Sublayer

Sublayer is a Ruby framework that provides a consistent, model-agnostic DSL for building generative AI agents. It abstracts away differences between LLM providers so developers can focus on constructing agents from composable components: Generators, Actions, Tasks, and Agent triggers. The framework enforces a clean separation of concerns through base classes, making it easy to create AI-powered applications with Ruby. Designed for Ruby developers, Sublayer streamlines the integration of large language models into existing Rails or Ruby applications. It supports dynamic output structures, custom trigger types, and adapter-based model switching, enabling flexible AI workflows without vendor lock-in. The library's core philosophy is self-assembly—agents automatically determine which actions to take based on context. What sets Sublayer apart is its Ruby-native design and emphasis on maintainable, testable code. Unlike Python-centric frameworks, Sublayer feels familiar to Rubyists and works seamlessly with Ruby's object-oriented patterns. It also supports running LLMs locally via llamafile or Ollama, catering to developers who prioritize data privacy or offline capabilities. Sublayer is open-source and free to use, with no paid tiers or enterprise plans announced. It's best suited for Ruby developers who want to experiment with AI agents without leaving their comfort zone, though it may lack production-grade orchestration features found in heavier tools.