ASCAI

Imagine that you need to stake something in a contract, but you lack significant knowledge about smart contracts. How would you determine what's happening in the backend? How would you verify whether it's safe or not? To address this issue, we've developed a MetaMask snap. This snap acts as a shield, enhancing transparency for users who interact with the contract without a deep understanding of its functionality. We have used Tenderly to simulate the transaction and inform the user beforehand about what will happen if the transaction gets executed. We offer crucial insights through visual representations and also issue alerts if vulnerabilities are detected within the smart contract.

For our Minimum Viable Product (MVP), we are utilizing prebuilt AI models. However, for future iterations, we intend to create and fine-tune our custom model. This AI model will perform Natural Language Processing (NLP) and present its output to the user, enabling even non-technical individuals to comprehend the ongoing processes.

In constructing this solution, we have harnessed a combination of advanced tools and technologies. To ensure a comprehensive understanding of the process, we employed the following key components:

Tenderly for Transaction Simulation: We have integrated Tenderly into our framework, leveraging its capabilities to simulate transactions. By doing so, we allow users to gain insights into the potential outcomes of executing transactions. This proactive approach helps users make informed decisions before proceeding with any actions.

MetaMask Snap for Transaction Details: The MetaMask snap plays a pivotal role in our solution. It acts as a conduit, enabling users to retrieve intricate transaction details prior to their execution. This feature enhances transparency and empowers users to comprehend the impending consequences of their actions within the contract.

OpenAI for Natural Language Processing (NLP): OpenAI's advanced Natural Language Processing (NLP) technology forms the core of our solution's backend. We harness the power of OpenAI to process the simulated transactions generated by Tenderly. This processing results in clear and coherent explanations of the transaction processes, even for individuals without a strong technical background.

Front-End Development with React: For the user interface, we turned to React, a widely acclaimed JavaScript library. By employing React, we crafted an intuitive and responsive front-end experience. This allows users to effortlessly interact with the simulation outputs and the transaction details presented by the MetaMask snap and OpenAI's NLP.