Security at ORBTR

Security is our product, not an afterthought. Here's how we protect your infrastructure and data.

Security by design

Zero-Trust Model

Every connection is authenticated and encrypted. No implicit trust between devices, networks, or services. Identity is verified cryptographically at every layer.

Key Hierarchy

HSM-backed Platform Root, exportable Tenant Root, and device-local Agent Keys. Dual-signed Root Descriptors require both platform and tenant approval.

End-to-End Encryption

All mesh traffic uses Noise Protocol with Ed25519 key binding. Device keys are generated locally and never leave the OS keychain or TPM.

Signed Everything

Policy bundles, script manifests, update artifacts, and LAD snapshots are all cryptographically signed. Epoch-based anti-rollback prevents replay attacks.

Agent Rate Limiting

Built-in rate limits detect key compromise — policy updates capped at 10/hr, jobs at 100/hr. Platform override available for emergencies only.

Verified Updates

All agent updates are cryptographically verified with Ed25519 signatures and SHA-256 integrity checks. Staged rollouts with auto-rollback on failure thresholds.

Transport Resilience

Multipath failover across VL1, QUIC, and WebSocket transports. Automatic path selection with relay fallback using short-lived, audited tickets when direct P2P fails.

Agent Hardening

Device keys stored in OS keychain (DPAPI on Windows, Keychain on macOS). Encrypted local database, column-level encryption for sensitive fields, and HMAC integrity verification on cached state.

Compliance-Segmented Cryptography

Separate Platform Root keys per compliance domain — standard, HIPAA, FedRAMP, PCI, and IRAP. Prevents cross-contamination between regulatory boundaries at the cryptographic level.