Lighthouse Website Performance Monitor

Lighthouse Website Performance Monitor & Core Web Vitals Engine (Pro)

Lighthouse Website Performance Monitor is an enterprise-grade, 100% in-house autonomous web performance auditing engine, Core Web Vitals lab diagnostic suite, and site speed monitor for BrowserMesh. Built for modern 2026 web architectures, it measures real-world user experience and responsiveness without requiring any external Google API keys or third-party cloud dependencies.

The autonomous, private, and high-throughput alternative to Google PageSpeed Insights Web, Google Lighthouse CLI, WebPageTest ($150+/mo), and SpeedCurve ($1,800+/yr).

📑 Quick Navigation Table of Contents

• 🤝 Perfect Companion Ecosystem Pairing
• 🎯 Target Audience & Industry Solutions
• 🚀 Key Features & Architectural Capabilities
• 📊 Output Data Contract (Pristine 19 Top-Level Fields)
• ⚙️ Example Input Configurations
• 🏆 Industry Benchmark Comparison Table
• ❓ Frequently Asked Questions (FAQ)

🤝 Perfect Companion Ecosystem Pairing

Lighthouse Website Performance Monitor forms the ultimate performance, accessibility, and SEO tri-cluster when paired with companion BrowserMesh audit plugins:

• advanced-seo-audit: Flagship technical SEO and Generative Engine Optimization (GEO/AEO) engine — audits Princeton University citability, AI search bot access (OAI-SearchBot, Claude-SearchBot), client-side hydration parity, hreflang reciprocity, and Schema.org 2026 deprecation linting.
• accessibility-auditor: Comprehensive digital inclusion scanner — evaluates WCAG 2.1 & 2.2 Level AA/AAA, European Accessibility Act (EAA) legal compliance, live computed color contrast ratios (4.5:1 / 3:1), and keyboard focus traps.
• third-party-script-detector: Deep network forensics tool — tracks analytics tags, marketing beacons, trackers, and CSP security compliance.

🎯 Target Audience & Industry Solutions

• Frontend Engineers & Performance Architects (Next.js, Nuxt, Remix, Astro, SvelteKit): Pinpoint exact bottleneck vectors with 4-phase LCP decomposition and 3-phase INP interaction breakdowns. Eliminate client-side hydration lag and main-thread blocking.
• Technical SEO Directors & Web Agencies: Run batch site-wide audits across client domains via XML Sitemaps (.xml and .xml.gz). Prevent Core Web Vitals ranking degradation before Google search algorithms trigger penalization.
• E-Commerce & Shopify Plus Store Owners: Enforce strict performance budgets on total page weight and JavaScript payloads. Maximize conversion rates by maintaining sub-2.5s LCP and sub-200ms INP.
• Enterprise DevOps & CI/CD Teams: Integrate automated site performance regressions checks into build pipelines with calibrated 0–100 health scoring and automated catastrophic failure ceiling caps.

🚀 Key Features & Architectural Capabilities

1. 100% In-House Autonomous Engine (Zero External API Keys)

• Runs entirely on host Playwright and Chrome DevTools Protocol (CDP) headless browsers.
• No Google PageSpeed Insights API key, no quota rate limits, no third-party data sharing, and 100% offline air-gapped capability.

2. Official 2026 Core Web Vitals Standards

• Interaction to Next Paint (INP): Permanent Core Web Vital replacing legacy FID ($\le 200\text{ms}$ Good threshold).
• Largest Contentful Paint (LCP): Precise measurement of the largest viewport render candidate ($\le 2.5\text{s}$ Good threshold).
• Cumulative Layout Shift (CLS): Unprompted layout instability score ($\le 0.10$ Good threshold).
• Time to First Byte (TTFB) & First Contentful Paint (FCP): Backend latency ($\le 800\text{ms}$) and visual population ($\le 1.8\text{s}$).
• Total Blocking Time (TBT): Lab proxy for user interactivity and main-thread CPU starvation ($\le 200\text{ms}$).

