SEO/GEO Auditor
Expert SEO and GEO (Generative Engine Optimization) auditor for Next.js sites. Uses fan-out query methodology to measure semantic coverage and identify content gaps.
What This Skill Does
Produce a structured audit with 3 scores, fan-out queries grouped by semantic cluster, gap analysis with specific file paths, and actionable code deltas. Works on any Next.js site: comparators, SaaS, blogs, e-commerce, directories.
Core Methodology
Phase 1: Site Detection
Read the codebase to determine:
- Thematic: Main topic/industry (from page content, data files, metadata)
- Site type: Comparator, SaaS, blog, e-commerce, directory
- Architecture: App Router vs Pages, static vs dynamic, data sources
- Existing pages: Route inventory from
app/directory +generateStaticParams
Priority order: read files first (app/, lib/, components/). Use crawl or fetch only when files are insufficient.
Phase 2: Fan-Out Query Generation
Generate 25-50 realistic search queries that real users would type, grouped by semantic clusters.
Cluster types by site type:
| Site Type | Typical Clusters |
|---|---|
| Comparator | solutions, vs comparisons, use-cases, company size, features, pricing |
| SaaS | features, alternatives, integrations, use-cases, how-to, pricing |
| Blog | informational, how-to, tools, trends, definitions |
| E-commerce | products, categories, buyer intent, comparisons |
Volume indicators per query:
- High: likely >1K searches/month
- Medium: 100-1K/month
- Niche: <100/month but high intent
For detailed methodology, consult references/fan-out-methodology.md.
Phase 3: Scoring & Gaps
Calculate 3 mandatory scores:
- GEO Score /100 — How well can AI engines extract and cite the site's content
- Fan-Out Coverage % — Percentage of generated queries covered by existing pages
- Next.js SEO Score /100 — Technical SEO implementation quality
For detailed scoring rubrics, consult references/geo-scoring.md.
Output Modes
| Mode | Trigger | Queries | Detail |
|---|---|---|---|
| Standard | default | 25-50 | Balanced: scores, queries, gaps, quick wins, roadmap |
| Express | --express | 15-25 | Ultra-concise: scores + top 5 quick wins + 3 pages |
| Deep | --deep | 50-80 | Exhaustive: per-page analysis, internal linking map, competitor analysis |
Output Structure (strict order)
## Scores
GEO: XX/100 | Fan-Out: XX% | Next.js SEO: XX/100
## Fan-Out Queries (grouped by cluster, with volume indicators)
## Gaps (each pointing to a specific file path or missing route)
## Quick Wins (<48h, code deltas only)
## Routes to Create (path, title, H1, target queries)
## Templates (FAQ schema, comparison table — if relevant)
## Roadmap (J1-7, J8-30)
## Bonus (1-2 prompts to generate missing content)For a real output example, consult references/example-output.md.
Code Output Rules
- Deltas only: Never output complete files — only the diff to apply
- Actionable: Every recommendation includes a specific file path or route
- Copy-paste ready: Code snippets work as-is when pasted into the project
Adaptation by Site Type
| Type | Focus |
|---|---|
| Comparator / pSEO | Maximize generateStaticParams, cross-comparison pages, FAQ per entity, schema SoftwareApplication |
| Blog | Topical clusters, internal linking, schema Article, author pages |
| SaaS | Feature pages, /vs/ pages, /for/ pages, FAQ product, schema SoftwareApplication |
| E-commerce | Schema Product, category pages, buyer FAQ |
Next.js Technical Checks
For the full checklist (30+ items), consult references/nextjs-seo-checklist.md.
Key checks: generateMetadata on every route, generateStaticParams for dynamic routes, sitemap.ts, robots.ts with AI bot permissions, JSON-LD structured data, OpenGraph images, loading.tsx, next/image optimization.
Additional Resources
Reference Files
references/fan-out-methodology.md— Fan-out query generation, cluster design, volume estimationreferences/geo-scoring.md— Detailed scoring rubrics for all 3 scoresreferences/nextjs-seo-checklist.md— Complete Next.js App Router SEO checklistreferences/example-output.md— Real audit output example (comparator site)