SEO Audit Skill
Purpose
Perform a comprehensive, full-site SEO audit that evaluates every dimension of search performance. This skill orchestrates 6 specialist subagents to analyze technical infrastructure, content quality, schema markup, sitemap health, performance metrics, and visual/image optimization. The output is a single health score (0-100) and a prioritized action plan sorted by impact.
Scoring Model
Category Weights
| Category | Weight | Subagent | Focus |
|---|
| Technical SEO | 25% | seo-technical | Crawlability, indexability, security headers, Core Web Vitals, AI crawler management |
| Content Quality | 25% | seo-content | E-E-A-T signals, readability, keyword optimization, AI citation readiness |
| On-Page SEO | 20% | seo-page | Title tags, meta descriptions, H1s, URL structure, internal linking |
| Schema Markup | 10% | seo-schema | Structured data coverage, validation, type accuracy |
| Performance | 10% | seo-performance | Core Web Vitals (LCP, INP, CLS), TTFB, resource optimization |
| Images & Visual | 5% | seo-visual | Alt text, file sizes, formats, responsive images, lazy loading |
| AI Search Readiness | 5% | seo-content | GEO signals, AI Overview optimization, citation structure |
Score Interpretation
| Score Range | Rating | Meaning |
|---|
| 90-100 | Excellent | Industry-leading SEO; maintain and iterate |
| 75-89 | Good | Strong foundation with optimization opportunities |
| 50-74 | Needs Work | Significant gaps impacting rankings |
| 25-49 | Poor | Fundamental issues blocking organic growth |
| 0-24 | Critical | Severe problems requiring immediate intervention |
Industry Detection
Before starting the audit, detect the site's industry to apply context-specific scoring adjustments and recommendations.
Supported Industries
| Industry | Detection Signals | Scoring Adjustments |
|---|
| SaaS | Pricing page, signup/login flows, app subdomain, API docs | Weight schema for SoftwareApplication; prioritize conversion path SEO |
| Local Business | NAP data, Google Maps embed, service area pages, location schema | Weight LocalBusiness schema heavily; check Google Business Profile alignment |
| E-commerce | Product pages, cart, checkout, product schema, price elements | Weight Product/Offer schema; check faceted navigation, pagination |
| Publisher | High article count, author pages, date stamps, Article schema | Weight content freshness and E-E-A-T author signals; check pagination |
| Agency | Portfolio/case studies, service pages, team/about, testimonials | Weight service page optimization; check local SEO if applicable |
Detection Protocol
- Crawl homepage and top 10 pages by internal link count
- Scan for industry-specific patterns (schema types, page templates, URL patterns)
- Classify into primary industry
- Apply industry-specific scoring multipliers and recommendation templates
Crawl Protocol
Configuration
| Parameter | Value | Rationale |
|---|
| Max Pages | 500 | Sufficient for most sites; prevents runaway crawls |
| Concurrency | 5 | Respectful of server resources |
| Delay Between Requests | 1 second | Prevents rate limiting and server strain |
| Respect robots.txt | Yes (mandatory) | Always honor crawl directives |
| User-Agent | AntigravitySEOBot/1.0 | Identifiable, non-deceptive |
| Timeout Per Page | 30 seconds | Skip unresponsive pages |
| Follow Redirects | Yes (max 5 hops) | Track redirect chains |
| Parse JavaScript | Optional | Note JS-dependent content for manual review |
Crawl Phases
Phase 1: Seed Discovery
Fetch robots.txt and sitemap.xml
Extract seed URLs from sitemap
Add homepage as primary seed
|
Phase 2: Breadth-First Crawl
Crawl seeds first, then discovered internal links
Track: status codes, redirects, canonical tags, meta robots
Record: response times, content types, content hashes
|
Phase 3: Data Collection Per Page
- HTTP status code
- Response time (ms)
- Title tag (content + length)
- Meta description (content + length)
- H1 tag(s) (count + content)
- Canonical URL
- Meta robots directives
- hreflang tags
- Schema markup (types + validation)
- Image count, alt text coverage, file sizes