3. LCP 4-Phase Decomposition (The 2026 Diagnostic Standard)

Pinpoints the exact micro-phase responsible for sluggish LCP rendering:

• Phase 1: Time to First Byte (TTFB): Server processing and initial HTML network transmission ($\le 800\text{ms}$).
• Phase 2: Resource Load Delay: Time between initial HTML receipt and when the browser begins downloading the LCP hero asset ($< 10%$ target).
• Phase 3: Resource Load Duration: Network transfer duration of the hero image/video poster.
• Phase 4: Element Render Delay: Time from asset download completion until pixels are visually painted on display.

4. INP 3-Phase Interaction Latency Breakdown

Breaks responsiveness into three actionable sub-components:

• Input Delay: Wait time from user physical click/tap until event listener execution begins.
• Processing Duration: Execution runtime of event handler callbacks.
• Presentation Delay: Time between event completion and the next physical screen display frame.

5. Long Animation Frames (LoAF) API (Chromium 123+)

• Diagnostic successor to legacy longtask.
• Flags any frame taking $> 50\text{ms}$ to execute with exact script attribution (sourceURL, functionName, style/layout recalculation duration).

6. Speculation Rules API & Modern Prerendering

• Inspects <script type="speculationrules"> for next-generation prerender and prefetch directives.
• Validates state-altering mutation exclusions to ensure safe instant navigation.

7. Back/Forward Cache (bfcache) Eligibility

• Detects anti-patterns that revoke instant browser back/forward restorations:
  • Legacy window.onunload event listeners (recommends pagehide).
  • Active WebSocket connections during navigation.
  • HTML responses served with Cache-Control: no-store.

8. WCAG 2.2 Target Size Standards (SC 2.5.8)

• Audits interactive targets (buttons, links, form inputs) against the minimum 24 by 24 CSS pixels touch target standard.

9. Google Lighthouse 13.5.0 Agentic Browsing Category (3/3 Audits)

Audits the website for modern autonomous AI agent discoverability across 3 dedicated checks:

1. Accessibility Tree is Well-Formed: Validates semantic heading hierarchies, landmark tags, and accessible control names so autonomous AI agents can reliably interpret and interact with the page.
2. Cumulative Layout Shift (CLS): Verifies that viewport stability ($\le 0.10$) prevents UI mis-clicks by AI browsing agents.
3. llms.txt Follows Recommendations: Checks for a properly structured Markdown document at /llms.txt containing at least one # H1 heading describing site context and resources for LLM web crawlers.

10. Authentic Mobile CDP Throttling (Google Lighthouse 13.5.0 Emulation)

• Implements real Chrome DevTools Protocol (CDP) session throttling for the mobile strategy:
  • 4x CPU Slowdown (Emulation.setCPUThrottlingRate: 4): Accurately reflects mobile JavaScript parse/compile times, yielding true-to-life Total Blocking Time (TBT).
  • Slow 4G Network Profile (Network.emulateNetworkConditions): 1.6 Mbps download throughput, 750 Kbps upload, and 150ms round-trip latency.
  • Can be toggled on/off via the emulateThrottling configuration parameter.

11. Calibrated 0–100 Health Score & Log-Normal Formulation

• Mathematically replicates official Google Lighthouse log-normal curves:
  • Largest Contentful Paint: 25%
  • Total Blocking Time: 30%
  • Cumulative Layout Shift: 25%
  • First Contentful Paint: 10%
  • Speed Index: 10%
• Abramowitz & Stegun polynomial approximation of error function erf(z) with 90+ score boost.
• Catastrophic ceiling caps:
  • HTTP 4xx/5xx: score capped at 0 (Grade F, status critical).
  • LCP $> 8,000\text{ms}$: score capped at 35 (Grade F, status critical).
  • TBT $> 1,500\text{ms}$: score capped at 45 (Grade F, status critical).
  • JavaScript $> 3,000\text{ KB}$: score capped at 55 (Grade D, status poor).
  • CLS $> 0.50$: score capped at 60 (Grade D, status poor).
  • Extreme budget breaches ($3+$ limits): score capped at 65 (Grade C, status needs_improvement).

