karalabe/darkbio-crypto-fl

This project, darkbio-crypto-fl, brings post-quantum cryptography to Flutter, Google's popular toolkit for building mobile and desktop apps from a single codebase. In plain terms, it helps developers protect data inside their apps against future threats, specifically, the risk that powerful quantum computers could eventually break the encryption we rely on today. At a high level, the library is a thin layer that sits on top of existing cryptography tools written in Rust, a language known for performance and safety. Rather than building new cryptographic algorithms from scratch, it carefully selects and combines proven ones, then exposes them through a single, consistent interface for Flutter developers. The design is deliberately opinionated: instead of offering endless configuration options, it picks specific security parameters and algorithms that are believed to withstand quantum attacks. The goal is to make the tools easy to use correctly and difficult to use incorrectly. The library covers the essentials of secure communication: digital certificates and signatures for proving identity, encryption for keeping data confidential, key derivation for turning passwords into strong keys, and serialization for packaging data into a standard format. It uses emerging standards that combine traditional cryptography (like Ed25519 and X25519) with new post-quantum algorithms (like ML-DSA and ML-KEM), often pairing them together so that data remains secure even if one approach turns out to be weaker than expected. This would appeal to developers building Flutter apps that handle sensitive information, think healthcare records, financial transactions, or private messaging, where there is a need to anticipate future security threats. A startup building a privacy-focused notes app, for example, might use this to ensure that encrypted user data stays protected for years, even as quantum computing advances. It is worth noting that the project is built for a specific initiative called the Dark Bio project, so the feature set is intentionally narrow and focused on exactly what that effort requires.