38,000 lines of Rust, one developer, three AI models doing the heavy lifting. That's the recipe behind eden-skills, a new open-source CLI that takes a Terraform-style approach to managing AI agent skills.
The project grew out of frustration with existing skill managers like skills.sh, which rely on Node.js runtimes and break across environments due to fragile symlink handling. Eden-skills replaces all of that with a single ~10MB Rust binary built on Tokio (Rust's async runtime). You define your agent skills in a skills.toml config file, and the tool handles installation, versioning, and environment setup deterministically - same config in, same result out, every time.
The more interesting story here isn't the tool itself but how it got built. The developer treated three separate AI models as an engineering team, dividing responsibilities across them throughout the project. At 38,000 lines, this isn't a weekend hack or a demo app - it's a substantial systems tool with real architectural decisions baked in.
What This Actually Tells Us About AI-Assisted Development
We're past the phase where "I built X with AI" means someone auto-completed a few functions. Projects at this scale require the developer to act more like an engineering manager than a line-by-line coder: defining architecture, reviewing AI-generated code for correctness, catching subtle bugs that models introduce, and maintaining consistency across a large codebase.
The fact that the result is a compiled Rust binary - not a Python script or a JavaScript app - adds another layer. Rust's strict compiler catches a lot of errors before runtime, which likely made AI-generated code easier to validate. That's a pattern worth watching: languages with strong type systems may turn out to be better targets for AI-assisted development than dynamically typed ones, precisely because the compiler acts as a second reviewer.
Eden-skills is open source and available now. For anyone managing AI agent configurations across multiple environments, it's a leaner alternative to Node.js-based tooling.