agentsclimarketplace

Web seo

Skill ahtishamshahzad/agent_dev_flow/.ai/skills/web/web-seo

Tool-neutral AI Engineering System: 174 reusable skills (installable as Claude Code plugins) for planning and building software with AI agents. Classify → plan → approve → build under quality gates. Works with Claude Code, Codex, Cursor, Windsurf, Copilot & Antigravity. Canonical in .ai/.

Install
npx -y skills add ahtishamshahzad/agent_dev_flow --skill web-seo

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • 23 days oldThe repository was created 23 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Use to plan SEO for web apps where search visibility matters — per-page metadata, Open Graph/Twitter cards, sitemap, robots, canonical URLs, structured data (JSON-LD), and rendering strategy for indexable content. Applies to public/marketing surfaces; skip for authenticated dashboards.

SKILL.md

5.3 KB, ~1.1k tokens by cl100k_base, as published. Nobody here has run it

Web SEO

Purpose

Plan search visibility for the surfaces that need it: metadata and social cards per page, crawlability (sitemap, robots, canonicals), structured data, and a rendering strategy that serves real HTML to crawlers. SEO effort is scoped to the apps where it matters — a marketing site, yes; an internal dashboard, no.

When to Use

  • For public web / marketing / indexable content surfaces (per ../../application-selection and web-stack-selection).
  • When an existing public app underperforms in search or unfurls badly when shared.
  • Not for authenticated SPAs or admin dashboards — record "SEO: not applicable" and move on.

Inputs

  • Which pages/routes are public and should rank; content/keyword intent per page.
  • Rendering strategy from nextjs-foundation (SEO-relevant apps normally chose Next.js — content must be server-rendered or static).
  • Domain/URL structure plans (web-deployment).

Discovery Questions

  • Which pages carry the search value (home, product/landing pages, blog, docs), and what should each rank for?
  • What social-share behavior is expected (OG/Twitter cards per page type)?
  • Which structured-data types fit (Organization, Product, Article, FAQ, Breadcrumb)?
  • Are there duplicate-content risks needing canonicals (params, pagination, staging)? Multiple locales (hreflang)?

Responsibilities

  • Plan per-page metadata: unique title/description per indexable page, templated by page type; Open Graph/Twitter card fields with correctly sized images.
  • Plan crawlability: XML sitemap generation (kept current), robots rules (block staging/preview and non-indexable areas, never block what should rank), canonical URLs for parameterized/duplicate views.
  • Plan structured data (JSON-LD) per page type where it earns rich results — valid and truthful, not markup spam.
  • Confirm the rendering strategy serves complete HTML for indexable content (static/ISR/SSR per nextjs-foundation) — client-only rendering for content that must rank is a finding to escalate to web-stack-selection.
  • Align with Core Web Vitals (web-performance) and semantic heading structure (web-accessibility) — both feed ranking.
  • Define URL conventions: stable, readable, lowercase paths; redirects for moved content.

Required Workflow

  1. Scope: list indexable pages; explicitly exempt non-SEO surfaces.
  2. Define metadata + social-card templates per page type.
  3. Plan sitemap/robots/canonical mechanics.
  4. Select structured-data types per page and their data sources.
  5. Verify rendering serves indexable HTML; record the plan and validation approach (Lighthouse SEO, rich-results test — run or mark "unverified until run").

Decision Rules

  • SEO scope follows the application decision — spend nothing on surfaces that will never be indexed.
  • Content that must rank is rendered server-side/static; no exceptions built on "Google runs JS."
  • One canonical URL per piece of content; parameters/variants point at it.
  • Structured data only where the page genuinely is that thing.

Rules

  • Staging/preview environments are never indexable (robots + auth, both).
  • Metadata is data-driven per page — no site-wide duplicate titles.
  • No cloaking, keyword stuffing, or markup describing content that isn't on the page.

Anti-Patterns

  • Retrofitting SEO onto a client-only SPA that should have escalated the framework decision.
  • Same title/description on every page.
  • Sitemap listing 404s/redirects; robots.txt blocking the whole site in production (or staging left indexable).
  • JSON-LD claiming reviews/ratings the page doesn't show.

Validation Checklist

  • Indexable pages scoped; non-SEO surfaces exempted explicitly.
  • Per-page metadata + social cards planned by page type.
  • Sitemap, robots, and canonical mechanics planned (staging blocked).
  • Structured data planned per page type, truthfully.
  • Rendering verified to serve indexable HTML (or escalated).
  • Validation runs planned or executed (marked if unrun).

Definition of Done

A recorded SEO plan scoped to the surfaces that need it — metadata templates, crawlability mechanics, structured data, and confirmed indexable rendering — with validation steps run or explicitly marked unverified.

Related Skills

web-stack-selection, nextjs-foundation, web-performance, web-accessibility, web-routing, web-deployment.

Related Knowledge

../../../knowledge/ (content strategy, target queries).

Related References

../../../references/web/seo/ (metadata templates, schema examples — when populated).

Context Loading Guidance

  • Requires: indexable-page list, rendering strategy, URL/domain plans.
  • Does not require: dashboard skills, state/data internals.
  • May load: web-performance for Core Web Vitals; web-deployment for domain/redirect mechanics.
  • Stop when: the SEO plan and validation status are recorded.

Token Efficiency Guidance

Template metadata by page type instead of enumerating every page. Keep schema examples in references; the plan lists types and sources only.

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Keep looking

Skills are one crate of 326,984. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.