SEO Audit Skill

Run a full SEO audit on any website in seconds. Get a 0-100 score, 50+ individual checks, industry compliance flags, and a prioritized fix list — no agency required.

Free vs Premium

Free tier (no API key needed):

• --demo — full audit output on a built-in static sample, zero network calls, see exactly what the output looks like
• --compliance-only <url> — fast HTTP fetch + forbidden word/MLO compliance scan, no AI analysis
• --version — print version

Premium tier (ANTHROPIC_API_KEY):

• Full 50+ check crawl + AI analysis (Claude Haiku, ~$0.003/audit)
• --competitor — side-by-side gap analysis against a competitor URL
• --generate-article — AI-written 800-word SEO article with H1/H2/FAQ schema, ~$0.003/article
• --format=json — pipe results into agent workflows or dashboards
• Industry-specific scoring modes: --industry=mortgage, --industry=medical, --industry=legal

The free compliance scan catches NMLS violations, missing disclosures, and forbidden words before they cost you — worth installing on that alone.

What this skill does

Crawls a URL and runs 50+ checks across 5 categories. Returns a weighted 0-100 score, per-check PASS/FAIL/WARN results, and a ranked fix list ordered by score impact.

Check categories

Scoring: Weight-based (not binary pass/fail count). Critical checks (NMLS, HTTPS, H1) weight 3×. Minor checks weight 1×. Score = pass_weight / total_weight × 100.

Input contract

Tell me:

1. URL to audit — the page you want scored
2. Target keyword (optional) — for on-page keyword density check
3. Industry (optional): mortgage / real-estate / crypto / saas / healthcare / legal / general
4. Competitor URL (optional) — for gap analysis

Example prompts:

• "Audit https://example.com for 'mortgage broker [city]'"
• "Audit https://example.com --competitor https://rocket.com"
• "Check https://example.com compliance only"
• "Generate an article for 'first time home buyer [city]'"

Output contract

Standard audit output:

SEO Audit — https://example.com
Score: 74/100  |  Industry: Mortgage  |  Keyword: "mortgage broker [city]"

TECHNICAL (8/10)
  ✓ HTTPS enforced
  ✓ robots.txt found
  ✓ sitemap.xml found
  ⚠ No canonical tag — add <link rel="canonical"> to prevent duplicate content
  ✗ Missing mobile viewport meta

ON-PAGE (6/10)
  ✓ Title: 52 chars, keyword present
  ✗ Meta description: 210 chars (too long, truncates at 160)
  ✓ H1 present: "[City] Mortgage Broker — Get Pre-Approved Today"
  ⚠ Only 380 words — thin content, target 800+
  ✗ 4 images missing alt text

SCHEMA (3/10)
  ✗ No LocalBusiness schema — quick win, adds map pack eligibility
  ✗ No FAQ schema — competitors using this get 2× SERP real estate
  ✗ No Review schema

SOCIAL (9/10)
  ✓ OG title, description, image all present
  ✓ Twitter card configured

COMPLIANCE (10/10)  [Mortgage]
  ✓ NMLS number visible
  ✓ Phone number present
  ✓ Physical address in footer
  ✓ No forbidden words detected

TOP 3 QUICK WINS (ranked by score impact):
  1. Add LocalBusiness + FAQ schema  (+8 pts estimated)
  2. Expand content to 800+ words    (+6 pts estimated)
  3. Fix meta description length     (+4 pts estimated)

Competitor gap analysis (with --competitor):

GAP ANALYSIS — Your Site vs competitor.com

Your site WINS: NMLS compliance, OG tags, sitemap
Competitor WINS: FAQ schema, 2,400-word content, 3 review schema blocks

Top gaps to close:
  1. FAQ schema — competitor has 8 FAQ pairs, you have 0
  2. Content depth — competitor 2,400 words vs your 380
  3. Review schema — competitor shows star rating in SERP, you don't

Article generation (with --generate-article):

[800-word article with H1/H2 structure, FAQ section, target keyword density 1-2%]
Ready to paste into CMS. Estimated read time: 4 min.

How the skill works

Uses seo_audit_skill.py (in this directory). Makes HTTP requests with BeautifulSoup for crawling, Claude Haiku for AI analysis.

# Full audit
python3 seo_audit_skill.py --url https://example.com --keyword "mortgage broker [city]"

# Demo mode (no network, no API — see the output format)
python3 seo_audit_skill.py --demo

# Compliance scan only (free, HTTP fetch only)
python3 seo_audit_skill.py --compliance-only --url https://example.com --industry=mortgage

# Competitor gap analysis
python3 seo_audit_skill.py --url https://yoursite.com --competitor https://competitor.com

# Generate SEO article (premium)
python3 seo_audit_skill.py --generate-article "first time home buyer [city]" --output article.md

# JSON output (for pipelines)
python3 seo_audit_skill.py --url https://example.com --format=json | jq '.score'

# Industry-specific compliance
python3 seo_audit_skill.py --url https://medicalsite.com --industry=healthcare

# Version
python3 seo_audit_skill.py --version

Industry compliance modes

Why weight-based scoring

A missing NMLS number is not equal to a missing favicon. Weight-based scoring reflects real SEO impact:

• Critical (weight 3): HTTPS, NMLS/license number, H1, sitemap
• Moderate (weight 2): Meta description, canonical, mobile viewport, schema
• Minor (weight 1): OG tags, alt text on individual images, word count

A site with perfect compliance + schema but missing HTTPS scores lower than a site with HTTPS + good on-page but weak social.

Competitor intelligence note

Large mortgage sites (Rocket, LoanDepot, NerdWallet) use Cloudflare WAFs and will 403 bot requests. The audit reports the 403 as a FAIL on HTTPS-accessible content — which is honest. For auditing your own site, no issue. For competitor research on large players, use --format=json to capture what passes before the block.

Real finding: NerdWallet scores ~70/100 on this audit — they fail schema markup entirely and have a 22-char title. Their SEO advantage is domain authority (backlinks), not on-page quality. This gap is closeable for smaller operators who nail technical + schema.

Integration with agent infrastructure

# Via Telegram
@openclaw seo-audit "Audit https://example.com for '[city] mortgage'"
@openclaw seo-audit "Compare https://mysite.com vs https://competitor.com"

# Via Claude Code
openclaw run seo-audit "Audit https://example.com"

# In agent pipelines (JSON mode)
python3 seo_audit_skill.py --url https://example.com --format=json | jq '.score, .quick_wins'