01
Native liquidity
The fungible leg is ordinary ERC-20, so OBOL trades on a standard v4 pool — no wrapper, no bespoke router.
ERC-402: An experimental token standard, struck at the price you paid.
Every hybrid before this one mints coins blind to price — a transfer never sees the market. ERC-402 moves the mint into a Uniswap v4 hook, where it can.
1,111 obols / $OBOL / ERC-402 / Robinhood Chain
Contract
Not yet deployed — chain 4663
Protocol overview
Struck by the market, not by a team.
ERC-404 proved the idea worked but left every coin interchangeable. ERC-402 moves the machinery inside a Uniswap v4 hook, where the price is finally visible.
02
Fractionalisation
Hold 0.4 of an obol and you hold exposure to the collection. Cross 1.0 and a real coin is yours; fall below and it melts back.
03
Semi-fungible
One contract carries the balance and the ids. We report IERC721 honestly, so
marketplaces can actually find the collection.
04
The strike ●
The part no ERC-20 can do. The hook reads the execution price and stamps it into the coin forever — and that number sets the grade.
05
Gamble
Stake obols in provably-fair 1v1 duels. One flip settles it, the winner claims both coins from escrow. Zero house edge, no grade advantage.
The strike
A plain transfer has no idea what the price is.
A transfer knows balances and nothing else. Our hook runs
inside the swap, where the price lives. Follow one buy through it.
A buy arrives
ETH for OBOL, through any router. Nothing special is asked of the trader.
The fee bends to your grade
beforeSwap reads the best coin the trader already holds and overrides the fee. Better coinage, cheaper trade.
The price is published
afterSwap reads sqrtPriceX96 as executed and writes it to transient storage — alive for this transaction only.
The coin is struck
The balance crosses a whole unit. The token stamps that price into the new obol — permanently, as part of the token itself.
The hoard takes its cut
A slice of the fee lands in the treasury standing behind every circulating obol.
The hook publishes only a price; the token decides whether a coin is owed. Neither side has to trust caller-supplied data.
Standard comparison
ERC-402 against the lineage.
ERC-404 drew this matrix first. The bottom rows require a hook, which is why only the last column can tick them.
| Capability | ERC-20 | ERC-721 | ERC-404 | ERC-402 |
|---|---|---|---|---|
| Fungible transfers | ✓ | – | ✓ | ✓ |
| Non-fungible tokens | – | ✓ | ✓ | ✓ |
| Native DEX liquidity | ✓ | – | ✓ | ✓ |
| Fractional ownership | – | – | ✓ | ✓ |
| Automatic mint & burn on trade | – | – | ✓ | ✓ |
| Single contract | – | – | ✓ | ✓ |
| Reentrancy protected | – | – | – | ✓ |
| Launch trading gate | – | – | – | ✓ |
| Marketplace discoverable | – | ✓ | – | ✓ |
| No owner-managed exemption list | ✓ | ✓ | – | ✓ |
| Price-stamped provenance | – | – | – | ✓ |
| Holder-tiered swap fees | – | – | – | ✓ |
| Protocol-owned backing | – | – | – | ✓ |
Questions & answers.
An experimental, unofficial standard from Pandora Labs that merged ERC-20 and ERC-721 into a single contract — native liquidity, fractionalisation, and semi-fungible behaviour in one deployment. It was never accepted as an EIP, and the original implementation had well-documented gas and correctness problems. It is the ancestor of everything on this page.
Our designation for a semi-fungible standard that keeps the ERC-404 idea but relocates the mint-and-burn machinery to a Uniswap v4 hook. The number is a joke with a point: HTTP 402 is the status code reserved for payments and never used, and it sits below 404 in the lineage that produced it.
Like ERC-404 before it, ERC-402 is not an accepted EIP. It is our own name for our own line of work, published openly and offered on its merits.
Balance and coinage are two views of one number. When your balance crosses a whole unit upward, the contract strikes an obol to your address at the pool price in that moment — recorded on-chain, never mutable. When it falls back below, that obol melts into a recycle bank and is re-issued to the next buyer who crosses the same threshold. Circulating supply is therefore always struck − banked, and it goes down as well as up.
Grade is a pure function of the strike price, assigned once and frozen. Proof is the opening band, then Mint State, Extremely Fine, Very Fine, Fine, and Good as price climbs. The names are the real coin-grading vocabulary, not invented rarity tiers.
Grade also sets your swap fee: Proof trades at 0.25% where Good pays the 1.00% base. Nobody is allocated a grade — supply in each band is whatever the market actually produced, and once price leaves a band that band can never grow again.
Gamble is a provably-fair 1v1 dueling contract. Two holders stake one obol each — one fair flip settles it, and the winner claims both coins from escrow, strikes and grades included. There is no house edge and no grade advantage: a Proof and a Good meet as equals.
Gamble is not live. No dueling contract exists yet and nothing about it is deployed. It ships after launch as a separate contract, and it will be announced here before anywhere else.
A treasury funded by a slice of every swap fee, reported as a per-coin backing figure. It accrues from trading rather than from a team allocation or a tax on holders. It is deliberately a floor-support mechanism and not a redemption promise — see the risk note in the footer.
Two reasons, one of them unglamorous. The glamorous one is the strike price, which requires code running inside the swap.
The unglamorous one matters more in practice. On V3 a hybrid token has to maintain an owner-managed allowlist so the pool contract doesn't mint NFTs into itself, and the pool must be exempted before it ever receives tokens or the launch bricks. In v4 all pools share one singleton PoolManager, so it is exempt in code permanently. There is no list to maintain, no ordering to get wrong, and no pool address that can be registered incorrectly at deploy time and quietly break the launch.
Not yet. No third-party audit has been performed and nothing is deployed to mainnet. All contracts will be open source before launch.
Worth stating plainly: the two largest Uniswap v4 losses to date — Cork Protocol at roughly $12M and Bunni at $8.4M — were both application-level bugs in audited hook code, not flaws in v4 itself. An audit is necessary and is not sufficient. Participate only with risk you are willing to lose.