v2.0 WebmasterTools 2.0 • Powered by Claude Sonnet 5.5 Reasoning Engine

Autonomous Webmastering & Technical Site Optimization at Machine Speed

Replace manual audits with autonomous AI agents that diagnose Core Web Vitals bottlenecks, audit complex sitemaps, and generate validated pull requests in seconds.

Scan Your Website Free Read Documentation
https://

Instant quick audit runs live DNS, TLS-issuance and email-authentication checks via Cloudflare DNS-over-HTTPS. Full agentic scans run in your workspace.

app.webmastertools.co/sites/acme-store

Claude Agent Console

diffproposed patch

              

            


            
            
          

        

      

      
Interactive preview with illustrative demo data — switch between LCP, CLS and INP to see the agent's reasoning and patch.

    


    
    

      

        

          
// features

          
Your site's on-call engineer, SEO lead and schema expert

          
Agents that find problems, explain them in plain engineering terms, and ship the fix as reviewable code.

        

        

          

            
            
Autonomous Code Remediation

            
Generates exact git diffs that resolve LCP, INP and CLS regressions — image priority, font loading, layout reservation, main-thread yielding.

            
LCPINPCLS

          

          

            
            
Deep Semantic & Technical SEO

            
Reasons across hundreds of pages at once in Claude's 1M-token context to untangle canonical chains, hreflang conflicts and indexing anomalies.

          

          

            
            
Structured Data & JSON-LD Engine

            
Synthesizes Schema.org-compliant rich-result markup from your real content and validates it against schema before it ships.

          

          

            
            
Agentic CMS & Git Integration

            
Multi-step tool use across GitHub, Cloudflare and headless CMS APIs — with scoped permissions and every action logged.

          

          

            
            
AI Search Optimization (GEO)

            
Structures content, entities and citations so AI answer engines — Claude, Perplexity, ChatGPT Search — can quote your pages accurately.

          

          

            
            
Continuous Uptime & Anomaly Guard

            
Checks preview deployments and alerts before a merge degrades Core Web Vitals, breaks robots rules or drops pages from the sitemap.

          

        

      

    


    
    

      

        

          
// agentic fixes

          
An agent loop you can audit

          
Every fix follows the same loop. Nothing reaches production without passing your CI and your review.

        

        
    
          
  1. 01 crawl
    Crawl & measure
    Render pages headlessly, collect lab + field vitals, DOM and network traces.
    2. 
          
  2. 02 diagnose
    Diagnose
    Claude maps each bottleneck to the exact component and line in your repo.
    3. 
          
  3. 03 patch
    Patch
    A minimal diff is written on a branch — never on your default branch.
    4. 
          
  4. 04 verify
    Verify
    Re-measure on the preview deploy; your CI runs; regressions roll back automatically.
    5. 
          
  5. 05 review
    Pull request
    Opened with before/after metrics and reasoning. A human merges.
    6. 
        

      

    


    
    

      

      

      

        

          Built on Anthropic Claude
          
Why Claude powers WebmasterTools

          
Claude Sonnet 5.5 is the reasoning core for diagnosis and code generation. Claude Haiku 4.5 handles high-volume page classification and log triage during continuous crawls.

          

            

              
              
State-of-the-art code reasoning
Frontend fixes need framework awareness — React Server Components, hydration, bundler config. Claude writes minimal, idiomatic diffs that pass review.

            

            

              
              
Reliable structured output
Crawler findings, schema markup and fix plans are returned through structured outputs and strict tool definitions, so every response parses against our JSON schema.

            

            

              
              
Prompt caching for continuous crawling
Site maps, design-system rules and repo context are cached once and reused across thousands of page analyses, keeping always-on monitoring fast and affordable.

            

          

        

        

          

            

              agent/remediate.ts
              tool use · cached
            


// Repo context is cached; each page adds only its trace
const msg = await client.messages.create({
  model: "claude-sonnet-5-5",
  max_tokens: 16000,
  system: [{ type: "text", text: repoContext,
             cache_control: { type: "ephemeral" } }],
  tools: [readFile, proposePatch, runLighthouse],
  messages: [{ role: "user",
               content: pageTrace }],
});

          

          

            
1M
token context

            
Sonnet 5.5
diagnosis & diffs

            
Haiku 4.5
crawl triage

          

        

      

    


    
    

      

        

          
// pricing

          
Start free. Scale to your fleet.

        

        

          

            
Developer

            
For side projects and evaluation.

            
$0/mo

            
    
              
  • 500 pages crawled / month
    • 
              
  • Basic technical & vitals audits
    • 
              
  • JSON-LD validator
    • 
              
  • CLI access
    • 
            

            
          

          

            Most Popular
            
Pro Team

            
For product and growth engineering teams.

            
$49/mo

            
    
              
  • 50,000 pages crawled / month
    • 
              
  • Automated GitHub PR agent
    • 
              
  • Continuous Core Web Vitals fixes
    • 
              
  • Preview-deploy regression guard
    • 
              
  • Slack & email alerts
    • 
            

            
          

          

            
Enterprise

            
For multi-domain fleets and agencies.

            
$199/mo

            
    
              
  • Multi-domain fleet management
    • 
              
  • Custom crawler proxies & regions
    • 
              
  • SLA & SSO
    • 
              
  • Dedicated Claude rate limits
    • 
            

            
          

        

      

    


    
    

      

        

          
// docs

          
From zero to first PR in three commands

          
Use the CLI locally or in CI, install the GitHub App for automated pull requests, or call the REST API from your own tooling.

          

            
            
            
          

        

        

          

            

              ~/acme-store
              
            


$ npm i -g @webmastertools/cli
$ wmt login
$ wmt scan https://acme-store.com --vitals --seo --fix-pr


✔ crawled 48,210 pages · 6m12s
▲ LCP 4.1s on /products/* — hero image lazy-loaded
▲ CLS 0.21 — promo banner without reserved space
✔ patch verified on preview: LCP 4.1s → 2.2s
→ opened PR #482 "perf: prioritize product hero image"

          

        

      

    


    
    

      

        

          
// faq

          
Technical FAQ

        

        

          

            Will crawling slow down or overload my site?
            
No. The crawler honors robots.txt and crawl-delay, adapts its rate to your server's response times, and backs off automatically on 429/5xx. You can set a hard requests-per-second cap and crawl windows per domain.

          

          

            What git permissions does the agent need? Can it push to main?
            
The GitHub App requests only repository contents and pull request access on the repos you select. It works on its own branches, never force-pushes, and cannot merge — branch protection and human review stay in control.

          

          

            How deep does an audit go?
            
Full JavaScript rendering, Lighthouse lab metrics, Chrome UX Report field data where available, internal link graph, canonical/hreflang/robots analysis, sitemap reconciliation and structured data validation — then root-cause mapping to your source code.

          

          

            Is my source code used to train AI models?
            
No. We use Anthropic's commercial API, which does not train on customer inputs or outputs by default. Repository access is read-scoped to what a fix requires, and code context is not retained beyond the job.

          

          

            Which frameworks and hosts are supported?
            
Audits work on any public site. Code remediation supports Next.js, Nuxt, Astro, SvelteKit, Remix, plain HTML/JS and major headless CMSs, with deployment checks for Vercel, Netlify and Cloudflare Pages.

          

        

      

    


    
    

      

        

        

        
Ship faster pages without the audit backlog.

        
Run a free audit now, or connect a repo and let the agent open its first pull request.

        

          Scan Your Website Free