Back to directory

Sparrow

Sparrow is a spec-driven, domain-driven-design framework for AI coding assistants that turns raw business requirements into production code through an eight-step pipeline, using each tool's own native AI rather than a separate agent framework.

Sparrow

Sparrow introduces an Interaction Context as a first-class concept alongside the usual DDD bounded context, handling frontend UI and backend-for-frontend aggregation through the same standardized design, model, plan and apply workflow that backend contexts use, with no dependency between the two so they can be worked on in parallel. Its workflow skills are split into a sequential process pipeline that carries a change from requirement through verified, archivable code, and supporting workflows such as a harness and a reconcile command that can run independently at any point.

It works directly with Claude Code, OpenCode, Cursor and Pi without pulling in an agent framework such as CrewAI or LangChain, produces version-controlled Markdown artifacts at every step so decisions stay traceable, and supports Java, Python, Node.js and TypeScript, Go, Rust and C++, with each bounded context free to use a different stack.

It is published on npm as sparrow-ddd, written in TypeScript, and released under the MIT license.

More in Agent skills

See all 581
  • Provenance gate

    github.com

    A deterministic authorization gateway that sits between an AI agent and its tools, tracing where each parameter of a side-effect call came from, and blocking the call or asking a human when the source does not check out.

  • A markdown framework of system rules that restates Isaac Asimov's Four Laws of Robotics as a set of prioritized loss calculations for large-language-model agents to follow, published as a discovery file for agents and crawlers to read.

  • Vitruvius

    github.com

    An engineering research agent that runs a discover, read, synthesize, verify and review loop over technical questions, reading sources directly rather than inferring from titles, and publishing its own scored verifier benchmark, weaknesses included.