Google's official agent skill for writing production Maps Platform code — grounded in freshly fetched docs, with a demo key path that needs no billing account.
Stitch: shadcn/uiSkill
Summary
Google's official Stitch build skill for shadcn/ui — component discovery, installation, customisation and the conventions that keep a generated UI idiomatic.
Features
- Discovers which shadcn/ui components fit a given interface
- Handles component installation through the shadcn CLI
- Covers customisation patterns rather than overwriting generated component source
- Documents shadcn/ui conventions and common mistakes
- Pairs with the Stitch design skills to turn a generated design into real components
Install This Skill
Add this skill to your favorite AI agent in a few steps.
Skill Content
Usage Instructions
Learn how to use this skill with different AI agents.
Description
shadcn/ui is not a dependency you install and import — it copies component source into your
project, which is exactly why agents get it wrong. They re-install components that already exist, edit generated
files in ways the next add will overwrite, or hand-roll a component the registry already provides.shadcn-ui is Google Labs' official skill for getting that right.
It covers the full cycle: discovering which components in the registry match the interface being built, installing
them through the shadcn CLI, and customising them along the grain of the library's conventions instead of against
them. It is the "build" half of the Stitch suite — where the design skills produce a screen, this one turns that
screen into real, idiomatic React components.
The skill declares access to the shadcn MCP tools (shadcn*:*, mcp_shadcn*) alongside file and shell tools, so it
can query the registry directly rather than relying on a model's recollection of which components exist and what
props they take — a meaningful difference for a library that adds components regularly.
Useful on its own for any React project using shadcn/ui, not just Stitch work. Apache-2.0 licensed.
Related Skills
Railway's official agent skill: create projects, provision databases and buckets, deploy, manage variables and domains, and read build failures back — from the CLI, API or MCP server.
Diagnoses wrong gradients in differentiable NVIDIA Warp programs by measuring first — comparing autodiff against finite differences on a shrunk reproduction before proposing any fix.
Google's official skill for the gws CLI — drive Gmail, Drive, Calendar, Sheets, Docs, Chat and Admin APIs from an agent, with Model Armor screening.