Skip to content

Security: Inpriv/inpriv

Security

SECURITY.md

Security Policy

Reporting a Vulnerability

Inpriv is a privacy-first project — we take security reports seriously.

Do NOT open a public GitHub issue for security vulnerabilities.

Instead, please report privately:

  1. Email: security@inpriv.xyz
  2. Subject: [SECURITY] Inpriv — <brief description>
  3. Include: description, steps to reproduce, impact assessment, and suggested fix (if any).

Response Timeline

Stage Target
Acknowledgement Within 48 hours
Initial assessment Within 5 business days
Fix or mitigation Within 30 days (severity-dependent)
Public disclosure After fix is released, coordinated with reporter

Security Model

Inpriv tools are designed around zero-knowledge architecture:

  • Client-side only — all cryptographic operations happen in your browser via the Web Crypto API
  • No servers processing your data — static HTML/JS served from Cloudflare's edge
  • No telemetry, analytics, or tracking — verified by design, not by promise
  • No accounts or databases (except Hush signaling and Burn ephemeral notes, which store only encrypted blobs)

Cryptographic Primitives

Purpose Algorithm
Key exchange X25519 (ECDH)
Symmetric encryption AES-256-GCM
Key derivation HKDF-SHA-256, PBKDF2 (100k iterations)
Wallet key derivation BIP39 → BIP44 (Ed25519)
Wallet encryption AES-256-GCM + Argon2id
Randomness crypto.getRandomValues() / crypto.subtle.generateKey()

Modules with Server Components

Module What the server sees What it never sees
Hush (E2E chat) Encrypted message routing, signaling Plaintext, keys, metadata content
Burn (ephemeral notes) Encrypted note blobs (auto-deleted) Plaintext, encryption keys
OSINT (intelligence engine) Query routing via SearXNG Personal data about the user

Dependency Policy

  • Minimize external dependencies — prefer vanilla browser APIs
  • All dependencies are pinned in lockfiles
  • Python backends use pinned versions in requirements.txt
  • No CDN-loaded scripts at runtime (self-hosted fonts and assets only)

Disclosure Credits

We credit security researchers who report valid vulnerabilities (unless they prefer to remain anonymous).


Built with paranoia. © 2026 Inpriv Labs.

There aren't any published security advisories