AgentSkillsCN

Ct Site Auditor

Ct站点审核员

SKILL.md

ct-site-auditor

Comprehensive site-wide SEO, performance, and accessibility auditing skill based on the 25-section Website Audit Checklist.

Overview

This skill performs full website audits across 11 categories with 105+ checks:

Core Categories:

  1. Technical SEO (12 checks)
  2. Schema Markup (7 checks)
  3. E-E-A-T Signals (8 checks)
  4. On-Page SEO (20 checks, includes social meta)
  5. AI/LLM Optimization (5 checks)
  6. Performance (17 checks, includes budgets)
  7. Accessibility (8 checks)
  8. Security (8 checks)

Extended Categories: 9. Mobile (10 checks) 10. International (8 checks) 11. Competitor (8 checks, optional)

Capabilities

Inputs

  • SITE_URL - The website URL to audit
  • AUDIT_TYPE - full, quick, or category-specific
  • CATEGORY_FILTER - Optional category to focus on
  • AUTO_FIX - Whether to apply automatic fixes

Outputs

  • site_audit_report - Comprehensive audit results
  • category_scores - Individual category scores (0-100)
  • issues_list - Prioritized list of issues
  • task_suggestions - Auto-created task IDs

Dependencies

  • ct-metrics-store - For storing audit results
  • ct-task-manager - For creating tasks from issues

MCP Requirements

  • gsc - Google Search Console data
  • dataforseo - Lighthouse, on-page analysis, backlinks, competitor analysis, mobile CWV
  • scraperapi - HTML fetching, header analysis, hreflang parsing, social meta parsing

Audit Modules

Module 1: Technical SEO

Checks: TECH001-TECH012

Uses:

  • Local file reads for robots.txt, sitemap.xml
  • mcp__gsc__get_sitemaps for sitemap status
  • mcp__gsc__check_indexing_issues for crawl errors
  • mcp__scraperapi__scrape for redirect and HTTPS checks

Key Checks:

  • robots.txt properly configured
  • XML sitemap exists and submitted
  • Canonical tags present
  • HTTPS with no mixed content
  • No orphan pages

Module 2: Schema Markup

Checks: SCHEMA001-SCHEMA007

Uses:

  • mcp__dataforseo__on_page_content_parsing for schema extraction
  • mcp__scraperapi__scrape for HTML parsing
  • JSON-LD validation logic

Key Checks:

  • Organization/Person schema present
  • WebSite schema with SearchAction
  • Article/BlogPosting on blog content
  • BreadcrumbList for navigation
  • FAQPage on FAQ pages
  • Schema validation passes

Module 3: E-E-A-T Signals

Checks: EEAT001-EEAT008

Uses:

  • Local file scanning for page content
  • HTML parsing for author elements
  • mcp__scraperapi__scrape for live site checks

Key Checks:

  • Author bylines with credentials
  • About page comprehensive
  • Contact information accessible
  • Published/updated dates visible
  • External citations to authoritative sources

Module 4: On-Page SEO

Checks: ONPAGE001-ONPAGE010

Uses:

  • mcp__dataforseo__on_page_instant_pages for meta analysis
  • mcp__dataforseo__on_page_content_parsing for heading structure
  • Local file scanning for URLs

Key Checks:

  • Unique title tags (50-60 chars)
  • Meta descriptions (150-160 chars)
  • Single H1 per page
  • Proper heading hierarchy
  • Images have alt text
  • Clean, keyword-rich URLs
  • Strong internal linking

Module 5: AI/LLM Optimization

Checks: AI001-AI005

Uses:

  • Content structure analysis
  • mcp__dataforseo__ai_opt_llm_ment_search for LLM mentions
  • HTML parsing for FAQ elements

Key Checks:

  • Clear, quotable sections
  • Q&A format for common questions
  • Key takeaways section
  • Statistics with attribution
  • Technical terms defined

Module 6: Performance

Checks: PERF001-PERF008

Uses:

