Fast and scalable analysis of smart contracts

We design and implement a fast and scalable fuzzing framework for Ethereum smart contracts. It provides a way to encode custom properties into a contract that will automatically be checked. The fuzzer takes a truffle project as input, fuzzes contracts and for each contract it reports coverage, integer overflows, assertions, any violation of custom properties and statistics about failed/succeeded transactions for each method. It is configurable to adjust the fuzzing strategy for a specific domain. We report benchmarks by testing it in several settings including over 30k real-world contracts deployed on the main chain of Ethereum.