Magictools Website Analyzer

Website Analyzer
evidence before opinions

Audit one public page deeply or sample up to 25 same-origin pages for SEO, accessibility, performance, security, GEO, links, and site architecture.

Analyze server HTML and a temporary rendered browser pass.

Source + rendered evidence Page and 25-page site audits No permanent report history
example.comREPORT
example.comDesktop preview

Paste a URL to capture a desktop preview

Preview readyMulti-module scan
Screenshots, lab diagnostics, and automated checks are evidence snapshots—not ranking, security, performance, or accessibility certifications.

How it works

Three simple steps. No account or installation required.

  1. Step 1

    Enter your URL

    Choose a deep Page audit or a bounded 25-page Site audit, then enter a public URL.

  2. Step 2

    We analyze the page

    Review independent dimensions, source and rendered evidence, real previews, and the severity-ordered fix queue.

  3. Step 3

    Review your report

    Filter findings, open technical detail, or export the textual report as JSON, CSV, or a print-ready PDF.

More about this tool

Overview

Website Analyzer is an evidence-backed auditing suite for public websites. Page audit compares server HTML with a rendered browser view, captures desktop and mobile previews, runs automated accessibility checks, reports mobile lab diagnostics, and inspects SEO, security, GEO, infrastructure, social metadata, and technologies. Site audit samples up to 25 same-origin HTML pages for indexability, duplicates, broken links, sitemap coverage, content structure, and internal topology. It never blends these into one quality score: each dimension explains its evidence and unavailable checks remain unavailable. Reports are temporary, exports are created in your browser, and optional Google CrUX field data is requested only when you explicitly enable it.

What the Website Analyzer checks

Choose one deep page audit or a bounded 25-page site sample. Every report states its evidence source, limits, and unavailable checks.

Visual previewDesktop and mobile screenshots with capture context.

Temporary 1440 × 900 desktop and 390 × 844 mobile screenshots rendered with JavaScript in an isolated browser, plus a website identity card and explicit capture dimensions.

Source and rendered evidenceServer HTML and rendered DOM stay visibly separate.

Compares titles, descriptions, canonicals, headings, body text, crawlable links, structured types, and indexability without silently merging contradictory results.

Performance evidenceMobile lab diagnostics and optional CrUX field data.

Shows LCP, CLS, FCP, TBT, TTFB, requests, transfer bytes, and DOM size from one lab run. Optional phone and desktop CrUX p75 values are requested only after explicit opt-in and remain separate from lab data.

Accessibility automationRendered axe-core violations grouped by impact.

Runs bounded automated rules against the rendered page and reports violations, passes, and incomplete checks. Automation does not prove WCAG conformance.

Bounded site auditUp to 25 same-origin pages, links, and topology.

Samples crawlable HTML pages for duplication, indexability, link integrity, sitemap coverage, hreflang, content signals, status codes, and real sampled inbound links. Deep browser checks remain opt-in per page.

Social sharingOpen Graph and Twitter Card previews with fallbacks.

Approximate Open Graph and Twitter Card previews with detected values, labeled fallbacks, and safely normalized images.

MetadataTitle, description, canonical, robots, headings, and structured tags.

Title, description, canonical URL, robots directives, language, viewport, favicon, headings, image alt coverage, Open Graph, Twitter cards, and JSON-LD types.

Domain registrationRegistrar, dates, status codes, and nameservers.

Registrar, registration dates, registry status codes, and authoritative nameservers published through RDAP.

DNSA, AAAA, CNAME, MX, and TXT records.

A, AAAA, and CNAME records for the website host, plus MX and TXT records for its registered domain.

Hosting and deliveryIP, ASN, CDN, origin platform, and server software.

Connected IP, BGP origin ASN, public network registration, CDN or edge provider, likely origin platform, and visible server software.

SEO intelligenceTransparent 0–100 score across content, indexing, and technical checks.

A transparent 0–100 score for content, headings, canonical and indexability evidence, robots.txt, sitemaps, structured data, social previews, mobile signals, links, and HTTPS.

GEO / AI readinessCrawler access, semantic evidence, and optional AI files.

Reads named search, citation, user-fetch, and training crawler rules from robots.txt. Optional llms.txt, llms-full.txt, and agents.md files are factual, unweighted evidence rather than requirements.

Security intelligenceHTTPS, TLS, headers, and cookie security flags.

A transparent 0–100 score for HTTPS, certificate validity, negotiated TLS, browser security headers, and value-free cookie security flags.

What the evidence means

Page audit

A page audit examines one public URL in depth. Magictools keeps server HTML and rendered browser evidence separate, captures desktop and mobile views, and organizes technical findings into independent SEO, security, accessibility, GEO, and performance dimensions.

Site audit

A site audit is a bounded sample of up to 25 same-origin HTML pages. It follows crawlable links breadth-first, can use sitemap URLs to fill the remaining capacity, honors robots.txt, and reports duplication, indexability, broken links, sitemap coverage, content signals, and sampled site topology.

