본문으로 건너뛰기
cccskills
무료GitHub에 공개

icon-craft

Pick an icon strategy and produce a coherent icon set that doesn't look mass-produced. Solves the "Lucide everywhere" problem — every AI-generated UI uses the same 24px stroke-1.5 outline icons, which is the second-biggest vibecoded tell after Inter-on-purple-gradient. Generates custom SVG sets, tunes existing icon libraries to match archetype + typography, defines stroke weight, optical sizing, and corner-radius rules. Load when the user asks to pick icons, design an icon set, customize Lucide / Phosphor / Heroicons, generate SVG icons, make icons feel custom, says "the icons look generic", "design a custom icon system", "icons for this product", or when frontend-design routes here during icon craft. Sub-skill of frontend-design.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

설치 방법 보기

포함된 파일(4)

  • SKILL.md7.3 KB
  • DEPRECATION.md796 B
  • references/icon-strategies.md4.3 KB
  • references/svg-craft.md3.6 KB

SKILL.md(원문)

설치하기 전에 에이전트에게 전달되는 지시 내용을 확인할 수 있습니다.

Icon Craft

You are the Icon Designer. Generic icons make even careful UIs look mass-produced. You decide the icon strategy for the product, tune existing icon libraries to match its archetype and typography, OR generate a small custom SVG set when the archetype demands it.

Hard Rules

  • Read ARCHETYPE.md and TOKENS.md first. Icons must match the typography weight and the design's stroke language.
  • Pick ONE strategy per product. Mixing icon libraries is a vibecoded tell. If the system needs functional + decorative icons, both must come from the same family or hand-tuned set.
  • Stroke weight must match type weight. Light grotesque body → Phosphor Thin or Light. Bold characterful display → Phosphor Bold or solid icons. Mismatch reads as wrong.
  • Default Lucide is banned. Lucide is acceptable ONLY if every icon used has been retuned (stroke, corner radius, optical size). Default lucide-react drop-in fails this skill.
  • Icons drawn at the size they're used. A 16px icon scaled up to 32px reads pixelated and lazy. Hand-redraw or pick an icon family with optical sizing.

Workflow

Step 1 — Read Inputs

Read ARCHETYPE.md (note icon stance) and TOKENS.md (note typography weights and stroke-adjacent visual language).

Step 2 — Pick Strategy

Read references/icon-strategies.md. Map archetype → strategy:

ArchetypeDefault strategy
b2b-productivitytuned-phosphor (Light/Regular @ 14–16px) OR custom-svg
enterprise-trusttuned-phosphor (Regular) OR Heroicons solid (tuned)
premium-consumercustom-svg (bespoke) — almost always
playful-consumercustom-svg with personality OR Phosphor Bold/Fill
editorialhand-drawn or sparse — icons rare in this archetype
brutalist-distinctivemixed-metaphor (system glyphs, ASCII, hand-drawn marks)
dev-toolcustom-svg light strokes OR Phosphor Light at 14–16px
marketing-landinginherits product archetype's icon stance
creative-toolcustom-svg OR tuned-phosphor Regular at 14–18px
social-feedcustom-svg for action rail (reply/repost/like/share) — these drive the product
conversational-aicustom-svg or tuned-phosphor Light/Regular — small set, high frequency
spatial-canvascustom-svg for toolbar + mixed-metaphor for shape/connector glyphs

If the user has named a constraint ("we want it to feel like Apple's icons"), let that override.

Step 3 — Inventory Required Icons

List every icon the build will need. Categorize:

  • Navigation icons (sidebar, nav bar) — typically 6–12
  • Action icons (buttons, menus) — typically 8–20
  • Status icons (success, error, warning, info, loading) — typically 4–6
  • Decorative icons (feature illustrations, marketing) — typically 0–6

Total: most products need 20–40 icons. If list explodes >50, audit — many "icons" are probably better as text.

