Blind Slippage Pools

Darklake’s Blind Slippage Pools use a two-step execution model to prevent frontrunning without compromising liquidity efficiency.

At the heart of most MEV exploits is intent leakage. Traditional AMMs are fully transparent, exposing trade data to the public mempool before execution. This creates opportunities for frontrunners to profit at the expense of honest market participants. Blind Slippage Pools break that vulnerability by keeping the minimum output of each swap confidential until after execution.

How It Works

The trade process is split into two phases:

1. Commitment Phase

After a user expresses intent to trade asset Y for asset X with a specified minimum output (min_out) of Y, a hash of the min_out is generated together with a unique cryptographic salt, and both are submitted with the trade. Input tokens are transferred from the user to the pool and the price action is calculated.

2. Proof phase

A Groth16 proof is created by Darklake's Proof Generator to show that the calculated return either matches or exceeds the hidden min_out. If valid, the trade settles. If not, the transaction is reverted and funds returned.

The diagram below illustrates the execution sequence in more detail:

Key advantages

Blind Slippage Pools add a cryptographic commitment layer to the AMM that makes slippage data invisible to searchers but verifiable post-trade. This mechanism neutralizes frontrunning at the source by removing the signal MEV bots rely on to manipulate transaction ordering. At the same time, the underlying AMM model is unchanged, meaning liquidity remains fully usable, with no idle capital, batching delays, or sequencing bottlenecks.