Tldraw-docs skill

tools
A Claude Code skill that teaches an agent how to navigate the tldraw canvas SDK documentation without hallucinating API signatures or grabbing the wrong bundle.
Author

vade-coo

Published

2026-05-11

Open source: coo-labs/skills/v0.3.0/skills/tldraw-docs

The tldraw team publishes its documentation in markdown bundles explicitly designed for language-model consumption (tldraw.dev/llms*.txt plus per-page .llms.md URLs). The naive failure modes when agents work with tldraw are hallucinating Editor.createShape() argument shapes from memory, grabbing the full mega-bundle when a narrow page would do, and guessing at topic names that don’t exist in the docs.

This skill teaches which bundle to fetch when, which pages cover which features (ShapeUtil, BindingUtil, persistence, snapshots, side effects, the store/signals system), and how to anchor a doc lookup so the answer comes from the docs rather than the agent’s prior.

Companion: canvas-ui

The sister skill canvas-ui is the patterns layer: anti-patterns extracted from a production tldraw-based codebase. The two compose — tldraw-docs tells you which page to read; canvas-ui tells you which mistakes are already documented in our cycles.

Install

Drop the skill directory into your project’s .claude/skills/. The setup script in the public repo does this for you.

Back to top

Reuse

CC-BY-4.0