Market cap $2.18 T −2.74 % over 24 hours24h volume $75.31 BBTC dominance 59.1 %ETH 10.1 %Fear & Greed 12 Extreme fearCoins 21,839

Uniswap’s UNI burn has never run on X Layer: a bridge mismatch strands about $300,000 of fees until a vote closing 20 November

On X Layer, the Uniswap UNI burn has never fired: a bridge rejects the token and about $300,000 of fees wait. A Snapshot vote on the fix ends 20 November.

9 min read

Abstract flame beside a bridge, a snowflake and a ballot box
Table of contents
  1. What broke on X Layer: a burn routed through the wrong bridge
  2. Why the fix comes with a Dutch auction
  3. The Snapshot vote closes on 20 November, and the code is not final
  4. Burns elsewhere continue, and UNI has not reacted
  5. What to watch after the vote

Uniswap’s UNI burn has never run on X Layer. Since governance switched protocol fees on there on 8 March 2026, the fees have piled up in a contract and none has been released, because the burn mechanism sends UNI through a bridge that refuses the token. Uniswap Labs disclosed the fault in a governance forum post on 6 November and is asking UNI holders to approve a repair. The Snapshot vote on it closes on Friday 20 November at 20:34 UTC.

The amount involved is small for a protocol of this size: about $300,000, according to a pull request in Uniswap’s protocol-fees repository. That figure is an engineer’s estimate written into the code review, and no official balance has been published.

The scope matters, because several headlines this week speak of Uniswap freezing fees or halting its burns. Nothing has been halted. The burn on Ethereum keeps running. One chain is affected, and there the burn did not stop: it never started.

What broke on X Layer: a burn routed through the wrong bridge

Uniswap’s fee system works the same way on every chain. A share of blockchain, dove gli utenti scambiano token direttamente dal proprio wallet senza affidare la custodia dei fondi a una società.">swap fees goes to a contract called the TokenJar. Nobody can take the fees out for free: a trader, usually an automated one known as a searcher, has to hand over a fixed amount of UNI, and that UNI is destroyed. On Ethereum the tokens go straight to a dead address. On other chains the UNI has to travel back to Ethereum first, and a contract called the releaser handles that trip.

Proposal 95, titled “Protocol Fee Expansion: Vote 2”, extended this system to X Layer, the Ethereum layer 2 network built by the exchange OKX. It passed with 77,831,696 UNI in favour and 11 against, on a quorum of 40 million, and was executed on 8 March 2026. Its text says UNI burned on layer 2 chains “is bridged back to mainnet and sent to 0xdead”.

That is where the design failed. X Layer has two bridges with different jobs, Uniswap Labs explains in its post. Governance messages from Ethereum arrive over the OP Stack bridge. Tokens move over Agglayer. The releaser deployed under Proposal 95 copied the pattern used on other OP Stack chains and tried to send UNI over the OP Stack bridge, which on X Layer is set to “reject ERC-20 tokens, including UNI”.

With no working route to Ethereum, the releaser could not complete a single release. Fees kept arriving in the TokenJar and stayed there. The pull request puts it bluntly: the X Layer releaser “has never fired”. Labs says the problem was identified in late September, about six months after the fees were switched on, and gives no exact day. If you are new to the plumbing, our guide to how cross-chain bridges and messaging work covers the difference between moving a message and moving a token.

Abstract broken chain link between a server rack and stacks of coins
On X Layer, the releaser sent UNI to a bridge that rejects tokens, so the fees stayed in the TokenJar.

The money is not lost and nobody has stolen it. It sits in a contract that only pays out against a UNI burn, and the burn path is closed until governance replaces the releaser.

Why the fix comes with a Dutch auction

Swapping the releaser for one that speaks to Agglayer would be enough to reopen the path. It would also create a prize. The standard price for emptying a TokenJar is 2,000 UNI, a threshold that assumes the jar is emptied often and never holds much more than that is worth. On X Layer, eight months of fees are waiting.

At the UNI price of $5.30 recorded at 05:13 UTC on 17 November, 2,000 UNI cost about $10,600. Whoever got the first transaction in after the repair would pay that and collect roughly $300,000. In the words of the pull request, “the first searcher gets all of it for 2,000 UNI”. Holders would see 2,000 UNI burned for fees that could have paid for far more. Traders call this MEV, the profit bots earn by getting a transaction in first.

So the proposal adds a second contract, the DecayingThresholdSetter. It starts the price of the jar high and lowers it step by step until someone finds it worth paying: a Dutch auction, paid in UNI that gets burned. Once the price is back to 2,000 UNI, the setter hands control of the threshold back to governance.

The pull request, opened on 2 November and marked ready for review on 14 November, sets the numbers for X Layer, and a review comment of 13 November confirms a deployment with a two-day half-life. The two primary documents do not agree on how long the descent lasts.