12. Performance Budget Assertions

• Evaluates real measurements against configurable threshold limits:
  • LCP Budget Limit (budgetMaxLcpMs, default 2,500ms).
  • JavaScript Size Budget (budgetMaxJsKb, default 450 KB).
  • Total Page Weight Budget (budgetMaxTotalKb, default 2,000 KB).
• Produces exact breach logs with overage quantities.

13. Prioritized Opportunities & Byte-Efficiency Savings

• Prioritizes high-impact optimizations sorted by estimated savings in milliseconds and bytes:
  • Eliminate render-blocking resources.
  • Reduce unused JavaScript and CSS.
  • Serve modern next-gen image formats (AVIF / WebP).
  • Defer offscreen imagery and replace animated GIFs.

14. Third-Party Script & Tracker Profiling

• Automatically identifies and groups external vendors into categorized breakdowns (Analytics, Tag Managers, Advertising, Social, Utilities) with cumulative byte weight.

15. XML & Plaintext Sitemap Batch Auditing with Native Gunzip

• Automatically detects .xml, .xml.gz, and plaintext .txt sitemaps, with auto-discovery from robots.txt.
• Recursively expands <sitemapindex> child trees.
• Decompresses gzipped sitemaps using native node:zlib.

16. Digital Carbon Footprint & ESG Sustainability (SWD Model v4)

• Employs the official Sustainable Web Design (SWD v4) methodology ($0.81\text{ kWh/GB}$, $442\text{ g CO}_2\text{/kWh}$).
• Reports exact grams of $\text{CO}_2$ per visit, environmental letter grades (A+ to F), and global website percentile ratings.

17. Enterprise Security Headers Analysis

• Inspects HTTP response headers for enterprise defensive compliance:
  • HTTP Strict Transport Security (Strict-Transport-Security / HSTS).
  • Content Security Policy (Content-Security-Policy / CSP).
  • MIME-Sniffing Prevention (X-Content-Type-Options: nosniff).
  • Clickjacking Protection (X-Frame-Options: SAMEORIGIN / DENY).
  • Modern Referrer Policy (Referrer-Policy).

18. Frontend Framework & JavaScript Library Detection

• Audits client-side scripts to identify frontend libraries and frameworks (React, Next.js, Vue, Tailwind CSS, Bootstrap, jQuery, GSAP) matching the DoBetterWeb web library database.

19. Comprehensive AI Search Bot & Crawler Directives

• Evaluates robots.txt access permissions across the top 8 AI bot user agents: GPTBot, ClaudeBot, PerplexityBot, DeepSeekBot, Google-Extended, Meta-ExternalAgent, Applebot-Extended, and Bytespider.

📊 Output Data Contract (Pristine 19 Top-Level Fields)

