#nil $NIL @Nillion
A trading instruction can stay encrypted and still execute badly.
That distinction made me look more closely at Nillion’s Covenants.
With Dusk, Nillion’s first Covenants release on Ethereum mainnet, an instruction can remain sealed until its release condition is met.
The mechanism matters: the author selects a committee of m nodes and a threshold k. The payload is split using Shamir secret sharing, with each share encrypted for its recipient. Any k shares can reconstruct it; fewer than k cannot.
For example, a 3 of 5 committee can tolerate two unavailable nodes. But if one operator controls three shares, it can reconstruct early. I would check operator independence before treating node count as decentralization.
When nodes publish enough shares, anyone can reconstruct the payload. An onchain commitment checks that the revealed instruction matches the original. A settlement hook can then pass it to an application.
Here is the part I find most important: revealing an instruction does not guarantee a successful trade.
The receiving contract still needs authorization, replay protection and execution limits such as minimum output. Gas funding matters too: nodes can delay publishing shares if the base fee exceeds the author’s reimbursement ceiling.
NIL connects to this infrastructure through node staking and protocol fees. What I want to watch next is repeat application usage, committee diversity and successful settlements.
For me, the strongest Covenants thesis is protecting an instruction before disclosure while making its execution verifiable afterward. Dusk’s adoption will depend on developers getting both sides right.
A trading instruction can stay encrypted and still execute badly.
That distinction made me look more closely at Nillion’s Covenants.
With Dusk, Nillion’s first Covenants release on Ethereum mainnet, an instruction can remain sealed until its release condition is met.
The mechanism matters: the author selects a committee of m nodes and a threshold k. The payload is split using Shamir secret sharing, with each share encrypted for its recipient. Any k shares can reconstruct it; fewer than k cannot.
For example, a 3 of 5 committee can tolerate two unavailable nodes. But if one operator controls three shares, it can reconstruct early. I would check operator independence before treating node count as decentralization.
When nodes publish enough shares, anyone can reconstruct the payload. An onchain commitment checks that the revealed instruction matches the original. A settlement hook can then pass it to an application.
Here is the part I find most important: revealing an instruction does not guarantee a successful trade.
The receiving contract still needs authorization, replay protection and execution limits such as minimum output. Gas funding matters too: nodes can delay publishing shares if the base fee exceeds the author’s reimbursement ceiling.
NIL connects to this infrastructure through node staking and protocol fees. What I want to watch next is repeat application usage, committee diversity and successful settlements.
For me, the strongest Covenants thesis is protecting an instruction before disclosure while making its execution verifiable afterward. Dusk’s adoption will depend on developers getting both sides right.