Hi HN,
We built Blockrand, a randomness API for applications where “trust us” is not good enough (games, lotteries, simulations, on-chain/off-chain hybrids).
Most “provably fair” RNGs fail under closer inspection:
The server can influence outcomes
The client can brute-force seeds
Or verification only works after you already trust the operator
Blockrand uses a double-blind model:
The client commits to a seed hash (unknown to us)
The server commits to entropy (unknown to the client)
Final randomness is derived only after both are locked
Anyone can independently verify the result later
No party can bias the outcome, even if malicious.
Why we built this
We couldn’t find a randomness service that:
Works off-chain
Is cryptographically verifiable
Doesn’t require trusting the operator
Is simple enough to integrate into existing systems
So we built one.
What’s live today
Simple REST API
Deterministic, replay-verifiable results
SDKs (JS, more coming)
Open verification logic (no black boxes)
What we’d love feedback on
Threat model assumptions
Attack vectors we may have missed
Whether this solves a real problem for you
API ergonomics (especially for game devs)
This is early and opinionated.
Happy to answer questions...
https://blockrand.net
GitHub: https://github.com/blockrand-api/blockrand-js
1 comment
Show HN: Blockrand – Verifiable randomness API with a double-blind threat model | Heykuki News