Back to directory

AI Design Skills Collection

A collection of 44 agentic skills and 18 commands across six plugins that give an AI agent working vocabulary and practices for designing AI products, from model interaction patterns through alignment, evaluation and prompt architecture.

AI Design Skills Collection

AI Design Skills Collection packages the vocabulary and practices of what it calls Agentic Experience Design, an emerging discipline covering ideas such as mixed-initiative flow, harm anticipation, handoff protocols and error personality, into skills an AI agent can load while a designer is working on or auditing an agentic product.

The six plugins map to six layers of the discipline: model interaction, alignment reasoning, system behavior, evaluation, agent orchestration and prompt architecture, each with seven skills and three commands. The underlying ideas come from alignment research and HCI scholarship, and the repository documents, paper by paper, what was adopted, what was reframed, and how each academic concept became a practitioner skill.

It installs as a Claude Code or Gemini CLI plugin, and is written in Python.

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.