- Internal links (count + anchor text)
- External links (count + nofollow status)
- Word count
- Content hash (duplicate detection)
|
Phase 4: Aggregate Analysis
Pass collected data to 6 specialist subagents
Each subagent scores its category independently
Compute weighted overall score
Crawl Limits and Safety
- Stop at 500 pages regardless of discovered URLs
- Skip non-HTML resources (PDFs, images, videos) unless analyzing image SEO
- Skip external domains entirely
- Log but do not follow
nofollow links
- Detect and break infinite crawl loops (parameter-based URL variations)
- Respect
Crawl-delay directive in robots.txt if present
Subagent Delegation
Delegation Flow
+-------------------+
| SEO Audit Lead |
| (This Skill) |
+-------------------+
|
+------------------+------------------+
| | | | | |
+-----------+ +------+ +------+ +--------+ +------+ +------+
| seo- | | seo- | | seo- | | seo- | | seo- | | seo- |
| technical | | cont | | sche | | sitemap| | perf | | visu |
+-----------+ +------+ +------+ +--------+ +------+ +------+
Subagent Responsibilities
| Subagent | Input | Output |
|---|
seo-technical | Full crawl data, robots.txt, server headers | Technical score (0-100), issues list |
seo-content | Page content, word counts, readability metrics | Content score (0-100), E-E-A-T assessment |
seo-schema | Extracted JSON-LD/microdata per page | Schema score (0-100), missing types, validation errors |
seo-sitemap | sitemap.xml contents, crawled URL list | Sitemap health report, orphan pages, missing URLs |
seo-performance | Response times, resource sizes, CWV data | Performance score (0-100), bottleneck identification |
seo-visual | Image inventory (URLs, sizes, formats, alt text) | Image optimization score (0-100), fix list |
Subagent Invocation Template
You are the [SUBAGENT ROLE] specialist for the SEO Audit skill.
Site: [URL]
Industry: [DETECTED INDUSTRY]
Crawl Data: [RELEVANT SUBSET OF CRAWL DATA]
Your task:
1. Analyze all provided data from your domain expertise
2. Score 0-100 with clear justification
3. List every issue found, categorized by severity
4. Provide specific, actionable fixes for each issue
5. Note any items requiring manual verification
Output format:
- Score: XX/100
- Critical Issues: [list]
- High Priority Issues: [list]
- Medium Priority Issues: [list]
- Low Priority Issues: [list]
- Recommendations: [prioritized list with estimated impact]
Output Artifacts
1. FULL-AUDIT-REPORT.md
Complete audit report containing:
# SEO Audit Report: [Domain]
**Date:** [YYYY-MM-DD]
**Pages Crawled:** [N]
**Industry:** [Detected Industry]
**Overall Score:** [XX/100] ([Rating])
## Executive Summary
[3-5 sentence overview of site health and top priorities]
## Score Breakdown
| Category | Score | Weight | Weighted Score |
|----------|-------|--------|----------------|
| Technical SEO | XX/100 | 25% | XX |
| Content Quality | XX/100 | 25% | XX |
| On-Page SEO | XX/100 | 20% | XX |
| Schema Markup | XX/100 | 10% | XX |
| Performance | XX/100 | 10% | XX |
| Images & Visual | XX/100 | 5% | XX |
| AI Search Readiness | XX/100 | 5% | XX |
| **Overall** | | | **XX/100** |
## Technical SEO
[Full seo-technical subagent report]
## Content Quality
[Full seo-content subagent report]
## On-Page SEO
[Full seo-page analysis]
## Schema Markup
[Full seo-schema subagent report]
## Performance
[Full seo-performance subagent report]
## Images & Visual
[Full seo-visual subagent report]
## AI Search Readiness
[GEO signals and AI citation assessment]
## Crawl Summary
- Total URLs discovered: [N]
- Pages crawled: [N]
- 200 responses: [N]
- 301/302 redirects: [N]
- 404 errors: [N]
- 5xx errors: [N]
- Blocked by robots.txt: [N]
- Average response time: [N]ms
2. ACTION-PLAN.md
Prioritized action plan sorted by impact:
# SEO Action Plan: [Domain]
**Generated:** [YYYY-MM-DD]
**Current Score:** [XX/100]
**Target Score:** [XX/100] (after completing Critical + High items)
## Critical Priority (Fix Immediately)
Impact: Blocking indexing or causing significant ranking loss
- [ ] [Issue description] — [Specific fix] — [Estimated impact: +X points]
- [ ] ...
