A Polkadot forum proposal swaps re-execution for proof checking
Serguey Asael Shinder1 min readA post on the Polkadot forum dated 23 September 2026 sets out WonderZK, a proposed protocol architecture in which validators verify zero-knowledge proofs instead of re-executing computation, with heavy execution and proving moved off the validator set: https://forum.polkadot.network/t/polkadot-wonderzk/18742.
This is a proposal by an individual, not a project announcement. The author, posting as ultracoconut, describes designing the architecture and is asking for collaborators with backgrounds in zero-knowledge systems, cryptography, proving systems and Polkadot's architecture.
What it means
The idea is the oldest trade in the scaling literature, applied to a specific validator set. Re-execution gives every validator first-hand knowledge at the cost of doing all the work; proof verification gives second-hand knowledge cheaply. Polkadot's design already leans on validators checking parachain work, so moving that check from execution to verification changes what a validator needs to be - and therefore what it costs to be one.
What a forum post cannot tell you is whether the economics survive. Proving is not free; it moves cost from many validators to fewer provers, and who pays them, and how their availability is guaranteed, is the part that turns an architecture into a protocol.
⚠️ Filed here because the architecture is interesting and the venue is the project's own forum, not because anything has been adopted. No implementation, benchmark or timeline is reported, and a request for collaborators is the opposite of a roadmap.