  • mcp__dataforseo__on_page_lighthouse for Core Web Vitals
  • Lighthouse performance audit

Key Checks:

  • LCP < 2.5s
  • FID/INP < 100ms
  • CLS < 0.1
  • Page load < 3s
  • TTFB < 600ms
  • Images optimized (WebP/AVIF)
  • CSS/JS minified
  • Lazy loading enabled

Module 7: Accessibility

Checks: A11Y001-A11Y008

Uses:

  • mcp__dataforseo__on_page_lighthouse accessibility audit
  • HTML parsing for ARIA, landmarks

Key Checks:

  • Color contrast meets WCAG AA
  • Keyboard navigable
  • Focus indicators visible
  • ARIA labels present
  • Landmark regions defined
  • Form labels associated
  • Skip link present
  • Videos have captions

Module 8: Security

Checks: SEC001-SEC008

Uses:

  • mcp__scraperapi__scrape for header analysis
  • SSL certificate checks

Key Checks:

  • Content-Security-Policy configured
  • X-Frame-Options set
  • HSTS enabled
  • X-Content-Type-Options set
  • SSL valid and not expiring
  • TLS 1.2 or higher
  • Privacy policy exists
  • Cookie consent implemented

Module 9: Mobile

Checks: MOBILE001-MOBILE010

Uses:

  • mcp__dataforseo__on_page_lighthouse with mobile emulation
  • mcp__scraperapi__scrape for viewport meta parsing

Key Checks:

  • Mobile LCP < 2.5s
  • Mobile INP < 200ms
  • Mobile CLS < 0.1
  • Viewport meta configured
  • Tap targets 48x48px+
  • Font sizes >= 16px
  • Content fits viewport
  • Mobile-first indexing ready
  • Mobile content parity
  • Mobile resources optimized

Module 10: International SEO

Checks: INTL001-INTL008

Uses:

  • mcp__scraperapi__scrape for hreflang parsing
  • Multi-URL validation for reciprocal links

Key Checks:

  • hreflang tags present
  • x-default hreflang exists
  • hreflang reciprocal valid
  • Language codes valid (ISO)
  • HTML lang attribute set
  • Content-Language header
  • Geo-targeting consistent
  • Regional content localized

Module 11: Competitor Benchmarking

Checks: COMP001-COMP008

Uses:

  • mcp__dataforseo__dataforseo_labs_google_domain_rank_overview
  • mcp__dataforseo__dataforseo_labs_google_domain_intersection
  • mcp__dataforseo__backlinks_competitors

Key Checks:

  • Domain authority comparison
  • Keyword overlap analysis
  • Keyword gap opportunities
  • Referring domains comparison
  • Backlink quality comparison
  • Link gap opportunities
  • Performance vs competitors
  • Content coverage analysis

Requires competitors to be configured via /competitor add <domain>.

Scoring System

Category Weights

code
Technical:      12%
Schema:          9%
E-E-A-T:        12%
On-Page:        12% (includes social meta)
AI:              8%
Performance:    12% (includes budgets)
Accessibility:   9%
Security:        8%
Mobile:          8%
International:   5%
Competitor:      5% (optional)
─────────────────
Total:         100%

Score Calculation

python
overall_score = (
    technical_score * 0.12 +
    schema_score * 0.09 +
    eeat_score * 0.12 +
    onpage_score * 0.12 +
    ai_score * 0.08 +
    performance_score * 0.12 +
    accessibility_score * 0.09 +
    security_score * 0.08 +
    mobile_score * 0.08 +
    international_score * 0.05 +
    competitor_score * 0.05  # optional, re-weighted if skipped
)

Severity Classification

  • Critical: Security issues, major indexability problems, severely broken UX
  • High: Missing required elements, performance failures, accessibility violations
  • Medium: Best practice violations, optimization opportunities
  • Low: Nice-to-have improvements, polish items

Auto-Fix Capabilities

The following checks support automatic fixes:

