descriptive and the hierarchy unambiguous; do not fail a page solely because valid HTML contains more than one <h1>
Logical hierarchy (don't skip levels)
Include keywords naturally
Descriptive, not generic
Image SEO
<img src="IMG_12345.jpg">
<img src="blue-widget-product-photo.webp"
alt="Blue widget with chrome finish, side view showing control panel"
width="800"
height="600"
loading="lazy">
Image guidelines:
Descriptive filenames with keywords
Alt text describes the image content
Compressed and properly sized
WebP/AVIF with fallbacks
Lazy load below-fold images
Internal linking
<a href="/products">Click here</a>
<a href="/widgets">Read more</a>
<a href="/products/blue-widgets">Browse our blue widget collection</a>
<a href="/guides/widget-maintenance">Learn how to maintain your widgets</a>
Linking guidelines:
Descriptive anchor text with keywords
Link to relevant internal pages
Reasonable number of links per page
Fix broken links promptly
Use breadcrumbs for hierarchy
Structured data (JSON-LD)
Read the structured data reference when the user requests schema markup or an audit surfaces a structured-data issue. It contains Organization, Article, Product, FAQ, and Breadcrumb examples plus validation links.
Describe visible, accurate content. Do not add a type or claim solely to obtain a rich result.
Use the most specific applicable type. Keep identifiers and absolute URLs stable across renders.
Validate rendered output. Passing syntax does not guarantee search-engine eligibility or display.
Agentic browsing and AI discoverability
Keep these concepts separate:
Lighthouse Agentic Browsing measures technical signals that help an assistant understand and interact with the rendered page. Current checks include the agent-facing accessibility tree, optional llms.txt, and WebMCP registrations, schemas, and form coverage when present.
Search indexing and ranking depend on search-engine systems and cannot be inferred from the Agentic Browsing score.
AI ingestion or citation is product-specific. A technically browsable page or valid llms.txt file does not prove that an AI product will ingest, rank, or cite it.
Prioritize semantic HTML, descriptive labels, crawlable content, accurate metadata, and clear page structure because they benefit people, search engines, and agents. Add WebMCP tools only when the application has useful actions to expose and the user wants that integration; validate tool names, descriptions, schemas, and form annotations with Lighthouse.
Crawler controls are product-specific
Audit each documented user agent separately instead of applying a blanket "AI bot" rule:
Control
Documented purpose
Effect of blocking
OAI-SearchBot
ChatGPT search discovery
Prevents page content from being included in ChatGPT summaries and snippets; a link and title may still surface through third-party discovery
PerplexityBot
Perplexity search indexing
Prevents that crawler from indexing the blocked content for search results
Claude-SearchBot / Claude-User
Claude search indexing / user-directed retrieval
May reduce search visibility / prevents retrieval for user-directed requests
Google-Extended
Controls certain Gemini training and grounding uses of content Google crawls
Does not affect Google Search inclusion or ranking
Training controls such as GPTBot and ClaudeBot are distinct from search and user-fetch controls. GoogleOther is a generic crawler, not an AI-search visibility switch. Verify current names and consequences in the vendors' maintained documentation: OpenAI, Perplexity, Anthropic, and Google.
llms.txt is optional
llms.txt is an experimental proposal, not a cross-vendor discovery standard. Lighthouse can validate the availability and shape of /llms.txt, but that does not show that a target product reads it. Add one only when the user requests it or a documented consumer supports it; do not recommend it ahead of crawlability, semantic HTML, accurate metadata, and useful content. Never treat it as a ranking or citation factor, duplicate the sitemap, or reorganize content solely to raise this audit.