Skip to content

Security: travisjneuman/neumanos

Security

SECURITY.md

Security Policy

Supported Versions

Version Supported
2.0.x Yes
1.5.x Yes
< 1.5 No

Reporting a Vulnerability

If you discover a security vulnerability in NeumanOS, please report it responsibly.

Email: os@neuman.dev

Please include:

  • A description of the vulnerability
  • Steps to reproduce
  • Potential impact
  • Any suggested fixes (optional)

You will receive an acknowledgment within 48 hours. We aim to provide a fix or mitigation within 7 days for critical issues.

Security Architecture

NeumanOS is a local-first application with a zero-data-host architecture:

  • No backend server — all data stored locally in IndexedDB
  • No accounts or authentication — nothing to breach
  • BYOK (Bring Your Own Key) — AI provider API keys are encrypted with AES-256-GCM using PBKDF2 (600k iterations) before storage
  • Encryption passwords are never persisted — they exist only in memory during the session
  • Content Security Policy — strict CSP headers enforced via Cloudflare Pages _headers file

Known Limitations

  • unsafe-eval is required in CSP for WebContainers (Phantom Shell feature)
  • AI provider API keys transit the browser network when making API calls (inherent to BYOK browser apps)
  • Dependency audit status after the April 29, 2026 maintenance pass: npm audit --audit-level=low reports 0 vulnerabilities.
  • Third-party dependency reports remain out of scope when there is no exploitable path in NeumanOS or the issue is browser/runtime-level rather than application-level.

Scope

The following are in scope for security reports:

  • XSS, injection, or data exfiltration vulnerabilities
  • Encryption implementation weaknesses
  • Privacy violations (data leaving the device without user action)
  • CSP bypass vectors

The following are out of scope:

  • Vulnerabilities requiring physical access to the user's device
  • Issues in third-party dependencies with no exploitable path in NeumanOS
  • Browser-level vulnerabilities

There aren't any published security advisories