Skip to content

Web AI Tools

Jinny You edited this page Apr 16, 2026 · 1 revision

WebCanvas ships with built-in support for AI-assisted development. Two complementary pieces cover different use cases.


SKILL: thorvg-webcanvas

SKILL.md is a structured prompt guide for AI coding agents (Claude Code, Cursor, Copilot, etc.). It tells the agent:

  • When to reach for @thorvg/webcanvas
  • Renderer Selection Guidance (gl / wg / sw)
  • Advanced patterns to follow (initialization, animation callbacks, memory cleanup)
  • Common tasks with copy-paste code snippets based on Playground

Place skills/**/SKILL.md in your project root (or point your AI tool at it) so the agent has full context when writing WebCanvas code.

File: packages/webcanvas/skills


Machine-Readable API Docs

llms.txt and llms-full.txt follow the llms.txt standard — a convention for exposing library documentation in a format AI tools can reliably consume.

  • llms.txt — a structured index of all APIs, classes, and methods with one-line descriptions
  • llms-full.txt — the complete API reference in plain markdown, including parameters, return types, and examples

Both files are auto-generated from TypeDoc on every build and included in the npm package. AI tools and MCP servers can fetch them directly to answer questions about the API without hallucinating method signatures.

URLs:

Clone this wiki locally