Back to directory

baoyu-design

An Agent Skill that runs Claude Design locally inside Cursor, Claude Code, or another file-capable coding agent, producing UI mockups, prototypes, wireframes, and decks as self-contained HTML.

baoyu-design

baoyu-design packages Claude Design, the design engine behind the claude.ai design product, as a portable Agent Skill that runs on a local coding agent instead of the website. Dropped into Cursor, Claude Code, Claude Desktop, or another file-capable agent, it produces UI mockups, interactive prototypes, wireframes, landing pages, dashboards, mobile app designs, and slide decks, all as self-contained HTML files that stay in the user's own repository rather than requiring a website visit, a separate subscription, or an upload step.

It drives a full design process of clarifying questions, gathering design context, producing one or more HTML deliverables, and previewing and verifying the result, backed by a set of built-in skills for areas such as high-fidelity design, interactive prototypes, wireframes, decks with speaker notes, mobile and motion design, and website and landing page layouts. Because the output is plain HTML served locally, an agent's built-in browser preview and element-annotation tools can be used to point at something in the live preview and describe a change, which the agent then applies to the underlying source.

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.