| name | test-seo |
| description | Run SEO and meta data testing on a URL. Use when user wants to test SEO, meta tags, Open Graph, structured data, sitemap, robots.txt, or crawlability. |
| allowed-tools | Bash(playwright-cli:*) Bash(npx:*) Bash(mkdir:*) |
SEO & Meta Data
Run 27 SEO test cases across 3 parallel agents covering meta tags, content structure, and technical SEO.
Setup
URL="${ARGUMENTS:-http://localhost:3000}"
TIMESTAMP=$(date +%Y%m%d-%H%M%S)
OUTDIR="test-results/test-seo/$TIMESTAMP"
mkdir -p "$OUTDIR/screenshots" "$OUTDIR/snapshots" "$OUTDIR/traces"
Before running, verify prerequisites per references/prerequisites.md.
Parallel Execution — 3 Subagents
Agent 1: Meta (-s=seo-meta)
Tests meta tags and head content.
playwright-cli -s=seo-meta open "$URL"
playwright-cli -s=seo-meta tracing-start
playwright-cli -s=seo-meta snapshot --filename="$OUTDIR/snapshots/seo-meta-initial.yaml"
Test Cases:
| TC | Test | How to Check |
|---|
| TC-SEO1 | <title> present, unique, 50-60 chars | eval "document.title" — verify non-empty, length in range |
| TC-SEO2 | <meta name="description"> 150-160 chars | run-code to get meta description content, check length |
| TC-SEO3 | Open Graph tags present | run-code to check for og:title, og:description, og:image, og:url meta tags |
| TC-SEO4 | Twitter Card tags present | run-code to check for twitter:card, twitter:title, twitter:description meta tags |
| TC-SEO5 | Canonical URL | run-code to find <link rel="canonical">, verify href is valid |
| TC-SEO6 | Favicon present | run-code to find <link rel="icon"> or <link rel="shortcut icon"> |
| TC-SEO7 | Viewport meta tag set correctly | run-code to check <meta name="viewport"> content includes width=device-width |
| TC-SEO8 | <html lang> attribute | eval "document.documentElement.lang" — verify present |
| TC-SEO9 | hreflang tags for multi-language | run-code to find <link rel="alternate" hreflang="..."> tags if multi-language site |
playwright-cli -s=seo-meta tracing-stop
playwright-cli -s=seo-meta close
Agent 2: Content (-s=seo-content)
Tests content structure and semantics for SEO.
playwright-cli -s=seo-content open "$URL"
playwright-cli -s=seo-content tracing-start
Test Cases:
| TC | Test | How to Check |
|---|
| TC-SEO10 | Single <h1> per page | run-code: document.querySelectorAll('h1').length === 1 |
| TC-SEO11 | Heading hierarchy (no skips) | run-code to get all headings in order, verify no level gaps (h1→h3 without h2) |
| TC-SEO12 | Images have descriptive alt text | run-code to find images without alt or with empty/generic alt |
| TC-SEO13 | Internal links use descriptive anchor text | run-code to find internal links with generic text ("click here", "here", "read more") |
| TC-SEO14 | No broken internal links | run-code to collect internal links, fetch each with HEAD, check for 404s |
| TC-SEO15 | Structured data (JSON-LD) present + valid | run-code to find <script type="application/ld+json">, parse JSON, check for @context and @type |
| TC-SEO16 | Breadcrumb markup | run-code to find breadcrumb elements (nav with aria-label="breadcrumb" or BreadcrumbList schema) |
| TC-SEO17 | Content-to-HTML ratio | run-code to measure text content length vs total HTML length, flag if ratio < 10% |
playwright-cli -s=seo-content tracing-stop
playwright-cli -s=seo-content close
Agent 3: Technical (-s=seo-technical)
Tests technical SEO and crawlability.
playwright-cli -s=seo-technical open "$URL"
playwright-cli -s=seo-technical tracing-start
Test Cases:
| TC | Test | How to Check |
|---|
| TC-SEO18 | robots.txt accessible | run-code to fetch {origin}/robots.txt, check for 200 status and valid content |
| TC-SEO19 | sitemap.xml accessible and valid | run-code to fetch {origin}/sitemap.xml, check for 200 status and valid XML |
| TC-SEO20 | Correct HTTP status codes | Check current page returns 200, test known pages for proper status codes |
| TC-SEO21 | No noindex on important pages | run-code to check for <meta name="robots" content="noindex"> — flag if found on main pages |
| TC-SEO22 | Clean URLs (no query string mess) | Check URL structure is clean: no excessive query params, no session IDs in URLs |
| TC-SEO23 | Mobile-friendly viewport | Verify viewport meta tag has width=device-width, initial-scale=1 |
| TC-SEO24 | Page loads under 3 seconds | run-code to check performance.timing.loadEventEnd - performance.timing.navigationStart < 3000 |
| TC-SEO25 | No duplicate content indicators | run-code to check canonical tag matches current URL, no duplicate title/description |
| TC-SEO26 | 301 redirects (not 302) for permanent moves | run-code to follow redirect chain, verify permanent redirects use 301 |
| TC-SEO27 | HTTPS used throughout | Verify page URL is HTTPS, no mixed content (HTTP resources on HTTPS page) |
playwright-cli -s=seo-technical tracing-stop
playwright-cli -s=seo-technical close
Fix Loop
After execution, if any tests failed, follow the fix loop process in references/fix-loop.md.
Max 3 attempts. Only re-run failed test cases.
Include fix attempt history in the report.
Report Generation
After all 3 agents complete, merge results into $OUTDIR/report.md.
Include an SEO scorecard:
| Category | Score | Issues |
|---|
| Meta Tags | X/9 | ... |
| Content | X/8 | ... |
| Technical | X/10 | ... |
| Overall | X/27 | ... |
Group results:
- Meta Tags (TC-SEO1 through TC-SEO9)
- Content Structure (TC-SEO10 through TC-SEO17)
- Technical SEO (TC-SEO18 through TC-SEO27)
Cleanup
playwright-cli -s=seo-meta close 2>/dev/null
playwright-cli -s=seo-content close 2>/dev/null
playwright-cli -s=seo-technical close 2>/dev/null
