EVERYONE HAS TO SELL. THE QUESTION IS HOW.
You hold a position large enough that your own sell order moves the price. The whole memecoin tool stack was built for the way in — new pair detection, sniping, copy trade. On the way out you get one market-sell button and a thin pool.
Exit Desk reads the pool onchain and cuts your sell into slices sized to real depth. Assets never leave your wallet. You sign every slice.
Read the pool first.
Do the arithmetic second.
Onchain pool state reader
Two-sided depth, pool fees and available routes across DEXes, read at this moment. Without this layer there is nothing to slice against. Every number on the simulation table starts from real pool state, not a displayed price.
Slicer bound to your slippage ceiling
You enter the size you need out and the maximum slippage you accept per slice. The system derives slice count, each slice size and the spacing between them from measured depth.
Simulation table, next to the single-order case
Estimated slippage per slice, the total, and the expected average price — placed directly beside what happens if you sell in one market order. The default sell button does not show you that number before you sign. This table does.
You sign each slice from your own wallet
The system builds the transaction, pushes it to your wallet, waits for your signature, waits for confirmation, and only then builds the next slice. No keys held, no deposits taken, no internal matching. Every quote is re-simulated right before it is pushed; drift past the limit rebuilds instead of letting you sign a stale transaction. Each transaction carries an onchain slippage guard.
Stop when the pool runs dry, and record all of it
Depth below your floor stops the schedule. You continue, stretch the spacing, shrink the slice, or cancel the remainder. At the end the execution log lists every slice: size, price achieved, realised slippage against the estimate, fee paid. Exportable for your books.
Cut it before you sign it
| avg price | {{ singleAvg }} |
| total slippage | {{ singleSlip }} |
| avg price | {{ slicedAvg }} |
| total slippage | {{ slicedSlip }} |
| # | Slice size | Est. slip | Exec price | Proceeds | Wait |
|---|---|---|---|---|---|
| {{ s.n }} | {{ s.size }} | {{ s.slip }} | {{ s.exec }} | {{ s.proceeds }} | {{ s.wait }} |
Everyone leaves through one door. In order.
Slicing is not free. Many small, evenly spaced, predictable slices widen your sandwich surface and invite bots to trail the schedule — in some configurations the total cost lands above a single order.
So the desk scores MEV surface for the configuration you chose and warns before it runs, randomises slice size and spacing, and can route through a private path instead of the public mempool. The log then records real cost on both scenarios, so you conclude from your own data instead of taking our word.
Estimate on the left. Reality on the right.
| Slice | Size | Price achieved | Est. slip | Real slip | Route | Status |
|---|---|---|---|---|---|---|
| {{ r.n }} | {{ r.size }} | {{ r.price }} | {{ r.est }} | {{ r.real }} | {{ r.route }} | {{ r.status }} |