Check CodeFix Action
SCHEMA001Create OrganizationSchema component
SCHEMA002Create WebSiteSchema component
SCHEMA004Add Article schema to blog layout
SCHEMA005Add FAQPage schema to FAQ page
ONPAGE003Generate meta descriptions from content
PERF008Add loading="lazy" to images
A11Y007Add skip link to layout

Data Storage

Tables Used

  • site_audits - Overall audit results (11 category scores)
  • audit_checks - Individual check results (105+ checks)
  • check_definitions - Check metadata and severity
  • lighthouse_snapshots - Performance history
  • backlink_snapshots - Link profile history
  • schema_validations - Schema validation results
  • security_headers - Security header checks
  • mobile_audits - Mobile-specific results
  • hreflang_analysis - International SEO data
  • social_meta_analysis - Open Graph/Twitter Card data
  • performance_budgets - Budget thresholds
  • budget_compliance - Budget compliance history
  • competitors - Tracked competitors
  • competitor_snapshots - Competitor comparison data

Views Used

  • latest_site_audit - Most recent audit with trend
  • site_audit_trends - Historical score changes
  • failing_checks_summary - Current failing checks
  • category_scores - Latest category breakdown
  • budget_status - Current budget compliance
  • competitor_summary - Latest competitor metrics

Task Integration

Auto-Created Tasks

ALL audit issues automatically create tasks, regardless of severity:

Audit SeverityTask Priority
Criticalcritical
Highhigh
Mediummedium
Lowlow
json
{
  "content": "Add Organization schema to site layout",
  "activeForm": "Adding Organization schema",
  "status": "pending",
  "priority": "critical",
  "labels": ["audit", "schema"],
  "source": "audit-site",
  "audit": {
    "checkCode": "SCHEMA001",
    "category": "schema",
    "severity": "critical",
    "impact": "+15 points",
    "fixSuggestion": "Add Organization JSON-LD to site layout"
  }
}

This ensures full visibility of all issues. Tasks can be filtered by priority when viewing.

Task Linking

  • audit_checks.cleo_task_id links check to task
  • When task completed, check can be marked fixed_at
  • Re-audit verifies the fix

Usage Examples

Full Site Audit

code
Input: /audit site --full
Process:
  1. Verify MCPs (gsc, dataforseo, scraperapi)
  2. Detect site URL from config
  3. Run all 8 audit modules
  4. Calculate scores
  5. Store results
  6. Create tasks for critical/high issues
  7. Display comprehensive report

Category Audit

code
Input: /audit site --category=performance
Process:
  1. Verify dataforseo MCP
  2. Run only Module 6 (Performance)
  3. Run Lighthouse for all pages
  4. Store results
  5. Display focused report

Quick Audit (at /start)

code
Input: (automatic at session start)
Process:
  1. Check last audit timestamp
  2. If stale (>24h), run critical checks only
  3. Store results
  4. Display summary in /start output

Dispatch Triggers

This skill activates on:

  • /audit site command
  • /start (quick mode if stale)
  • "run site audit"
  • "check website health"
  • "audit the whole site"
  • "technical SEO audit"

Error Handling

MCP Unavailable

code
ERROR: Required MCP not available: dataforseo

The site audit requires the DataForSEO MCP for:
- Lighthouse performance testing
- On-page content analysis
- Backlink profile analysis

Setup instructions:
[... setup steps ...]

Partial Failure

If one module fails, continue with others:

code
WARNING: Performance module failed (Lighthouse timeout)

Completed modules: 7/8
Skipped: Performance

Overall score calculated from available data.
Re-run /audit site --category=performance to retry.

Rate Limiting

DataForSEO has rate limits. For large sites:

code
Site has 150 pages. Running batched audit...

Batch 1/3: Pages 1-50 ✓
Batch 2/3: Pages 51-100 ✓
Batch 3/3: Pages 101-150 ✓

Audit complete.