{
  "url": "https://stripe.com",
  "title": "Stripe | Financial Infrastructure for the Internet",
  "score": 92,
  "grade": "A",
  "status": "optimal",
  "strategy": "mobile",
  "loadTimeMs": 2180,
  "scores": {
    "performance": 91,
    "accessibility": 95,
    "bestPractices": 100,
    "seo": 98,
    "agenticBrowsing": 100
  },
  "coreWebVitals": {
    "lcpMs": 2140,
    "lcpStatus": "good",
    "inpMs": 85,
    "inpStatus": "good",
    "cls": 0.024,
    "clsStatus": "good",
    "fcpMs": 1120,
    "fcpStatus": "good",
    "ttfbMs": 210,
    "ttfbStatus": "good",
    "tbtMs": 115,
    "speedIndexMs": 2250,
    "lcpBreakdown": {
      "ttfbMs": 210,
      "loadDelayMs": 95,
      "loadDurationMs": 1650,
      "renderDelayMs": 185
    },
    "inpBreakdown": {
      "inputDelayMs": 18,
      "processingDurationMs": 42,
      "presentationDelayMs": 25
    }
  },
  "fieldData": {
    "available": true,
    "overallCategory": "FAST",
    "originFallback": false,
    "lcpMs": { "percentile": 2100, "category": "FAST" },
    "inpMs": { "percentile": 92, "category": "FAST" },
    "cls": { "percentile": 0.02, "category": "FAST" },
    "fcpMs": { "percentile": 1100, "category": "FAST" },
    "ttfbMs": { "percentile": 220, "category": "FAST" }
  },
  "resourceSummary": {
    "totalBytes": 1425890,
    "totalRequests": 48,
    "jsBytes": 385400,
    "cssBytes": 68200,
    "imageBytes": 842100,
    "fontBytes": 85400,
    "documentBytes": 32400,
    "otherBytes": 12390,
    "thirdPartyBytes": 142000,
    "thirdPartyRequests": 9
  },
  "budgetStatus": {
    "passed": true,
    "totalBudgets": 3,
    "failedBudgets": 0,
    "breaches": []
  },
  "opportunities": [
    {
      "id": "render-blocking-resources",
      "title": "Eliminate render-blocking resources",
      "savingsMs": 340,
      "savingsBytes": 45200,
      "displayValue": "Potential savings of 340 ms",
      "description": "Resources are blocking the first paint of your page. Consider delivering critical JS/CSS inline and deferring all non-critical JS/styles.",
      "fix": "Add defer or async attribute to non-critical script tags in <head> and preload critical fonts/CSS."
    }
  ],
  "diagnostics": {
    "domNodes": 842,
    "domDepth": 14,
    "renderBlockingScriptsCount": 0,
    "renderBlockingCssCount": 1,
    "heroLazyLoaded": false,
    "heroMissingFetchPriority": false,
    "missingDimensionsCount": 0,
    "thirdPartyVendors": [
      "Google Tag Manager",
      "Cloudflare Insights"
    ],
    "longAnimationFramesCount": 1,
    "criticalRequestChainsCount": 3,
    "criticalPathLatencyMs": 420,
    "speculationRulesDetected": true,
    "bfcacheEligible": true,
    "targetSizeViolationsCount": 0,
    "httpProtocol": "h2",
    "redirectsCount": 0,
    "userTimingsCount": 4,
    "preconnectOrigins": [
      "https://js.stripe.com"
    ],
    "carbonFootprint": {
      "co2GramsPerVisit": 0.082,
      "ecoGrade": "A+",
      "cleanerThanPercent": 95
    },
    "securityHeaders": {
      "hsts": true,
      "csp": true,
      "xContentTypeOptions": true,
      "xFrameOptions": true,
      "referrerPolicy": "strict-origin-when-cross-origin"
    },
    "detectedLibraries": [
      {
        "name": "React",
        "category": "UI Framework"
      }
    ],
    "pastePreventingInputsCount": 0,
    "agenticReadiness": {
      "llmsTxtFound": true,
      "aiCatalogFound": true,
      "ardManifestFound": false,
      "webMcpToolsCount": 4,
      "aiBotDirectives": {
        "gptBotAllowed": true,
        "claudeBotAllowed": true,
        "perplexityBotAllowed": true,
        "deepSeekBotAllowed": true,
        "googleExtendedAllowed": true,
        "metaExternalAllowed": true,
        "applebotExtendedAllowed": true,
        "bytespiderAllowed": true
      }
    }
  },
  "thirdPartyBreakdown": [
    {
      "vendor": "Google Tag Manager",
      "category": "tag-manager",
      "requestCount": 4,
      "transferBytes": 86400
    }
  ],
  "issues": [
    {
      "id": "render-blocking-css",
      "severity": "warning",
      "metric": "FCP",
      "title": "1 render-blocking stylesheet detected",
      "description": "External stylesheet /assets/main.css blocks the first contentful paint by 180ms.",
      "fix": "Incorporate critical styles into an inline <style> block and load the stylesheet asynchronously.",
      "location": "link[rel='stylesheet'][href*='main.css']"
    }
  ],
  "recommendations": [
    "Inline critical CSS and load secondary stylesheets asynchronously to recover ~340ms FCP.",
    "Maintain LCP hero image fetchpriority='high' to safeguard sub-2.5s Core Web Vitals rating."
  ],
  "timings": {
    "dnsLookupMs": 24,
    "tcpHandshakeMs": 48,
    "tlsNegotiationMs": 62,
    "responseDownloadMs": 76,
    "domInteractiveMs": 620,
    "domContentLoadedMs": 850,
    "domCompleteMs": 1940,
    "loadEventMs": 2180
  },
  "auditMeta": {
    "engine": "in_house_cdp",
    "lighthouseVersion": "13.5.0-compat",
    "userAgent": "Mozilla/5.0 (Linux; Android 11; moto g power (2022)) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/136.0.0.0 Mobile Safari/537.36",
    "timestamp": "2026-10-22T13:40:00.000Z"
  }
}

