Submitted by falconhoof, also found by josephdara, inzinko (1, 2), zhaojie, forgebyola, Rhaydden, J4X, BiasedMerc, jesjupyter, and cats
https://github.com/code-423n4/2024-01-salty/blob/53516c2cdfdfacb662cdea6417c52f23c94d5b5b/src/dao/Proposals.sol#L162-L177 
https://github.com/code-423n4/2024-01-salty/blob/53516c2cdfdfacb662cdea6417c52f23c94d5b5b/src/dao/Proposals.sol#L81-L118
The creation of token whitelisting proposals isn limited to 5 proposals; after which it is DOSd until proposals are voted on and finalized.
In proposeTokenWhitelisting() when a proposal is created to whitelist a new token; the ballotId is added to _openBallotsForTokenWhitelisting.
There is then a check to ensure that the length of _openBallotsForTokenWhitelisting does not exceed daoConfig.maxPendingTokensForWhitelisting().
Where maxPendingTokensForWhitelisting has been reached, new proposals will not be created for whitelisting tokens. This could happen by accident or by malicious actions on the part of users manipulating the system.
By default maxPendingTokensForWhitelisting is set to 5 but it could be decreased via vote to 3 which would make the issue even more common place.
Add the test function below to DAO.t.sol and run:
Malicious users can clog the whitelisting queue with fake token proposals, blocking the addition of genuine tokens and DOSing core functionality of the protocol.
This can restrict the ability of the protocol to operate and its popularity with users.
Although a user can only create one proposal per address at a time; a coordinated group of just five could block the functionality indefinitely.
Foundry
Allow a trusted authority to remove proposals which they deem to be malicious such as proposals for fake tokens.
othernet-global (Salty.IO) confirmed, but disagreed with severity and commented:
Picodes (Judge) decreased severity to Medium and commented:
othernet-global (Salty.IO) commented:
Status: Mitigation confirmed. Full details in reports from zzebra83, 0xpiken, and [t0x1c].
