From itsme
How to design and create itsme project icons. Load this when the user asks to design / create / add / modify any icon.
How this skill is triggered — by the user, by Claude, or both
Slash command
/itsme:iconsThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
> When loaded, follow this workflow **verbatim**. Do not freehand SVG.
When loaded, follow this workflow verbatim. Do not freehand SVG.
1. STOP — read this whole file before touching SVG.
2. READ — open docs/ICONS.md §3 §4 §5.
3. COPY — duplicate docs/assets/icons/_template.svg → <scope>-<name>.svg.
4. EDIT — only the inner shapes. Never touch viewBox / stroke / fill.
5. VERIFY — run `python3 scripts/lint-icons.py`. Must pass.
6. PREVIEW — append/update row in docs/assets/icons/README.md table.
docs/ICONS.md §4)| Attribute | Required value |
|---|---|
viewBox | 0 0 24 24 |
Root <svg> width / height | omit (let consumer size) |
fill (root) | none |
stroke (root) | currentColor |
stroke-width | 2 |
stroke-linecap | round |
stroke-linejoin | round |
Allowed fill="currentColor" | only on intentional accent dots / stars |
Hard-coded colors (#hex, rgb(), named) | ❌ banned |
<linearGradient> / <radialGradient> / <filter> | ❌ banned |
Root <svg transform> | ❌ banned |
Embedded fonts / <text> | ❌ banned (paths only) |
docs/assets/icons/<scope>-<name>.svg
Allowed scopes: brand (no prefix — just itsme.svg), engine, verb, hook.
Examples — good: verb-remember.svg, engine-aleph.svg, hook-before-exit.svg.
Bad: Remember.svg, verb_remember.svg, aleph-icon.svg.
Before reporting done, eyeball the icon and confirm:
currentColor)If the user says "make me an icon" without specifying scope/name, ask:
Then propose 1-2 concept sketches in chat first before writing files.
After creating one or more icons, end your reply with:
✅ Created: <list>
✅ Lint: PASS / FAIL (if FAIL, show errors)
✅ Preview: docs/assets/icons/README.md updated
npx claudepluginhub coxon/itsme --plugin itsmeCreates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.