HN Showcase
Back to browse
CipherStash Stack – Data Level Access Control in TS/JS

CipherStash Stack – Data Level Access Control in TS/JS

by dandraper·May 21, 2026·17 points·0 comments
Visit ProjectView on HN

AI Analysis

●●●BangerBig BrainWizardrySolve My Problem

Searchable encryption running queries over ciphertext faster than AWS KMS direct.

Strengths
  • •Value-level encryption with EQL allowing Postgres to filter/sort over ciphertext directly
  • •ZeroKMS key management achieving 14x speedup over direct AWS KMS calls
  • •Transparent SQL proxy enables legacy app integration with zero code changes
Weaknesses
  • •Postgres-only support limits adoption for teams using MySQL, MongoDB, or other databases
  • •Complex cryptography stack may require deep security audits before production trust
Category
Security
Target Audience

Backend developers, security engineers, SaaS founders handling sensitive data

Similar To

CryptDB · Microsoft Always Encrypted · CipherStash (previous versions)

Post Description

Hi HN, I'm Dan Draper, founder of CipherStash.

We've spent 5 years building per-value searchable encryption tech. We realized that encryption like this is incredibly powerful but the engineering required is generally not worth the effort. So we built a platform and opensource SDK to take the cry out of cryptography :)

CipherStash Stack is: * A searchable encryption SDK for TS and Postgres * Fast, key management with Zerokms (backed by HSM) * A transparent SQL proxy * Auth library and OIDC federation service * Skills for data security

Integrations include Prisma Next, Drizzle, Auth0, Clerk and Supabase.

Install and setup is done via a single npx stash init command. And can handoff implementation to your agent.

You can start by encrypting just a single field and we've got the whole process down to around 15 minutes.

Oh and its fast: most queries run in a few ms with many faster than that.

We published our full benchmarks here: https://github.com/cipherstash/benches#headline-numbers

I'd love to know what you think!

Similar Projects