create-svg-illustrations — Claude Code skill: install & use
create-svg-illustrations — a skill for Claude Code and Codex that generates, tests, and embeds accessible SVG diagrams. Runs when building graphics for slide decks, static docs, or web pages.
MIT 211 installs ★ 18 updated 2026-10-22
npx skills add narumiruna/skills --skill create-svg-illustrations -a claude-codeWhat create-svg-illustrations does
The skill creates editable SVG graphics with accurate viewBox attributes, explicit styling, and semantic title tags for screen readers. It keeps files portable by avoiding embedded font files, unexplained emoji, heavy filters, and base64 strings.
Code checks run through svglint when the utility is available locally. When preparing assets for Marp presentations, it pairs the vector markup with host-level alt descriptions, since built-in SVG metadata is not picked up by every deck viewer.
Output starts with the raw SVG markup followed by validation notes and font warnings. The skill does not commit changes or modify repository files without an explicit prompt.
What it helps with
- generate accessible SVG flowcharts for slide presentations
- embed vector assets into Marp Markdown decks with alt text
- validate SVG markup and viewBox dimensions with svglint
- strip base64 blobs and embedded fonts from vector assets
Example prompts for the agent
- Create an accessible SVG architecture diagram for my documentation
- Embed this vector chart into a Marp slide deck
- Inspect this SVG for scaling issues and validate it with svglint
How to install create-svg-illustrations
1Skills CLI (Vercel)
npx skills add narumiruna/skills --skill create-svg-illustrations -a claude-code2Manually: put the files in place
~/.claude/skills/create-svg-illustrations/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/create-svg-illustrations/
1Skills CLI (Vercel)
npx skills add narumiruna/skills --skill create-svg-illustrations -a codex2Manually: put the files in place
.agents/skills/create-svg-illustrations/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add narumiruna/skills --skill create-svg-illustrations -a cursor2Manually: put the files in place
.cursor/skills/create-svg-illustrations/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install narumiruna/skills visuals/skills/create-svg-illustrations2Manually: put the files in place
.github/skills/create-svg-illustrations/SKILL.md1Skills CLI (Vercel)
npx skills add narumiruna/skills --skill create-svg-illustrations -a gemini-cli2Platform CLI
gemini skills install https://github.com/narumiruna/skills --path visuals/skills/create-svg-illustrationsManually: put the files in place
.devin/skills/create-svg-illustrations/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add narumiruna/skills --skill create-svg-illustrations -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about create-svg-illustrations
How do I install create-svg-illustrations in Claude Code?
Run npx skills add narumiruna/skills --skill create-svg-illustrations -a claude-code or save the rule file to ~/.claude/skills/create-svg-illustrations/SKILL.md.
What does create-svg-illustrations do?
It builds and fixes accessible SVG diagrams for slides and documents. The skill structures valid viewBox attributes, avoids bloat like base64 payloads, and adds screen-reader metadata.
What tools are required for create-svg-illustrations?
The workflow uses svglint for validation when the linter is installed in your local environment.
Source text
SKILL.md · MIT
--- name: create-svg-illustrations description: Create, revise, embed, or troubleshoot accessible SVG diagrams and illustrations for slide decks, documents, and other static artifacts, with portable sizing, layout, style, and validation guidance. --- # SVG Illustrations Start with `references/core-rules.md`; load only the extra detail the artifact needs. | Need | Read or use | | --- | --- | | Canvas, style, text, accessibility, validation | `references/core-rules.md` | | Reusable compositions | `references/pattern-examples.md` | | Marp embedding | `references/embedding.md` | | Rendering failures | `references/troubleshooting.md` | | Known-good deck example | `assets/examples/with-diagrams.md` | ## Workflow 1. Inspect the target slide or document, required dimensions, surrounding palette, and renderer. 2. Create the smallest editable SVG that communicates the idea. Use a correct `viewBox`, explicit text styles, consistent geometry, and SVG title/description where the visual carries meaning. When embedding, also provide descriptive host-level alt text or an adjacent semantic equivalent as required by `references/embedding.md`; internal SVG metadata alone is not sufficient. 3. Keep assets portable: avoid embedded fonts, unexplained emoji, unnecessary filters, and base64 content unless the output must be self-contained. 4. Validate with `svglint` when available, then open or embed the SVG in the target artifact and inspect clipping, scaling, contrast, text, and relative paths. 5. Return the SVG artifact first, followed by validation performed and any renderer or font caveat. Use slide-authoring or color guidance only when the request also requires it. Do not claim visual inspection when only source validation ran, and do not commit or publish the artifact unless requested.
© narumiruna · MIT · narumiruna/skills