← All insights
RESEARCHZcash Forum · strahncryptography

Trustless shielded ZEC-EVM swap demonstrated

Using Zcash's property that transaction IDs can be predicted and constrained in advance, a team said it demonstrated an atomic swap between shielded ZEC and an EVM chain without any new signature scheme.

A team announced on the forum that it demonstrated a fully trustless atomic swap between shielded ZEC and an EVM chain without a new signature scheme. The earlier difficulties were that Zcash has no scripts, adaptor signatures are vulnerable to front-running, and there is no notion of finality. The team deployed a Skyhook bridge contract on the Ethereum Sepolia testnet.

The key idea is that because Zcash uses a segregated-witness-style approach (ZIP-244 in v5, ZIP 229 in Ironwood), a transaction ID can be predicted and constrained in advance, committing to the intent of a trade beforehand. Whether the trade then happened is verified with SPV. The team confirmed that the txid the contract expected was actually included in a Zcash testnet block, and explained that in an ordinary light wallet transaction they simply fixed psi, rho and rcm in advance.

The demo went step by step. A published pk_d and g_d and placed an order of 10,000 zatoshi for 5,000 szabo; B made an offer and A accepted, locking 5,000 szabo in the contract. Ethereum then verified the existence of the block and the Equihash solution, and B submitted a Merkle proof and received the ETH. However, because it is not wrapped in a zero-knowledge proof, the cmx blinding factor is revealed, and the team said it is working on replacing this with a BLS12-381 Groth16-based NIZK.

The author claimed the approach goes beyond simple swaps. Even if the orderer reveals only an address without a signature, a deposit to that address can be proven without exposing identity, which other contracts could use to verify shielded deposits. A trustless wrapper for multisig (such as FROST) guardian structures could be built, and institutional users could use ordinary transfer signatures as they are. A programmability feature that uses the first 52 bytes of encCiphertext for cross-chain messages is also being considered, but it is not implemented because its security proof is pending.

The team asked the community whether it wants this introduced to Zcash and whether it should apply for a community grant to commercialize it. A follow-up post explaining the programmability design and a security paper are due soon. The team said it deployed on Sepolia rather than the original target, Base, because it has almost no funds, and also asked for sponsorship.

This summary is based on the original; check the original for exact details.