Genotype

Genotype is a programming language built for teams that work with multiple programming languages, such as a TypeScript front-end and a Rust back-end. It facilitates the creation of API clients, file formats, standards, and supports gradual language migration.

The language is designed to be small and quick to learn, with a stated learning curve of 15 minutes. Despite its small size, Genotype is feature-rich, supporting complex data structures, generics, and recursive data types. It also allows for target-specific fine-tuning through annotations.

Genotype is unopinionated and adapts to existing ecosystems rather than creating new ones. It can generate Zod schemas instead of plain types, produce ready-to-ship packages, or integrate code directly into existing modules. It also provides support for legacy Python versions.

The code generated by Genotype is clean, follows idiomatic naming conventions for types and files, and adheres to common programming practices. Furthermore, Genotype is designed to be AI agent-ready, providing 'skill' and 'llms.txt' files, and embedding code comment breadcrumbs to guide Large Language Models (LLMs) in its correct usage.

Unlike Protocol Buffers, Genotype does not impose a transport or storage format. Its primary focus is on generating clean, idiomatic types with minimal to no runtime overhead, rather than optimizing data transfer performance. For example, the TypeScript output has no runtime.