Getting to Ethereum

Part of a series covering all 72 Ekubo DAO proposals. This post covers how Ekubo got onto Ethereum without moving its governance there.

Period covered: December 2024 – November 2025.

A DAO on one chain, contracts on another

Ekubo’s governance lives on Starknet. Its EVM deployment lives on Ethereum. Those are separate execution environments, and a vote on one does not move a contract on the other.

The bridge is the Starknet Owner Proxy: an account on Ethereum that only the Starknet governor can direct. Upgrade the governor to v2.3.1 in January 2025 gave the governor the ability to make L1 calls; Fund and Demonstrate StarknetOwnerProxy, later the same month, proved it worked.

The demonstration is almost comically modest: send 1 ETH to the L1 proxy, transfer 0.01 ETH of it back to refund the deployer. The point was never the 0.01 ETH. The point was a Starknet vote causing an Ethereum transaction, verifiably, with the contract verified on Etherscan before anything valuable depended on it.

Every DAO action on Ethereum since — the $1.25M of protocol-owned liquidity, the incentive campaigns, the Recovery Fund — runs through that proxy.

Two audits before a line of it was public

December 2024 — Fund additional audit, a Plainshift engagement at $125k base with $175k set aside as bounties for high-severity findings. Ekubo, Inc. escrowed the funds and returned the unused portion.

February 2025 — Fund audit for solidity codebase, a second Plainshift engagement, maximum 225k USDC against a 95k base, starting February 24.

Both are structured the same way: a modest base fee and a much larger contingent bounty pool. That is a deliberate choice about what you are buying. A flat audit fee buys attention; a bounty pool buys findings. Setting aside $175k against a $125k base says the DAO would rather overpay for a real vulnerability than underpay for a clean report.

The December proposal also flagged what was coming: “a subsequent proposal will be created regarding making the audited source code available.”

The licence

That subsequent proposal arrived in October 2025: EVM protocol license and audit competition.

It settled the open-source question for the next EVM version, and the mechanism is the interesting part. The licence text lives at an IPFS contenthash pointed to by ekubo-license-v1.eth — an ENS name owned by the Ekubo DAO. The terms of use are themselves a governed object: changing the licence means passing a proposal.

The same vote funded a $183.5k audit competition: $100k for high severity, $76.8k for medium, $3.2k for low/QA, $3.5k for an independent judge, with up to $176.8k refunded to the DAO in USDC afterwards. Roughly $240k of assets were transferred with only the necessary portion to be liquidated.

Three separate security engagements — two private audits and one public competition — before the deployment the DAO intended to put its treasury behind.

Clearing the runway

November 2025 — Halt Ethereum liquidity incentives. Wave One and the mev-resist campaigns stopped, with unspent tokens returned to the DAO’s Ethereum proxy, in preparation for the next version of the EVM DEX.

Stopping an incentive programme early is a harder vote than starting one, because the LPs earning those rewards notice immediately. Paying people to supply liquidity to a version you are about to replace is worse, and the DAO chose the visible cost over the quiet one.

The proposals

All five executed. This is the part of the DAO’s history with the least disagreement in it, which is worth contrasting against the incentive campaigns — where a third of the proposals failed.

Next: renting liquidity.