⚙️ Example Input Configurations

Input Settings Specification

Example 1: Single URL Audit (Default Mobile Strategy, Budgets Enforced)

{
  "url": "https://stripe.com",
  "strategy": "mobile",
  "mode": "single",
  "checkBudgets": true,
  "budgetMaxLcpMs": 2500,
  "budgetMaxJsKb": 450,
  "budgetMaxTotalKb": 2000,
  "emulateThrottling": true,
  "timeout": 45
}

Example 2: Full Site-Wide Batch Monitoring via XML Sitemap (Dual Form Factor)

{
  "url": "https://stripe.com/sitemap.xml",
  "strategy": "both",
  "mode": "sitemap",
  "maxPages": 25,
  "checkBudgets": true,
  "budgetMaxLcpMs": 2200,
  "budgetMaxJsKb": 400,
  "budgetMaxTotalKb": 1800,
  "emulateThrottling": true,
  "concurrency": 2,
  "timeout": 60
}

🏆 Industry Benchmark Comparison Table

❓ Frequently Asked Questions (FAQ)

1. Does this plugin require a Google Cloud API Key or PageSpeed Insights credentials?

No, absolutely not. This engine is 100% autonomous and self-contained. It executes directly inside host Playwright browsers using native Chrome DevTools Protocol (CDP) Performance Timeline APIs, completely independent of external cloud APIs or third-party quotas.

2. How are the Performance category score and Core Web Vitals calculated?

The engine mathematically replicates official Google Lighthouse log-normal distribution curves using the Gauss error function erf(z) and Abramowitz & Stegun polynomial approximations. The weighted scoring adheres to official Lighthouse standards: LCP (25%), TBT (30%), CLS (25%), FCP (10%), and Speed Index (10%).

3. What is the difference between INP and TBT in this audit?

Interaction to Next Paint (INP) observes real user interaction latency. Total Blocking Time (TBT) serves as the mathematically validated laboratory surrogate for INP during synthetic page loads. The plugin calculates lab TBT and models 3-phase INP interaction components (input delay, processing duration, and presentation delay).

4. Can I crawl an entire website using an XML Sitemap?

Yes. Provide any standard .xml, .xml.gz, or plaintext .txt sitemap URL. The plugin automatically traverses root <sitemapindex> feeds, decompresses gzipped archives using native Node.js zlib, and evaluates up to maxPages URLs in parallel. Auto-discovery from robots.txt is also included.

5. What are catastrophic failure ceiling caps?

Ceiling caps prevent misleadingly high composite scores on technically broken websites. Regardless of other factors, severe errors trigger hard caps: HTTP 4xx/5xx (0 score), LCP $> 8\text{s}$ (max 35), TBT $> 1.5\text{s}$ (max 45), JavaScript $> 3\text{ MB}$ (max 55), and CLS $> 0.50$ (max 60).

6. What is the 5th Agentic Browsing category in Lighthouse 13.5.0?

Lighthouse 13.5.0 introduces an explicit Agentic Browsing audit category that benchmarks how reliably autonomous AI agents (such as OpenAI Operator, Claude Computer Use, or Perplexity) can parse, navigate, and interact with the page. It audits accessibility tree cleanliness, viewport stability (CLS $\le 0.10$), and structured root /llms.txt crawler context.

7. How does the SWD v4 Digital Carbon Footprint score work?

The audit applies the official Sustainable Web Design Model (version 4) standard ($0.81\text{ kWh/GB}$, $442\text{ g CO}_2\text{/kWh}$). It calculates exact grams of $\text{CO}_2$ generated per page visit, assigns an international ESG letter grade (A+ to F), and determines the website's cleanliness percentile against global web benchmarks.