Implementation
Score & Summary
In practice, our AI Contract Report returns an overview & summary of the elements of the flagged parts of the contract. At its very core, the AI contract report returns a score, as seen in the top right, which is a standalone assessment of how safe the contract is. This score is generated through our proprietary Cognitive Framework that also takes into account how the contract ranks against other safe tokens. It also uses historical on-chain data, assessing the flagged functions against tokens previously that have 'rugged'. As an advanced framework, our AI Contract Report takes into account a multiplicity of different factors to determine the overall score, which consequently makes up a proportional aspect of overall score. Another aspect of this report is the explainability of complex functions, distilled into a readable, understandable format. Regardless of expertise, the majority of users are able to comprehend and understand what are complex vulnerabilities in tokens. This can allow users to make informed decisions on whether or not they want to interact with tokens, as well as more broadly improve understanding of how smart contracts work, and what the crucial elements are.
Within the AI Contract Report, we have embedded the ability to see the specific functions that have been flagged, and their relative severity. The severity is a means of explaining which flagged functions are worth noting with the cognitive framework giving an explanation as to the nature of the flagged function and its severity. Much akin to the overall explanation, our Cognitive Framework can simply explain what each function within a contract does, and if it has been flagged, why it's been flagged. On top of that, it's able to point out the code snippet in the source code for more advanced developers to review. Some function flags are less severe than others, and the ability to flag source code is an effective way to review whether the flagged functions are as severe as has been noted. Much of this proprietary technology is advanced in nature; the ability to flag dangerous contract functions, and then explain them to a range of users is a leap not yet seen in the field of auditing.
Last updated