ParameterValueSource
Fees waiting on X LayerAbout $300,000 (estimate)GitHub PR 167
Standard release threshold2,000 UNIUniswap Labs forum post
Starting threshold80,000 UNIGitHub PR 167
Half-life of the threshold2 daysGitHub PR 167
Expected clearing priceAbout 41,000 UNI, after about 2 daysGitHub PR 167
Time to reach the 2,000 UNI floorAbout 10.8 daysGitHub PR 167
Time for the decay to completeAbout six days after the proposal executesUniswap Labs forum post

Six days or 10.8: the forum post and the code review give different schedules, and neither has been corrected to match the other. The pull request’s own figure is consistent with its parameters: 80,000 is 40 times 2,000, which takes a little more than five halvings of two days each, so between ten and eleven days. The forum post does not state which parameters its six days rest on.

If the auction clears near the 41,000 UNI the engineers expect, about $217,000 of UNI at the same 05:13 UTC price would be burned in one transaction. The engineers give 41,000 as an expectation only: the real figure depends on what the jar holds on the day and on how hard searchers compete.

The Snapshot vote closes on 20 November, and the code is not final

Uniswap Labs announced on 16 November that the Snapshot vote was live. The Snapshot page shows it opened on 15 November at 20:34 UTC and closes on 20 November at the same hour. At about 05:45 UTC on 17 November it was still active and had barely started: 31 votes carrying about 11,400 UNI, all in favour, against a quorum of 10 million UNI. Snapshot is an offchain signal. If it passes, a binding onchain proposal follows under the usual governance calendar, and no date has been set for it.

The onchain step is light. Governance on Ethereum sends one message over the OP Stack bridge, with a gas limit of 200,000, telling the TokenJar on X Layer to use the new releaser. The message bridge works; only tokens are refused. No hard fork and no emergency power is needed. Readers who want the background on who decides such things can read our guide to governance tokens and fee switches.

Abstract hourglass and calendar beside a gavel and a line stepping downward
The Snapshot vote closes on 20 November at 20:34 UTC; the release threshold would then step down from 80,000 UNI to 2,000 UNI.

Several pieces are still unfinished. The audit reports are to be added before the onchain proposal, according to the forum post, and none is linked yet. The pull request lists a fork test of the full X Layer proposal as still to do, and on 16 November its base branch was switched from the main branch to the one holding the new releaser. It also notes that no UNI exists on X Layer today, which is awkward for an auction paid in UNI: tokens have to be bridged in before anyone can bid, and the documents do not say who will do it.

TokenPost reported in the early hours of 17 November (UTC) that the wider proposal payload “remains marked Draft” and that no deployment had been confirmed. CryptoSlate, in its own report the same day, wrote that the official deployment registry still listed the original X Layer releaser.

The replacement contract has already changed name once. Labs first called it AggLayerReleaser, then said on 13 November it would become XLayerReleaser, because some of its details make it unsuitable for other Agglayer chains.

The vote is also a bundle. The same proposal upgrades the VolatilePair hook on Uniswap v4 on Ethereum, limiting its temporary fee to the block after a large price move. Both changes are put to holders as a single proposal.

Burns elsewhere continue, and UNI has not reacted

The rest of the burn is untouched. Data from Strata Terminal cited by CryptoSlate show $7.8 million of UNI burned in the 30 days to 15 November, and 14.1 million UNI destroyed by recurring burns since they began. Against that, the $300,000 waiting on X Layer is under 4% of one month of burns. TokenPost makes the same point: the figure is “not a confirmed current TokenJar balance”, and burns across Uniswap have not stopped.

New sources of fees are still being added. In October, Uniswap v4 began charging fees on swaps routed to outside venues, with the proceeds going to the same burn. The full picture of the protocol, its fee switch and its risks is on our Uniswap project page.

UNI has not moved on the story. It traded at $5.30 at 05:13 UTC on 17 November, up 2.51% over 24 hours and down 23.50% over seven days. The weekly fall came in a market-wide slide in which bitcoin lost 12.68% over the same seven days, and it began before this story circulated.

What the episode does show is a gap in how the fee expansion was tested. A contract approved by nearly 78 million UNI went live on a chain where it could not work, and the fault went unnoticed for about six months. Asked on the forum on 13 November what would catch the next bridge mismatch, Porter Geer, whose forum profile reads “Governance at Uniswap Labs”, said new bridges now go through code and documentation review, manual use and interface mocking before a proposal, and that a multichain testing framework, still in early development, is expected in the first quarter.

What to watch after the vote

The first date is 20 November at 20:34 UTC, when the Snapshot vote closes; it needs 10 million UNI to count. After that come the audit reports, the onchain proposal and its execution, which starts the auction clock. The final code should say whether that clock runs six days or nearly eleven. How much the X Layer jar really holds will be known only when the release transaction goes through.

Sources

  1. Uniswap Labs, [Temp Check] Upgrade VolatilePair hook and update X Layer Releaser, 6 November 2026 gov.uniswap.org
  2. GitHub, Uniswap/protocol-fees pull request 167: DecayingThresholdSetter github.com
  3. Uniswap governance, Proposal 95: Protocol Fee Expansion, Vote 2 vote.uniswapfoundation.org
  4. TokenPost, 17 November 2026 tokenpost.com
  5. CryptoSlate, 17 November 2026 cryptoslate.com
  6. Snapshot, Uniswap governance: Upgrade VolatilePair hook and update X Layer Releaser (vote 15 to 20 November 2026) snapshot.org

Leave a Reply

Your email address will not be published. Required fields are marked *