@ottoai/browser-dna

One script. Complete visitor intelligence. You own the data.

Live demo: https://otto-ai-labs.github.io/browser-dna/

Drop-in browser intelligence for any website. Collect visitor identity, device, behavior, creator, and AI signals in a single payload.

No API key. No automatic network requests. No storage writes. Call scan(), get data, do whatever you want with it.

Works as a <script> tag (drop on any site, zero setup) or as an npm/ESM package (tree-shakeable, TypeScript-friendly).

Table of Contents

• Quick Start
• Script Tag Embed
• npm / ESM Usage
• Signals Collected
• Payload Shape
• Creator Score
• Config Reference
• Individual Collectors
• Build
• Security & Privacy
• Contributing

Quick Start

Option A — Script tag (fastest, zero install)

<script src="https://unpkg.com/@ottoai/browser-dna/dist/browser-dna.js"></script>

<script>
  window.BrowserDNA.ready(data => {
    console.log(data.creatorScore)
    // do whatever you want with data
  })
</script>

Option B — npm package

npm install @ottoai/browser-dna

import { scan } from '@ottoai/browser-dna'

const data = await scan()

console.log(data.creatorScore)   // { score: 72, tier: 'active', reasons: [...] }
console.log(data.device.browser) // 'chrome'
console.log(data.social)         // { inAppBrowser: null, referrer: { platform: 'Twitter/X', ... } }

Script Tag Embed

<script
  src="https://unpkg.com/@ottoai/browser-dna/dist/browser-dna.js"
  data-trigger="load"
  data-debug="false"
></script>

Receiving the data

// Callback — fires as soon as collection completes
window.BrowserDNA.ready(data => {
  console.log(data)
})

// Manual scan — returns a promise
const data = await window.BrowserDNA.scan()

Triggers

| data-trigger | Behavior | |--------------------|----------| | load (default) | Fires once when the page loads | | pageview | Fires on load + every SPA navigation (pushState / replaceState / popstate) | | manual | Does not fire automatically — call window.BrowserDNA.scan() yourself |

npm / ESM Usage

scan(options?) — collect all signals, return payload

import { scan } from '@ottoai/browser-dna'

const data = await scan()
// data.fingerprint, data.device, data.social, data.ai, data.creator, ...

// With debug logging
const data = await scan({ debug: true })

Signals Collected

social — Social platform signals

• inAppBrowser — detect Facebook, Instagram, TikTok, LinkedIn, Snapchat, WhatsApp, Pinterest, Telegram from UA string
• referrer.platform — map document.referrer hostname to platform (Twitter/X, Facebook, LinkedIn, Reddit, Discord, YouTube, WhatsApp, Telegram, and 30+ more)
• referrer.raw — original referrer URL

ai — AI tool signals

• referrer — detect ChatGPT, Claude, Perplexity, Gemini, Copilot, Grok, Meta AI, You.com, Phind, Poe from referrer + utm_source
• extensions — detected AI browser extensions via window globals and DOM selectors

extensions — Browser extension detection

• Scans for 20+ known extensions: AI assistants, creator tools (vidIQ, TubeBuddy, Loom), productivity tools
• Detection via window globals and DOM selectors (MV3-compatible)

fingerprint — Visitor identity

• id — stable ThumbmarkJS fingerprint hash (~90% accuracy, MIT)

device — Device and browser signals

• browser, browserVersion, os — parsed from UA string
• isBot, isWebView, isInAppBrowser
• ip, country, city, region, org — IP geolocation via ipapi.co
• timezone, language, userAgent, screen, deviceMemory, cpuCores, connection

hardware — Hardware-tier signals

• gpu, gpuVendor, gpuTier — WebGL renderer + tier
• displayType — ultrawide, 4k-retina, retina-laptop, tablet, mobile
• audioInputs, videoInputs, hasProAudio, hasProVideo
• sampleRate, maxChannelCount, baseLatency, audioTier — AudioContext signals
• p3ColorGamut, hdrDisplay, darkMode — CSS media query signals
• fonts — canvas-detected fonts by category (adobe-suite, premium-fonts, developer, system)
• micPermission, cameraPermission, hasRecordingAccess — Permissions API

creator — Creator platform signals

• referrer — 50+ creator platforms: YouTube Studio, Substack, Beehiiv, Gumroad, Stan Store, Patreon, Linktree, Beacons, etc.
• linkInBio — detected link-in-bio platform
• utmSignals — UTM parameter analysis with creator channel inference
• tier — inferred creator tier (monetized, active-creator, has-newsletter, beginner)
• storageArtifacts — localStorage keys from creator tools (vidIQ, TubeBuddy, Loom, etc.)

engagement — Page behavior

• scrollDepth — max scroll percentage (0–100)
• timeOnPage, activeTime — total and active time in seconds
• tabWasHidden, sessionHour, sessionDay