## High Priority (Fix Within 1 Week)
Impact: Directly affecting rankings or user experience
- [ ] [Issue description] — [Specific fix] — [Estimated impact: +X points]
- [ ] ...
## Medium Priority (Fix Within 1 Month)
Impact: Optimization opportunities for incremental gains
- [ ] [Issue description] — [Specific fix] — [Estimated impact: +X points]
- [ ] ...
## Low Priority (Fix When Possible)
Impact: Minor improvements, best practices alignment
- [ ] [Issue description] — [Specific fix] — [Estimated impact: +X points]
- [ ] ...
## Estimated Score After Fixes
| Phase | Actions | Projected Score |
|-------|---------|-----------------|
| After Critical fixes | [N] items | XX/100 |
| After High fixes | [N] items | XX/100 |
| After Medium fixes | [N] items | XX/100 |
| After All fixes | [N] items | XX/100 |
Priority Levels
| Level | Definition | Timeline | Examples |
|---|
| Critical | Blocking indexing, causing deindexation, or severe security issues | Immediate (same day) | noindex on key pages, site-wide 5xx errors, missing canonical causing massive duplication, security vulnerability |
| High | Directly impacting rankings or significant user experience degradation | Within 1 week | Missing/duplicate title tags, broken internal links, missing schema on key pages, slow LCP on landing pages |
| Medium | Optimization opportunities with measurable ranking potential | Within 1 month | Thin content pages, suboptimal meta descriptions, missing alt text on key images, redirect chains |
| Low | Best practice alignment and incremental improvements | When possible | Minor heading hierarchy issues, optional schema additions, image format upgrades, cosmetic URL improvements |
Execution Workflow
1. USER invokes: /seo-audit [URL]
|
2. VALIDATE URL (accessible, returns 200, is not blocked)
|
3. DETECT INDUSTRY (scan homepage + top pages)
|
4. FETCH robots.txt and sitemap.xml
|
5. EXECUTE CRAWL (breadth-first, max 500 pages, 5 concurrent, 1s delay)
|
6. DELEGATE to 6 subagents (parallel where possible)
- seo-technical: crawl data + headers + robots.txt
- seo-content: page content + word counts
- seo-schema: extracted structured data
- seo-sitemap: sitemap vs crawled URLs
- seo-performance: response times + resource data
- seo-visual: image inventory
|
7. COLLECT subagent scores and issue lists
|
8. COMPUTE weighted overall score
|
9. GENERATE FULL-AUDIT-REPORT.md
|
10. GENERATE ACTION-PLAN.md (sorted by priority, then by estimated impact)
|
11. PRESENT summary to user with score and top 5 actions
Usage
# Full audit of a website
/seo-audit https://example.com
# Audit with specific page limit
/seo-audit https://example.com --max-pages 100
# Audit with industry override
/seo-audit https://example.com --industry saas
Notes
- This skill coordinates all other SEO skills; it is the top-level orchestrator for site-wide analysis
- Individual page analysis should use the
seo-page skill instead
- Crawl data is ephemeral; it is not persisted between audit runs
- For sites with JavaScript-rendered content, note that the crawler may not execute JS; flag these pages for manual review
- Always check if the site has a staging vs production environment and audit the correct one