⚠ For educational and demonstration purposes only.

✨ Features

• Minimal bundle size, zero runtime deps
• Multi-format builds: ESM, CJS, Global (IIFE)
• Works in modern browsers; graceful fallbacks when signals are blocked
• Includes TypeScript types

📦 Installation

npm

npm i device-unique-keygen

Yarn

yarn add device-unique-keygen

📚 API

getCurrentBrowserFingerPrint(): Promise<string>

Returns a deterministic string for the current browser instance.

Under the hood, device keygen combines:

• Audio fingerprint (OfflineAudioContext render)
• Canvas fingerprint (styled text + canvas.toDataURL())
• Baseline browser signals (UA, platform, vendor, hardware concurrency, device memory, screen width/height/color depth, devicePixelRatio, languages, timezone)
• WebGL vendor/renderer (when WEBGL_debug_renderer_info is available)

All parts are concatenated and hashed via a compact 53‑bit hash (cyrb53). If the audio path fails, the library falls back to canvas + baseline signals (+ WebGL when available). If canvas is unsupported, a stable string plus baseline signals are used to avoid global collisions.

🔍 How it works

1. Generate audio entropy using an oscillator + compressor rendered in an OfflineAudioContext, summing a slice of the buffer.
2. Generate canvas entropy by drawing styled text and reading canvas.toDataURL().
3. Collect baseline signals and optional WebGL vendor/renderer.
4. Concatenate and hash to produce a compact, deterministic identifier.

Note: anti‑fingerprinting features (e.g., Brave, Firefox ETP, Safari ITP, extensions) can normalize/obfuscate signals. Broprint.js includes additional entropy to reduce collisions, but uniqueness cannot be guaranteed in all hardened environments.

🌐 Browser support

• Modern Chromium, Firefox, and Safari
• Graceful degradation when audio/canvas/WebGL are unavailable
• Some privacy modes/extensions may change stability or uniqueness

🧰 Troubleshooting

• Audio context errors: many browsers restrict audio contexts until user interaction; call after a click.
• Identical outputs: your environment may normalize canvas/audio/WebGL; baseline signals still help differentiate.
• Types: shipped via lib/index.d.ts.

🛠️ Build & contribute

Library build (root):

npm run build

Example app lives under example/ (Create React App).

PRs and issues welcome.

📄 License

MIT