Keybox.AI Pilot

This tool allows users to inspect smart contract source codes from more than ten blockchain explorers that offer accurate syntax, such as Etherscan. It provides a direct download feature, streamlining access across multiple chains.

The Google LLM AI service also deciphers the smart contract code into plain language and offers a preliminary risk analysis. However, it's worth noting that AI-driven risk evaluations for smart contracts have their limitations.

We archive the AI-generated reports in web3 storage to enhance our system's capabilities, aiding in subsequent machine learning data accumulation.

• Key Chains Supported: Ethereum Mainnet, Base Mainnet, Linea Mainnet, Scroll Testnet, Starknet Mainnet, XDC Network Mainnet, Mantle Mainnet, ETH Goerli Testnet, Polygon Mainnet, Polygon Mumbai Testnet, Arbitrum Mainnet, Arbitrum Goerli Testnet.
• Smart Contract Viewer with pretty syntax highlighter
• Google AI-powered Smart Contract Translator and Risk Assessment
• You can store the AI report in Web3 Storage for further research.

For this project, I leveraged Next.js, a distinguished React framework known for features such as hybrid rendering, smart bundling, and route pre-fetching. I kickstarted the development using create-next-app to ensure a seamless setup and maintain a cohesive project structure. The project taps into multiple blockchain data provider APIs, including Etherscan, to gather comprehensive smart contract data. This data is then processed through Google Cloud's AI service for translation and risk evaluation. Additionally, web3storage plays a crucial role in archiving the derived insights.

Example Addresses & Etherscan Integration: The project offers a collection of sample blockchain addresses. It leverages the Etherscan API to retrieve content related to these smart contracts.

Google AI's LLM Connection: The project is integrated with Google AI's LLM, a sophisticated language model.