Sell What
You Don't Own.

The borrow desk and short interest tape for Robinhood Chain.

Setup Score · v0 proxy
Pool liquidity
Holders
Oracle depth
5m vs 30m TWAP
Reading Robinhood Chain…
Scroll

Squeeze lets tradersbagholdersfundsdegensanyone borrow, short, and price fear on 50,000+ Pons tokens.

Inside Squeeze

Four parts.
One market.

01Vaults
Holders deposit tokens and get paid in ETH, not more tokens. Demand to short a bag is what pays the people holding it — at 90% utilization that's a triple-digit APR on a position they weren't selling anyway.
02The Desk
Post ETH at 150% margin, borrow the token, and the protocol sells it into the token's locked Uniswap V3 pool in the same transaction. Your loss is capped at your collateral. Never unbounded.
03The Oracle
Thin pools get manipulated — so a position only liquidates when the 30-minute and the 5-minute TWAP both break. A one-block pump moves nothing. Tokens only list if moving the price 50% costs more than the collateral at risk.
04The Tape
Once borrowing exists, short interest exists — and it's public. Short interest, days to cover, utilization, borrow APR. Live, per token, free. This is the part most people will use and never place a trade.
vault — squeeze
mockup · not deployed
$CASHCAT vault · illustrative
Illustrative figures — no vault exists
312%
Supply APR · in ETH
0% utilizationkink · 80%100%
Free to withdraw390.2K CASHCAT
Interest earned+2.184 ETH
Receipt tokensCASHCAT
short — squeeze
mockup · not deployed
Open short · $CASHCAT · illustrative
Borrow124,000 CASHCAT
Sold into V3 pool2.940 ETH
Entry price · 30m TWAP0.0000237
Borrow rate312% APR
Collateral4.410 ETH
Health factor
Liquidates at 1.00
1.25
Liq. price +25%
oracle — squeeze
live chain data
Read from the pool's own observation buffer
observations stored
30m TWAP
5m TWAP
Spot
5m vs 30m divergence
loading…
Both windows come from one observe() call on the live pool. Reproduce it: node indexer/verify-oracle.mjs
tape — squeeze
live chain data
TokenSetupPool ETHHoldersObs.5m/30m
loading…
The indexer

Run it yourself.

This is the actual output of the indexer in this repo, reading Robinhood Chain. Every market is checked against the listing criteria in the spec — nothing is hand-picked.

build-tape.mjs — squeeze
real output
score.mjs — squeeze
// v0. NOT short interest — that
// needs the borrow market to exist.

setup = 35 · thinness     pool/mcap, inverted
setup + 25 · concentration top-10 share
setup + 20 · volatility   5m vs 30m
setup + 20 · turnover     24h vol/mcap

// missing input → weight dropped,
// never guessed. below 60% cover:
null
Status

What is built,
and what isn't.

Squeeze covers three markets — $PONS, $CASHCAT and $AI. Half this page renders live chain data for them. The other half is a mockup of a protocol with no deployed contracts. Every panel says which it is.

live

The Tape (v0)

Real. An indexer reads pool reserves, holder counts and the oracle buffer from Robinhood Chain, checks each token against the listing criteria, and computes a Setup Score. Runs from this repo with no API key.

live

Oracle verification

Real. verify-oracle.mjs pulls both TWAP windows out of a live pool and rejects any pool whose observation buffer is too shallow to produce a genuine average.

not built

The Desk

Mockup. No contracts are deployed or written. Short interest, days to cover, utilization and borrow rates do not exist yet — and are shown as blank rather than filled with plausible numbers.

planned

Seed capital

Earmarked, not funded. $25,000 is intended to seed the first vault and the Backstop Fund. No contract holds it, nothing has been deposited, and it will not appear in any live panel until both are true.

There is no token, no presale and no contract address. Anyone showing you one for Squeeze is not us.
Risk

What can go wrong.

Shorting thin markets is genuinely dangerous. Every parameter in this protocol exists to survive the first row of this table.

Vertical pump on a thin pool highest

A 10× in thirty minutes can outrun liquidation and leave bad debt. Mitigated by conservative borrow caps that scale with the square root of pool liquidity, 150% initial margin, and the Backstop Fund. This is the most likely way the protocol dies.

Empty vaults medium

No supply means no shorts and no product. Seeded from the treasury at launch and targeted first at tokens where holders are already desperate for yield.

Oracle attack medium

Dual-window TWAP plus a manipulation cost floor: a token only lists if moving its price 50% costs more than the total collateral at risk in that market.

Socialised loss disclosed

If the Backstop Fund is ever exhausted, the remaining bad debt falls on lenders in that specific vault — never protocol-wide. Written here rather than buried in the docs, because that is how you keep a protocol's reputation.

FAQ

Frequently asked questions

Nothing bad. You lend tokens and you get tokens back — if it goes to zero the short buys them back for dust and returns them. In token terms you are whole, and you keep the ETH interest on top. Lenders are structurally protected against the exact scenario they fear most.

Not with one transaction. Liquidation requires both the 30-minute and the 5-minute TWAP to breach the threshold, so a single-block spike does nothing except cost the attacker money in slippage. Sustaining a fake price for thirty minutes against arbitrage is a different and far more expensive problem.

Right now, none — the Desk does not exist. When it does: any token that graduated on Pons with liquidity locked in the Uniswap V3 pool it launched in, holding at least 30 ETH of liquidity and 500 holders, more than 72 hours past graduation, with an observation buffer deep enough for a genuine 30-minute TWAP, and clearing the manipulation cost floor.

The indexer already evaluates every one of those criteria against live chain state — five markets currently pass. Listings would be curated by the team in v1 and permissionless with automatic caps in v2; we would rather say that plainly than pretend v1 is already trustless.

No. It is a collateralised borrow market — you borrow the actual token and it is actually sold into the actual pool. Real supply leaves the float, which is exactly why the short interest number means something. Synthetic perps would produce a number nobody could trust.

No, and most people won't. Today the Tape publishes what can actually be measured without a borrow market: pool depth, holder counts, oracle depth and the gap between the 5-minute and 30-minute TWAP, combined into a Setup Score. It is free, public and needs no wallet.

Short interest, days to cover, utilization and borrow APR are shown blank — not because the feature is unfinished, but because those numbers are meaningless until real supply is being borrowed and sold. They arrive with the Desk, not before.

Ten percent of all borrow interest goes to the protocol. Eighty percent of that buys back and burns $SQUEEZE; twenty percent funds the Backstop. Staking cuts your borrow rate in tiers and gets you early access to new listings. No governance theatre — two benefits you can price in ETH.

Check the work

Five markets on Robinhood Chain currently clear every listing criterion. The script that decides that is in the repo — run it yourself.

Squeeze is an independent protocol on Robinhood Chain. Not affiliated with, endorsed by, or connected to Robinhood Markets, Inc. or Pons. Shorting leveraged positions in illiquid markets can result in total loss of collateral. Figures shown on this page are illustrative.