How did you integrate our partners, if any?

What are the key links to share? (Ex. demo video, GitHub, deck)

Github: https://github.com/mishoko/uniswap-LP-queue Slides: https://github.com/mishoko/uniswap-LP-queue/blob/main/slides.html Project Link: https://github.com/mishoko/uniswap-LP-queue/blob/main/frontend/index.html Demo Video: https://youtu.be/oUqQS6FqvuM

Problem / Background: What inspired the idea? What problems are you solving?

Every protocol that needs deep liquidity buys it the same way: print your own token and hand it out as rewards. It works, and it is expensive in a way that does not show up as an expense. You dilute every existing holder, you manufacture sellers who never wanted to be sellers, and the day you stop paying the liquidity leaves. You are renting, and you are paying the rent in equity.

The idea came from looking at what an AMM actually does to the people funding it. Uniswap fills every LP in a position pro-rata - every trade takes a little from everybody, in proportion. That is not a setting somebody chose; it is the only thing an AMM does. You cannot opt out, go first, or go last.

Every real electronic market on earth prices queue position. Most of them do it implicitly, in latency spend that is burnt on infrastructure rather than paid to anyone in the market. Uniswap has never had a queue, so it has never had a price for one - hich does not make the ordering worthless. It makes it unpriceable inside the pool, and therefore captured outside it.

So: give each funder a rank and fill them in order. The surprising part, and the whole product, is that being first is bad. The front absorbs the adverse selection first, at the earliest and worst prices of every move. It is a first-loss position. and everyone behind it does measurably better because it is there.

That turns an unpriced externality into a thing a treasury can buy. It funds rank 1, takes the worse fills on purpose, and outside LPs get paid to stand behind it. Nothing is printed. Nobody is diluted. The capital is still on the treasury's balance sheet.

Impact: What makes this project unique? What impact will this make?

QUEUE is the first mechanism that orders LPs inside one position, at one distance. That ordering is what makes contract-enforced subordination possible: one LP takes structurally worse fills so another gets structurally better ones — and it is not a payment step somebody can withhold or a programme a vote can reverse. It is the order the contract fills people in, executed inside swap.

Each seat is an ERC-6909 token under a Harberger lease: you post your own price, you pay 10%/yr rent on that number, and anyone may buy the seat off you at exactly that number, at any moment. Post high and you are safe but you pay for it; post low and you will be taken. Rent flows forward - the back pays the front, because the back consumes the protection the front supplies.

Challenges: What was challenging about building this project?

  1. The hardest bug was in our own reasoning, not our own code. Six phases in, the mechanism was correct and the business case did not exist: all the seats share one Uniswap position, so they can at best tie with not using the hook. Fixing the distribution (29 of 32 seats losing → 0 of 32) created no reason for anyone to participate. That took a full phase to admit and answer in closed form. We had been optimising the wrong variable.

  2. Subordination is only real if you cannot walk away from it. A first-loss position whose holder can step aside the moment trouble arrives is worth nothing. We found and closed four separate evacuation doors

  3. Our own instruments were wrong as often as the code.

  4. v4 has sharp edges that a passing test suite will not show you.

What made all of this findable was mutation testing and an invariant campaign. The campaign found three real bugs in code that had already passed 135 tests and 61 mutations, all reachable through the ordinary public API and none visible to any correctness test. Mutation testing has found a defect every single time it has been run on this project.