Step 4 — Execute the Strategy

tuned-phosphor / tuned-existing

  • Pick ONE weight (Thin / Light / Regular / Bold / Fill / Duotone) and stick to it
  • Adjust stroke to match type weight if SVG-editable
  • Standardize size (e.g., all 16px for inline, all 20px for nav, all 24px for buttons)
  • Test optical balance — some Phosphor icons read heavier; hand-adjust outliers
  • Document the chosen weight + size in ICONS.md

custom-svg

  • Read references/svg-craft.md for craft rules
  • Define grid (24×24 typical, 16×16 for dense), keyline shapes (circle/square/rect baselines), stroke weight, corner radius
  • Draw each icon to the grid, optical adjustments where needed
  • Output as inline SVG components (React/Vue/Svelte) AND as standalone .svg files
  • Each icon ≤ 800 bytes minified

hand-drawn / mixed-metaphor

  • Decide marks: glyph chars (✶ ✻ → ↗ ●), ASCII boxes, hand-drawn doodles
  • Document the system in ICONS.md
  • If hand-drawn, generate via SVG paths or import scanned sketches → cleaned SVG

system-native

  • Use SF Symbols (web export) or Material Symbols
  • Pick ONE weight + ONE optical size
  • Document the choice

Step 5 — Write Icon Outputs

Write to:

  • src/icons/index.tsx (or framework equivalent) — the icon components
  • public/icons/*.svg — raw SVG fallbacks
  • .design/<feature>/ICONS.md — strategy, weight, size, list of icons used, rationale

Step 6 — Audit

Check before handoff:

  • All icons match the type weight visually (do a side-by-side comparison)
  • All icons share grid, stroke, radius, terminal style
  • No mixing of icon libraries
  • Each icon has accessible name (aria-label or title)
  • Decorative icons have aria-hidden="true"
  • No default Lucide drop-in present

Output Format (ICONS.md)

# Icons for [feature]

Strategy: [tuned-phosphor / custom-svg / hand-drawn / mixed-metaphor / system-native]
Source: [Phosphor Light / bespoke / SF Symbols / etc.]
Weight: [thin / light / regular / bold / fill]
Sizes used: [16, 20, 24] (each justified)
Grid: [24×24 / 16×16]
Stroke: [1.25px / 1.5px / 2px / N/A for fill]
Corner radius: [0 / 1 / 2 / 4 px]

## Icons inventoried
| Name | Size | Use |
|---|---|---|
| chevron-right | 16 | nav, accordion |
| ... | | |

## Files
- src/icons/index.tsx
- public/icons/*.svg

Gotchas

  • The most common vibecoded tell after fonts: Lucide at default 24px stroke-2 used everywhere. Even a perfect color system can't recover from it.
  • Phosphor's family of weights (Thin, Light, Regular, Bold, Fill, Duotone) gives you better range than Lucide for tuning to typography.
  • Custom doesn't mean elaborate. 8 well-drawn signature icons beat 80 stock ones.
  • For editorial, the right icon answer is often "no icons" — replace with type, hairlines, or photography.
  • For brutalist-distinctive, ASCII characters (→ ● ✶) often beat any drawn icon.

Reference Files

  • references/icon-strategies.md — strategy matrix per archetype + when to break the default
  • references/svg-craft.md — drawing rules for custom SVG sets (grid, stroke, optical sizing, terminal style, corner radius, exporting clean SVG)

Impact Report

Icon strategy: [name]
Archetype: [name]
Source: [tuned library / bespoke / system / mixed]
Icons inventoried: [count]
Files written:
  - src/icons/index.tsx
  - public/icons/*.svg ([count] files)
  - .design/<feature>/ICONS.md
Anti-Lucide-default audit: [✓ / list of issues]
Handoff to: frontend-design Step 6 (Build)

리뷰

아직 리뷰가 없습니다. 사용해 본 소감을 남겨 주세요.

댓글

로그인 없이 쓸 수 있습니다. 사용 후기, 질문, 작동하지 않은 환경 등을 편하게 남겨 주세요.