Lab and field performance

The mobile lab pass is one controlled diagnostic run and cannot measure real-user INP. Optional CrUX data contains eligible Chrome user p75 values and is fetched only after explicit opt-in. Lab and field results are never blended.

Automated accessibility checks

The rendered page is checked with axe-core against supported WCAG rule tags. Automated findings are useful evidence, but they cannot establish WCAG conformance and must be followed by keyboard, screen-reader, zoom, contrast, and task-based manual testing.

What a page report can include

A page report can include redirect, canonical, robots, hreflang, heading, link, sitemap, structured-data, social, image-alt, language, viewport, and HTTPS evidence; source-versus-rendered comparisons; accessibility violations; mobile lab metrics; optional CrUX field data; security headers, TLS and value-free cookie flags; hosting, DNS, domain and technology facts; and temporary desktop and mobile captures.

How to interpret recommendations

Start with the evidence, not the score. A noindex rule, cross-page canonical, blocked training crawler, short page, or absent schema can be intentional. Confirm the page's audience and purpose before changing it, then validate important fixes with the relevant search engine, browser, assistive technology, or security tooling.

Methodology references

Recommendations are aligned with official guidance from Google Search Central, Core Web Vitals, WCAG 2.2, RFC 9309, and the OWASP Secure Headers Project.

What this report does not cover

Magictools does not sign in, dismiss consent controls, bypass bot protection, submit forms, scan ports, exploit vulnerabilities, or certify rankings, security, accessibility, performance, or AI citations. Site audits are capped samples, screenshots can vary by capture environment, automated accessibility checks are incomplete by nature, and third-party field data may be unavailable.

How to use the findings

Work from the severity-ordered queue, then inspect the stated source—server HTML, rendered DOM, lab, field, infrastructure, or crawl. Use the affected-page links to run a deep page audit where a sitewide signal needs closer evidence. Exported JSON and CSV stay in your browser and intentionally omit screenshot data URLs.

Frequently asked questions

What is the difference between Page audit and Site audit?

Page audit deeply inspects one URL using server HTML, a rendered browser, desktop and mobile captures, automated accessibility checks, lab performance evidence, infrastructure, SEO, GEO, and security. Site audit samples up to 25 same-origin HTML pages for indexability, duplicates, links, sitemap coverage, and architecture; deep browser diagnostics still run only when you open a page audit.

Does the analyzer give one overall website score?

No. SEO and security keep separate published weighted scores. Accessibility is explicitly labeled as an automated-check score. GEO reports a state from crawlability and content evidence. Lab and field performance stay separate. Technologies, hosting, DNS, domain records, and screenshots are factual and unscored.

Does automated accessibility testing prove WCAG conformance?

No. The rendered browser pass uses axe-core rules associated with WCAG levels, but automation finds only some accessibility problems. The report includes violations and incomplete checks to guide a manual review; it is not a WCAG certification.

Are the performance results Core Web Vitals?

The mobile lab run reports LCP, CLS, FCP, TBT, TTFB, requests, transferred bytes, and DOM size from one controlled capture. Lab TBT is not INP and lab results are not real-user Core Web Vitals. When you explicitly enable optional CrUX field data, eligible LCP, INP, and CLS p75 values are shown separately.

What happens when I enable Google CrUX field data?

If the deployment has a CrUX API key, Magictools sends the submitted public URL to Google only after you opt in. It requests phone and desktop URL-level data, then origin-level data when exact-URL data is unavailable. The rest of the audit can finish if CrUX is unavailable or times out.

How is the 25-page site audit bounded?

The crawler stays on the final origin, honors robots.txt, ignores discovered query variants and fragments, skips non-HTML resources and nofollow discovery paths, and uses bounded requests, redirects, response sizes, concurrency, time, and sampled external-link checks. It is a representative sample, not proof that every site URL was checked.

What does the GEO section measure?

It reports robots.txt access for named search, citation, user-fetch, and training crawlers plus indexability, semantic, entity, citation, and structured-data evidence. Optional llms.txt, llms-full.txt, and agents.md files are shown without reducing a core score when absent. The report does not predict AI citations or rankings.

Does Website Analyzer perform a vulnerability scan?

No. It reviews TLS details, response security headers, value-free cookie flags, redirects, and observable mixed-content evidence. It does not scan ports, exploit vulnerabilities, submit forms, authenticate, or replace a professional security assessment.

How long are reports kept?

Page HTML and visual bytes are processed for the current request and are not stored as analysis history. Completed site-audit report data can remain in ephemeral worker memory for up to 15 minutes so your browser can poll and export it; raw HTML is discarded after each page is analyzed. Worker restarts clear temporary jobs.

Can Website Analyzer inspect private or authenticated pages?

No. It accepts public HTTP and HTTPS pages only. Local, private, reserved, authenticated, and non-standard-port destinations are blocked, and redirects are revalidated. It does not dismiss consent dialogs, solve bot protection, or send your login cookies.

Website analysis · Free, no account required

One task finished. The rest of the toolkit is ready.

Browse all tools