Spots a price gap and trades it in 1.6 seconds.

When two pools price the same token differently, buying on the cheaper one and selling on the other makes a profit. Nervo watches for that gap and makes the trade by itself, across two blockchains, with no server or bot behind it.

Measured on test networks with test tokens. Built for the Reactive Omni Buildathon, Every Second Counts track.

One real run, step by step
Trade done
  1. Base SepoliaBlock 59,089,529
    Someone swapped on Pool B (done)
    ConstantProductPool.swap0x5be4…3a46 ↗
  2. Reactive LasnaBlock 8,402,587
    The contract decided: worth trading (done)
    CallbackRequested0x695e…3a15 ↗
  3. Base SepoliaBlock 59,089,538
    The trade was made, profit kept (done)
This run, swap to trade: 9 Base blocks later1.8 seconds
Typical time, swap to trade1.6 secondsMedian of 73 measured runs.
Slow end2.0 secondsp95: 71 of 73 runs finished within it.
Trades that never arrived0 of 73Every trade the contract sent reached Base Sepolia.
Fees saved by checking first56 %Same 3 trades made, 9 wasted ones avoided.

One loop, two ways to use it.

Cross-chain arbitrage · Base Sepolia and Reactive Lasna

Make a price gap yourself, then watch the contract answer it.

Pick one of the two pools and swap test USDC. Your swap pulls that pool's price away from the other one. The contract on Reactive Lasna sees it, checks whether the gap pays for the fees, and trades it back on Base Sepolia. The page follows each step as it lands.

  1. Mint test tokens
  2. Swap on Pool A or Pool B
  3. Watch the trade arrive
Protective orders · Unichain Sepolia

A stop-loss that watches another chain.

Your order waits on Unichain Sepolia and fills when the price on Base Sepolia reaches your trigger. No keeper or server watches the price.

1.8 smedian trigger to fill, p95 2.8 s, 40 of 40 stop-losses filled

It checks twice before any money moves.

Each check leaves a public record of what was decided and why. The names in code are those records, exactly as they appear on the chain. A cancelled trade is recorded too, so you can always see what happened.

First check · on Reactive Lasna

Is it worth trading?

If the gap is too small to cover the fees, it stops here and nothing is sent. That keeps costs down.

CallbackRequested
Yes, send the trade
MarginBelowThreshold
No, the gap is too small
ExecutorSwap
That was our own trade, ignore it
Second check · on Base Sepolia

Is it still worth it now?

Prices can move in the second or two the trade takes to arrive, so it is checked again against the latest prices before any money moves.

Executed
Traded, profit kept
SkippedMarginGone
Too late, safely cancelled
Does checking first help?

Same trades, far fewer wasted ones.

We ran the same 12 price changes through a second copy that skips the first check and sends every possible trade. Both made the same 3 trades. The copy without the check also sent 9 that were cancelled on arrival, and paid Base Sepolia fees for each.

See the full numbers
Trades sent to Base Sepolia for the same 12 price changesTradedCancelled on arrival
With the first check3 sent, 3 traded, 0 wasted
Without it12 sent, 3 traded, 9 wasted
Base Sepolia fees, with the check
0.0000049 ETH
Base Sepolia fees, without it
0.0000113 ETH
Saved by checking first
56%

Everything is public on the test networks.

Open any address on its chain's explorer and read every decision the loop made.

ContractChainAddress
ArbReactiveReactive Lasna · 53180070x0f70a79D1a0DFCB61005AF03084d3bED0e71E932 ↗
ArbExecutorBase Sepolia · 845320xb568530a915a7Ae01FBE5917E1B21d666b24DFE4 ↗
Pool ABase Sepolia · 845320x82DEA86F73edF0CC76cd2Ff4Ae8612380e63CE6F ↗
Pool BBase Sepolia · 845320xf68dE754214b013c206AFD5753b9f26428C29228 ↗
Test WETHBase Sepolia · 845320x29583A3d63FD02f1bE462d4b6A1B1BeCeFcaFB87 ↗
Test USDCBase Sepolia · 845320xb790467931Ad6694747bB90c96ac9b07c0c4DE01 ↗