in the matter of the token $PAWL, the vault appearing as sole claimant
| filed2026 | drive fee2.2% of volume | reserveEthereum (Wormhole) |
| claimantthe vault, one address | fee destinationfixed at creation, no reassignment | withdrawal instructionnone exists |
A prior apparatus of this class is known. It paired a coin's liquidity with Ethereum, fed the pool from a protocol fee, and published proofs that the reserve could only grow. Its own specification recorded the defect in plain sight: "the key that deposits could also withdraw." The prior art, in the language of this office, disclosed a ratchet and omitted the pawl. The operator's key reversed the mechanism, and the reserve was taken.
Finding. A ratchet without a pawl is a wheel. The present apparatus is named for the part the prior art left out.
$PAWL trades on a Meteora dynamic bonding curve whose fee is written into the market's own configuration: 2.2% of every trade, collected entirely in SOL, payable to a fee claimer fixed in the transaction that created the market. The program contains no instruction to reassign that destination. Where the curve's creator is transferable by design, the claimant leg is not, and it is the claimant leg this apparatus is bolted to. If V is cumulative quote volume, revenue is
dR = 0.0220 · dV (1)
and that is the entire monetary base. No emission, no treasury, no second round. Equation (1) is indifferent to price, to holders, and to the operator. The only input is that people trade.
A block does not close on a clock; it closes when the vault holds enough for settlement to be worth its own cost. Below roughly 0.020 SOL the two swaps lose more to fees and slippage than they deposit. Creating the reserve pool the first time pays rent on six accounts, about 0.250 SOL, withheld from the first block's swap budget.
| Block | Target | Volume implied | Remarks |
|---|---|---|---|
| 0 | 0.285 SOL | ≈ 13 SOL | Opens the market: deployment floor, gas reserve, and rent for the pool's six accounts. |
| 1…n | 0.035 SOL | ≈ 1.6 SOL | Deposits into a pool that already exists. Rent drops out forever. |
If volume never comes, no block closes, and Claim 3 records what that costs the holder: nothing.
claim_partner_trading_fee; proceeds land as native SOL. Claiming and deploying are separate decisions, so nothing strands.The reserve is one side of a Raydium constant-product position, xy = k: x the $PAWL reserve, y the Ethereum reserve. Constant product takes a deposit at any price, never goes one-sided, and needs no management, which suits a position intended to outlive everyone's attention. The pool address is derived from the program and the two mints every block and asked of the chain, never read from our records; a stale ledger cannot cause a second pool. Trading fees paid to the pool accrue inside it, to the reserves themselves. The wheel advances when a block settles and when anyone trades against it.
What is claimed is:
Monotone depth. kn+1 ≥ kn for every block n, strictly whenever a block settles or a trade occurs.
A swap with fee φ moves the invariant to a value not below k, equal only at φ = 0. A deposit scales both reserves by 1+ε, ε>0, giving (1+ε)²k > k. A withdrawal would scale them down, and the LP required to perform one has been burned. A quantity acted on only by non-decreasing operations is non-decreasing. ∎
y = √(kp), x = √(k/p) (2)
The floor has no return stroke. Fix any price p. The Ethereum held whenever $PAWL trades at p is non-decreasing in time, and strictly increases with every settled block.
By (2) the Ethereum reserve at price p is √(kp), strictly increasing in k. By Claim 1, k is non-decreasing. Composition of an increasing function with a non-decreasing one is non-decreasing. ∎
Idleness is the worst case. If volume stops, the state does not move. No block settles, no reserve is spent, nothing is lost but time.
By (1) revenue is proportional to volume, so zero volume accrues nothing and the vault never reaches target. A settlement below target is not attempted, and no other instruction spends the pool. The failure mode is a pause, not a reversal. ∎
Not claimed: a redemption value (the Ethereum is in a trading pool, not an escrow), a floor price (a constant-product pool has none), or that a holder is made whole. The claims are narrower and survive arithmetic: the depth behind $PAWL at any given price ratchets, and the pawl is not ours to lift.
Meteora's program fixes the fee destination, an aggregator executes the swaps, Raydium's pool holds the reserve. A keeper runs the cycle each minute and holds the key to the launch wallet. That key cannot mint: the token authority is set immutable in the config. It cannot redirect the fee: the program lets a pool's creator be transferred, which is exactly why the creator's share here is zero and the whole 2.2% is payable to the claimant leg, written once, changeable by no instruction. It cannot freeze or tax. And it cannot withdraw the position, because the LP that withdrawing requires is burned the block it is minted. On the curve the same holds at graduation: migrated liquidity is 100% permanently locked, fee-bearing, irretrievable. One address to watch. Every lamport in leaves by one of two roads, both ending inside the wheel.
The reserve asset is Wormhole-bridged Ethereum. The ETH behind it sits in Wormhole's token bridge on mainnet, and the Solana mint authority belongs to the bridge program. Because this apparatus never withdraws from its pool, bridge failure is a permanent exposure no one here can exit. It is the largest single risk on this sheet and it is printed in the specification, not a footnote. Verify the reserve mint against the bridge's own documents before trusting anything above. This is an experiment in mechanism, not a security, and not a promise of return.