excalidraw-generation_skill

This skill helps you generate Excalidraw diagrams that convey concepts with hand-drawn aesthetics, spatial layouts, and accessibility-aware design.
  • Shell

10

GitHub Stars

2

Bundled Files

2 months ago

Catalog Refreshed

4 months ago

First Indexed

Readme & install

Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.

Installation

Preview and clipboard use veilstrat where the catalogue uses aiagentskills.

npx veilstrat add skill rhuss/cc-slidev --skill excalidraw-generation

  • README.md10.8 KB
  • SKILL.md30.4 KB

Overview

This skill generates hand-drawn, Excalidraw-style diagrams optimized for developer-facing presentations and informal visual explanation. It combines semantic analysis of the user's request with evidence-based design guardrails to produce accessible, minimal, and aesthetically consistent Excalidraw JSON. I output ready-to-render .excalidraw files and require automated SVG rendering via the provided script.

How this skill works

I analyze the user's prompt or slide content to extract core concepts (entities, relationships, flows) and classify the semantic type (containment, flow, comparison, hierarchy, grouping, annotation). I select an appropriate layout, build the diagram using element factories (rectangles, ellipses, arrows, frames, callouts) while enforcing cognitive and accessibility constraints, then save the JSON to diagrams/<slug>.excalidraw. Finally, I mandate running the supplied render-excalidraw.sh script to produce the SVG; I never embed raw JSON in markdown or attempt manual conversion.

When to use it

  • When the user asks to “create excalidraw diagram”, “generate excalidraw”, “hand-drawn diagram”, “sketch diagram”, or “whiteboard style diagram”.
  • When spatial or conceptual relationships (architecture, containment, hierarchy) are primary.
  • During brainstorming, ideation, or early-stage design reviews where informal visuals help.
  • When annotations or callouts will clarify intent without creating visual clutter.
  • When a single clear idea can be shown within strict cognitive limits.

Best practices

  • Strictly enforce the cognitive-load limit: max 9 visual elements (7±2 recommended).
  • Use the colorblind-safe palette only: Primary blue #3b82f6 and accent orange #f97316; never rely on color alone.
  • Keep total diagram text under 50 words and present one idea per diagram; split complex concepts into multiple diagrams.
  • Ensure text contrast meets WCAG AA (>=4.5:1) and use hand-drawn roughness=1 for informal aesthetic.
  • Always save JSON to diagrams/<slug>.excalidraw and run ${CLAUDE_PLUGIN_ROOT}/scripts/render-excalidraw.sh to produce the SVG — no manual SVG conversion.

Example use cases

  • High-level architecture diagram showing service boundaries and data flow between 3–7 components.
  • Mind map or brainstorming sketch with a central idea and 3–6 branches (radial layout).
  • Vertical containment diagram for platform vs. tenant boundaries using nested frames.
  • Annotated flowchart for an onboarding process with arrows and 3–6 labeled steps.
  • Side-by-side comparison diagram (pros vs cons) with minimal labels and color + shape cues.

FAQ

I generate Excalidraw JSON and save it to diagrams/<slug>.excalidraw; you must run ${CLAUDE_PLUGIN_ROOT}/scripts/render-excalidraw.sh to render the SVG. I never convert JSON to SVG manually.

What if my idea has many parts?

I split complex concepts into multiple diagrams so each remains within the 9-element cognitive limit and under 50 words of text; ask me to create a diagram series if needed.

Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational