Anything that touches plaintext is open source
The encryption engine, the daemon, and every language package can see secrets in plaintext - so all of them are MIT-licensed and auditable before any sales conversation. No NDA, no trust-us.
Kryptic's security model is designed to be verified, not believed: everything that can see a secret in plaintext is open source, and the server-side platform only ever stores ciphertext.
The encryption engine, the daemon, and every language package can see secrets in plaintext - so all of them are MIT-licensed and auditable before any sales conversation. No NDA, no trust-us.
AES-256-GCM from platform crypto libraries and proven Argon2id implementations, composed - never reimplemented. The engineering lives in the key hierarchy, envelope format, and rotation behavior.
The database only ever contains ciphertext envelopes and wrapped keys. Values are decrypted on explicit, authorized, audit-logged operations - and audit logs record actions, never values.
The daemon holds secrets in memory with a 5-minute TTL and serves them over a local OS socket that never crosses a network. Nothing is written to disk on developer machines.
The encryption engine ships with a SECURITY.md describing the full key hierarchy - what is derived from what, where each key lives, and exactly what the server can and cannot see - plus the test vectors that prove it. The daemon and all six language packages (.NET, Node.js, Python, Java, Go, Ruby) are MIT-licensed in the same organization.
Read the code on GitHubReport suspected vulnerabilities to [email protected]. We acknowledge within 48 hours, keep you informed while we fix, and credit reporters who want to be credited. Please do not open public issues for security reports.