Using Cornerstone with AI
Cornerstone publishes machine-readable documentation so AI coding assistants can understand its components and help you write better code. Whether you’re using Claude, ChatGPT, Copilot, Cursor, or another tool, you can give it context about Cornerstone’s APIs, properties, events, slots, and more.
AI-Ready Documentation
We provide two formats for giving AI tools context about Cornerstone. Both formats are generated automatically with every Cornerstone build and are available in your node_modules directory after installing via npm.
Two complementary skills that AI tools load progressively, fetching only the documentation a task needs.
A single text file holding the full API reference for every component, for any tool that takes pasted context.
AI Policy
We’re a small team, and like most people building things these days, we use AI tools regularly. They help us move faster: exploring ideas, writing boilerplate, catching edge cases, and getting more done in less time.
But Cornerstone is an open-source library that a lot of people rely on, so we want to be clear about how we use AI.
Human-Led, AI-Assisted
Everything in Cornerstone is built and owned by humans. AI is a capable assistant, but it never gets the final say. The humans on our team are always responsible for:
- The original ideas and vision
- The architecture and design decisions
- The critical thinking about what’s right for the library and its users
- The review, testing, and polish at the end
No component, page, or feature ever ships as pure AI output. There’s always substantial human direction, judgment, and final approval behind it.
This lets us ship better work more quickly, without giving up the human judgment that production-ready software demands.
Human ideas. Human architecture. Human verification. AI just helps us build what we were going to build anyway… only faster.