Aptos Pizza Drop

Description The function get_random_slice uses timestamp::now_microseconds() to generate randomness for pizza slice distribution:

move
let time = timestamp::now_microseconds();
let random_val = time % 401;
let random_amount = 100 + random_val;
This is insecure because:

Block timestamps are predictable.

Validators can manipulate timestamps within a range.

Attackers can exploit predictability to maximize rewards.

As a result, rewards that should be fairly distributed can be gamed by attackers, breaking fairness and undermining trust in the system.

##Risk
Attackers can:

Predict outcomes by calculating time % 401 locally.

Run a bot to submit claims only when rewards are highest.

Exploit validator timestamp manipulation to bias results.

Impact:

Unfair distribution of pizza slices.

Reward drain by a single attacker.

Loss of fairness and trust in the application.

##Proof of Concept (PoC)
Attacker bot simulation:

move

fun attacker_bot() {
let t = timestamp::now_microseconds();
let predicted = (t % 401) + 100;

// Only claim when reward is high
if (predicted > 480) {
    pizza_drop::claim_slice();
}

}
Expected Outcome:

Average user reward ≈ 300.

Attacker reward ≈ 480–500 consistently.

System fairness is broken.

##Recommended Mitigation
Avoid using block timestamps as a randomness source.

Option 1: Use Aptos’ VRF (Verifiable Random Function)
move

let random_seed = aptos_std::rand::u64();
let random_val = random_seed % 401;
let random_amount = 100 + random_val;
Option 2: Commit-Reveal Scheme
Users commit to a random seed in one transaction.

Reveal the seed in a later transaction.

Combine seeds from multiple users for final randomness.

This ensures randomness is unpredictable, unbiased, and unmanipulatable.