Quark

A production-grade, zero-dependency browser fingerprinting library. Quark generates stable, deterministic identifiers while prioritizing user privacy and avoiding invasive tracking techniques.

Core Philosophy

Quark is built for security and fraud prevention, not for invasive advertising tracking. It avoids deprecated APIs and heavy footprinting methods that trigger browser warnings.

• Zero Deps: ~4KB gzipped. No external bloat.
• Privacy First: Built-in "Privacy Mode" to skip high-entropy sources like Canvas and Audio.
• GDPR-Aware: Designed for legitimate interest use-cases and easy consent integration.
• Async Execution: Parallel collection ensures execution finishes in < 50ms on modern hardware.

Installation

npm install quark-fingerprint

Quick Start

import { getFingerprint } from 'quark-fingerprint';

// Simple collection
const fp = await getFingerprint();

console.log(fp.id);         // Stable SHA-256 hash
console.log(fp.confidence); // 0.0 to 1.0 uniqueness score

Advanced Config

import { createAgent } from 'quark-fingerprint';

const agent = createAgent({
    privacyMode: true,     // Skip high-entropy collectors (Canvas, Audio, etc.)
    componentTimeout: 100, // Timeout per collector
    debug: false
});

const result = await agent.collect();

Collectors

Quark gathers signals through modular collectors. Each signal is hashed or normalized before the final fingerprint generation.

| Collector | Signals Collected | Privacy | | :--- | :--- | :--- | | Browser | UA, Client Hints, Timezone, Language, Locale | Standard | | Screen | Dimensions, Color depth, Device pixel ratio, Orientation | Low Entropy | | Hardware | Logical cores (Concurrency), Device memory, Touch support | Standard | | Graphics | WebGL Vendor/Renderer, GPU precision parameters | Standard | | Canvas | Text & Geometry rendering behavior (Hash only) | High Entropy | | Audio | AudioContext oscillator and compressor dynamics (Hash only) | High Entropy | | Fonts | OS font detection via layout narrow/wide width checks | Standard | | Permissions | Feature states (Notifications, Midi, Geolocation, etc.) | Behavioral | | Media | Input/Output device counts (Camera, Mic, Speakers) | High Entropy | | Preferences | Color scheme, Reduced motion, Inverted colors, Webdriver | Standard | | Capabilities | PDF support, Storage quota, Modern CSS feature support | Low Entropy |

Privacy & Security

Entropy vs. Privacy

Quark computes a confidence score based on the total entropy gathered. High entropy (e.g., Canvas rendering) provides high stability but can be used for cross-site tracking. In privacyMode, these sources are omitted, yielding a lower confidence score but higher user anonymity.

Legitimate Interest

Under GDPR, fingerprinting for security and fraud prevention (e.g., preventing bot attacks or account takeovers) often falls under "Legitimate Interest." However, we recommend disclosing use in your privacy policy.

Running Locally

Since the library uses ES Modules, you must serve the test files via a local server to avoid CORS issues with file:// URLs.

# Using Node
npx serve .

# Using Python
python3 -m http.server

Open test.html in your browser to see a live breakdown of your device's fingerprint.

License: MIT