Author and manage user-facing documentation for the Grida Canvas editor. Covers writing style, tone, product screenshots, demo state preparation, and content structure for pages under docs/editor/. Use when creating or editing canvas editor documentation, capturing screenshots of canvas features, or planning visual demos. Trigger phrases: "write user docs", "document this feature", "screenshot for docs", "canvas user guide", "update help page".
80
100%
Does it follow best practices?
Run evals on this skill
Adds up to 20 points to the overall score
View guide
Passed
No findings from the security scan
Guide for pages under docs/editor/ — user-facing documentation for the
Grida Canvas editor at grida.co/docs.
Not for: docs/wg/, docs/reference/, forms, CLI, or architecture docs.
| Item | Value |
|---|---|
| Source | docs/editor/ |
| Site | Docusaurus in apps/docs/ |
| Product | Canvas editor at http://localhost:3000/canvas |
| Build | turbo build --filter=editor && turbo start --filter=editor |
See docs/AGENTS.md for linking rules and MDX caveats.
<kbd> tags. Values/filenames in backticks.https://grida.co/_/<path>.format: md frontmatter unless the page uses JSX/MDX features.---
title: Feature Name
description: One-line summary.
format: md
---
# Feature Name
Brief intro (1-2 sentences). What is it, why use it?
## How to Use / Creating a \_\_\_ / When to Use
Primary workflow or step-by-step.
## How It Differs from X (if applicable)
Comparison table.
## Nesting / Hierarchy Rules (if applicable)
## Default Appearance (if applicable)
## Keyboard Shortcuts (if applicable)Headings: ## for sections, ### for subsections. No ####+.
| Property | Value |
|---|---|
| Output size | 960 x 960 px (1:1) |
| Format | WebP q90 (preferred), PNG fallback |
| Theme | Default light |
| Naming | <feature>-<description>.webp (kebab-case) |
| Location | Co-located with the doc or sibling img/ |
960x960 is the cropped output, not the viewport. Capture at any size, then crop to the area of interest and resize. Crop aggressively — a single panel or dialog is better than a full-window capture.
Prefer scripting over manual interaction, in this order:
.grida or .svg from examples/fixtures/ or test/.globalThis.grida exposes the editor instance
(via WindowGlobalCurrentEditorProvider). Use javascript_tool or console
to create/modify nodes, set state, rename layers programmatically.?src=<url>, ?backend=canvas|dom.Before capture: name layers meaningfully, deselect (Escape), zoom to fit. Use realistic labels ("Login", "Dashboard") — never lorem ipsum.
localhost), not dev modeFor workflows, before/after, or relationships a screenshot can't show.
For SVG figures specifically — gestures, alignment, before/after diagrams that
recreate canvas UI — use the docs-svg-kit skill. It provides a
starter template, a primitives catalog (selection chrome, size badges, anchor
pins, resize cursors, ripples), and finished examples to crib from. Saves you
from re-deriving stroke widths, colors, and the half-pixel alignment tricks.
title and description in frontmatterformat: md (unless using MDX)pnpm --filter docs startdocs/wg/.2e0d276
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.