A tool that makes it easy to generate an ERC-7730 JSON file from a smart contract.
At ETHGlobal Cannes, some participants were scammed through blind signing, thinking they were minting an NFT granting access to exclusive side events.
To protect users from such attacks, Ledger introduced a clear signing system on its wallets, displaying human-readable transaction descriptions. This system relies on a JSON file following the ERC-7730 standard, which provides metadata to describe smart contract operations.
However, Ledger’s tools for generating these files are still complex to use. I took their open-source project, “ERC7730 JSON Builder,” and improved it to make it more accessible. I added support for multiple EVM-compatible chains (Ethereum, BNB Chain, Arbitrum, Base, Avalanche) and integrated AI-powered autocompletion that analyzes a contract’s ABI to suggest proper ERC-7730 fields.
I also included a feature that lets users instantly submit the generated JSON as a pull request to Ledger’s repository. With this tool, generating and publishing an ERC-7730 file takes just a few minutes, making it easier for developers to adopt and improving overall security across EVM blockchains.
I started with Ledger’s open-source project “ERC7730 JSON Builder”, with the goal of improving it and making it more accessible to developers.
First, I added the ability to import JSON files directly through the UI. I then implemented intelligent field autocompletion by integrating the OpenAI API (using the gpt-4.1-mini model), enabling the app to automatically infer and populate ERC-7730 fields from a smart contract’s ABI.
The original project only supported the Ethereum mainnet, so I extended it to support multiple EVM-compatible chains (BNB Chain, Arbitrum, Base, Avalanche, and more).
Finally, I added a feature that lets users automatically generate a pull request to Ledger’s official ERC‑7730 repository. AI is also used to write a summary description of the proposed metadata.
Tech stack: