Site Auditor

Comprehensive website audit for SEO, broken links, and accessibility.

Prerequisites

CF Browser MCP server must be configured (see the CF Browser quick start).

Workflow

Step 1: Crawl Site

browser_crawl(url, limit=50)
→ browser_crawl_status(job_id, wait=True)
→ Get list of all discovered pages

Step 2: Quick Analysis from Crawl Data

The crawl records include URL, record status, and page metadata such as final URL, title, and HTTP status. Parse only fields actually present before making extra calls:

• metadata.title → check for missing/duplicate titles
• metadata.status → identify 4xx/5xx error pages
• metadata.url versus record URL → identify a final-URL change; do not claim a full redirect chain without separate evidence

Step 3: Deep Scrape (only pages with issues)

For pages where crawl metadata is insufficient, use browser_scrape to extract:

• h1, h2, h3 headings (hierarchy check)
• img tags (check alt attributes)
• link[rel="canonical"] canonical URL
• meta[name="robots"] directives
• meta[name="description"] (if not in crawl metadata)

Batch scrape in groups of 5 to respect rate limits.

Step 4: Check Links

Using browser_links on 3-5 key pages (homepage, major sections):

• Categorize: internal, external, resource
• Cross-reference with crawl results to find orphan pages
• Identify links pointing to 4xx/5xx pages already evidenced by crawl records
• Treat external links as unverified until each destination is separately fetched or checked

Step 5: Generate Report

Output a markdown report with sections:

#### SEO Issues

• Missing or duplicate <title> tags
• Missing or duplicate meta descriptions
• Missing h1 or multiple h1 tags
• Missing canonical URLs
• Missing OpenGraph tags

#### Link Health

• Broken internal links (404s)
• Broken external links (only when separately verified)
• Redirect chains (only when hop-by-hop evidence is available)
• Orphan pages

#### Accessibility

• Images without alt text
• Missing form labels
• Missing lang attribute

#### Structure

• Sitemap coverage
• Page depth analysis
• URL structure issues

Each issue includes:

• Severity: Critical / Warning / Info
• Page: URL where issue was found
• Fix: Specific action to take

Output

Markdown report saved to audit-report-{domain}-{date}.md in the current directory.

Examples

"Audit my website for SEO issues"
→ Crawl + analyze + generate full report

"Check for broken links on docs.example.com"
→ Crawl + link check only

"Review OpenGraph tags across the site"
→ Crawl + extract OG tags + report missing ones