Back to directory

Taste Skill

A collection of installable AI-agent skills that add stronger layout, typography, motion and spacing rules to interfaces an agent builds, plus separate skills for generating reference images for web, mobile and brand-kit designs.

Taste Skill

Taste Skill is a set of portable Agent Skills that improve the visual quality of interfaces built by AI coding agents, replacing boilerplate-looking layouts with stronger typography, motion and spacing. The repository also includes image-generation skills that produce reference boards for web, mobile and brand-kit designs, meant to be paired with an image generator and then handed to a coding agent, such as Codex, Cursor or Claude Code, for implementation.

Skills install with a CLI that scans the repository's skills folder, either all at once or one at a time by its install name, and a skill file can also be copied directly into a project or pasted into a chat conversation. The current default skill is a version 2 rewrite of the original, and the previous version remains available under its own install name for projects that depend on its exact behavior.

It is MIT licensed.

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.