The distribute() function in CuratorRewardsDistributor distributes curator rewards accumulated through different artId mints under a credId. The balanceOf[credId] is distributed to users who hold non-zero amount of shares of that credId.
A malicious user could frontrun a distribute(credId) function call to buy a huge amount of shares of that credId. This would put the user into the curator addresses list that is eligible to receive the rewards.
If we check lines 114-115 in the snippet below, we can see that it uses the shares of a curator and divides it by the total number of shares to determine the % of distributeAmount to give to the curator. Since the malicious user frontran the distribute() call by buying a large amount of shares, the rewards would be diluted and distributed when the distribute() function call goes through. 
Through this, we can see that although the malicious user just bought the shares, the user is immediately eligible for the curator rewards. Many malicious users could exploit this opportunity to earn a good chunk of distributeAmount.
To note, there is obviously the 10 minute share lock period that the malicious user will have to adhere to, during which some curators can sell to profit from the malicious users buy that increased the price. Since there is a tradeoff introduced with the 10 minute delay, the net profit an attacker could make through this wont be much but the possibility still exists since previous curators may not sell.
Introduce a minimum time period, e.g., 10 minutes required before being eligible for curator reward distribution. 