아직 댓글이 없습니다. 첫 댓글을 남겨 보세요.

보내면 바로 공개됩니다. 이메일 주소나 전화번호 같은 개인 정보는 쓰지 마세요. URL은 2개까지입니다.

0 / 1,000자

이 양식은 reCAPTCHA로 보호되며 Google 개인정보처리방침 및 서비스 약관이 적용됩니다.

같은 저장소의 스킬

Put on the adversarial hat and systematically attack any document, plan, strategy, or idea to expose its weakest points before commitment. Structured devil's advocate with red team rigour — not pessimism, but evidence-based critique across three phases: diagnostic (are claims accurate?), creative (is the problem artificially constrained?), challenge (are solutions robust?). Load when the user asks to stress test a document, red team this plan, poke holes in this, devil's advocate this, challenge my assumptions, or when product-soul, brainstorming, prd-writing, or inversion calls for adversarial review. Also triggers on "what am I missing", "what could kill this", "find the flaws", or "critique this rigorously".

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

dvy1987/agent-loom32026년 8월 8일 업데이트

Design execution structure for decomposed processes: single agent or multi-agent topology. Load when user says "design an agent for this", "what agent structure do I need", "architect this", "should this be multi-agent", "what's the right execution structure", "agent topology", "how should agents be organized". Takes process-decomposer output as primary input. If triggered directly without a process entry, calls process-decomposer first.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

dvy1987/agent-loom32026년 8월 8일 업데이트

Internal skill. Called by setup-evaluation after a PASS. Launches agents from a validated architecture spec using Claude Code / Ampcode native parallelism (Task tool). Does NOT generate scripts or SDK code — it outputs structured spawn instructions that the platform executes natively. Never invoked directly by the user. Never launches without a setup-evaluation PASS.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

dvy1987/agent-loom32026년 8월 8일 업데이트

Sync library skills from an agent-loom upstream repo into this project's .agents/skills while preserving project-local and forked skills. Load when the user asks to sync agent-loom, update skills from upstream, rsync from ../agent-loom, pull new library skills, upgrade installed skills, or refresh the .agents folder without losing custom project skills. Also triggers on "sync skills from agent-loom", "update my agent skills", "pull skill library updates", or "merge agent-loom improvements into this repo".

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

dvy1987/agent-loom32026년 8월 8일 업데이트

Instrument a shipped product's AI agents with tracing and observability so you can see what they did, why outputs happened, and what each run cost. Plain-language primer plus free-tier-first backend selection (Langfuse, Phoenix, LangSmith, Braintrust) and OpenTelemetry/OpenInference instrumentation. Load when the user asks to add observability, add tracing, instrument my agents, see what my agent is doing in production, set up Langfuse or Phoenix or LangSmith, debug why my agent gave a bad answer, or track LLM cost per request. Also fires when agent-system-architecture or setup-evaluation requires an observability plan for an agent-chain product. NOT for tracing the coding agent itself — that is run-trace. Precondition for runtime-learning-loop.

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

dvy1987/agent-loom32026년 8월 8일 업데이트

Run a structured retrospective after development-phase runs of your product's agents — interview the owner in plain language about what went well and poorly, draft ranked improvement hypotheses, then design and run small n=1/n=2 experiments with pre-declared success criteria, guardrails, stop conditions, and a cost/ROI kill-switch. Load when the user says how did that run go, retro this run, the agent output was bad, what should we improve, draft hypotheses, run a small experiment, or after repeated dev runs of an agentic system produce uneven quality. Priority: output quality over performance over cost, each with diminishing-returns stops. NOT a product A/B test (experimentation), NOT coding-agent harness repair (harness-evolution), NOT production-scale learning (runtime-learning-loop).

한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.

dvy1987/agent-loom32026년 8월 8일 업데이트

dvy1987의 스킬 모두 보기

이 스킬의 문제 신고