Payload Shape

{
  "url": "https://yoursite.com/pricing",
  "timestamp": "2026-04-24T10:00:00Z",
  "fingerprint": { "id": "abc123def456" },
  "creatorScore": {
    "score": 72,
    "tier": "active",
    "reasons": ["YouTube Studio referrer", "vidIQ detected", "Pro audio interface"]
  },
  "social": {
    "inAppBrowser": null,
    "referrer": { "platform": "Twitter/X", "raw": "https://t.co/xyz" }
  },
  "ai": {
    "referrer": "ChatGPT",
    "extensions": [{ "name": "Claude Extension", "vendor": "Anthropic", "method": "global" }]
  },
  "extensions": [
    { "name": "vidIQ", "vendor": "vidIQ", "method": "global" }
  ],
  "device": {
    "browser": "chrome",
    "browserVersion": "124.0.0",
    "os": "macos",
    "isBot": false,
    "ip": "41.x.x.x",
    "country": "Nigeria",
    "city": "Lagos",
    "timezone": "Africa/Lagos",
    "language": "en-US",
    "screen": "1440x900",
    "deviceMemory": 8,
    "cpuCores": 8,
    "connection": "4g"
  },
  "hardware": {
    "gpu": "Apple M2",
    "gpuTier": "apple-silicon",
    "displayType": "retina-laptop",
    "hasProAudio": true,
    "audioTier": "pro",
    "p3ColorGamut": true,
    "darkMode": true,
    "hasRecordingAccess": true
  },
  "creator": {
    "referrer": { "platform": "YouTube Studio", "type": "video", "tier": "pro" },
    "tier": "active-creator",
    "storageArtifacts": [{ "tool": "vidIQ", "confidence": "high" }]
  },
  "engagement": {
    "scrollDepth": 68,
    "timeOnPage": 42,
    "activeTime": 31,
    "tabWasHidden": false,
    "sessionHour": 10,
    "sessionDay": "Thursday"
  }
}

Creator Score

creatorScore is a 0–100 score ranking how likely a visitor is a professional creator or influencer.

| Range | Tier | Description | |-------|------|-------------| | 75–100 | pro | Strong creator signals — pro tools, creator platform referrer, recording access | | 50–74 | active | Active creator signals — creator tools, content platform referrers | | 25–49 | aspiring | Some creator signals — social referrers, content consumption behavior | | 0–24 | consumer | General visitor — minimal or no creator signals |

High-confidence signals

| Signal | Points | |--------|--------| | YouTube Studio referrer | +20 | | vidIQ or TubeBuddy detected | +20 | | Pro audio interface (AudioContext) | +15 | | Recording access granted (Permissions API) | +12 | | Creator platform referrer (Substack, Gumroad, etc.) | +10 | | Creator tool localStorage artifact | +10 | | Loom extension detected | +8 |

Config Reference

Script tag attributes

| Attribute | Default | Description | |-----------|---------|-------------| | data-trigger | load | load, pageview, or manual | | data-debug | false | Log full payload to browser console |

scan(options)

| Option | Default | Description | |--------|---------|-------------| | debug | false | Log payload to console |

Individual Collectors

Import only what you need for tree-shaking:

import { getSocialSignals }      from '@ottoai/browser-dna'
import { getAISignals }          from '@ottoai/browser-dna'
import { getDeviceSignals }      from '@ottoai/browser-dna'
import { getFingerprintSignals } from '@ottoai/browser-dna'
import { getExtensionSignals }   from '@ottoai/browser-dna'
import { getCreatorSignals }     from '@ottoai/browser-dna'
import { trackEngagement }       from '@ottoai/browser-dna'
import { getHardwareSignals }    from '@ottoai/browser-dna'
import { computeCreatorScore }   from '@ottoai/browser-dna'

Build

npm install
npm run build   # builds both outputs
npm run dev     # watch mode

Outputs

| File | Format | Use | |------|--------|-----| | dist/browser-dna.js | IIFE (BrowserDNA global) | <script> tag embed | | dist/browser-dna.esm.js | ESM | npm / bundler import |

• Bundler: esbuild
• Target: ES2018 (IIFE) / ES2020 (ESM)
• No runtime dependencies except ThumbmarkJS (MIT, bundled in)

Testing locally

Live demo: https://otto-ai-labs.github.io/browser-dna/

npx serve .
# Open http://localhost:3000/test.html

Enable data-debug="true" to log the full payload to the console.

File Structure

services/CookieReader/
├── src/
│   ├── index.js         # IIFE browser entry — exposes window.BrowserDNA
│   ├── npm.js           # ESM entry — re-exports scan and all collectors
│   ├── core.js          # scan() implementation
│   ├── social.js        # In-app browser + social referrer detection
│   ├── ai.js            # AI tool referrer + extension detection
│   ├── extensions.js    # Browser extension registry + detection
│   ├── device.js        # IP geo, UA parsing, navigator signals
│   ├── fingerprint.js   # ThumbmarkJS visitorId
│   ├── hardware.js      # GPU, display, audio, fonts, permissions
│   ├── creator.js       # Creator platform referrers, UTM, tool artifacts
│   ├── engagement.js    # Scroll depth, time on page, session signals
│   └── score.js         # Creator score (0–100) + tier
├── dist/
│   ├── browser-dna.js       # Minified IIFE build
│   └── browser-dna.esm.js   # ESM build
├── build.cjs
├── package.json
└── index.html           # Browser test page

Security & Privacy

What data is collected @ottoai/browser-dna collects publicly available browser signals — UA string, referrer, screen size, IP geolocation, and behavioral signals. It does not collect passwords, form contents, clipboard data, or any PII beyond what is passively available to any web page.

No storage side effects The library does not write to localStorage, sessionStorage, or cookies. Every scan() call is independent and stateless.

No network requests scan() makes no network requests. The script tag never POSTs anywhere on its own.

Third-party dependency ThumbmarkJS (MIT) is bundled for fingerprinting. It runs entirely client-side and makes no external requests.

GDPR / CCPA Your obligations depend on what you do with the data after calling scan(). The library itself is neutral.

Contributing

Found a bug? Want to add a new signal or platform? Contributions are welcome.

See CONTRIBUTING.md for guidelines.

License

MIT License — free to use, modify, and embed.

Built by Otto AI