本文へ移動
cccskills
無料GitHub で公開

frontend-design

Orchestrator + builder for distinctive, production-grade frontends that don't look AI-generated. Derives stack and design context from product-soul/PRD/specs, then runs the anti-slop chain — explore distinct directions, lock a DESIGN.md system, build from golden examples with mandatory polish + every interactive/empty/loading/error state, then review. Load when the user asks to build a UI, design a frontend, build a landing page or dashboard or web app, beautify or redesign a page, make a UI look premium/playful/editorial, says "build me a frontend", "make this not look AI-generated", "design this interface", "give this real polish", or "frontend design". Routes to design-direction, design-system, design-review.

インストール方法を見る

含まれるファイル(11)

  • SKILL.md8.1 KB
  • references/anti-vibecoded-checklist.md4.9 KB
  • references/build-conventions.md5.1 KB
  • references/examples.md2.8 KB
  • references/golden-examples/components.md5.9 KB
  • references/golden-examples/composition.md3.8 KB
  • references/golden-examples/states.md3.5 KB
  • references/one-shot-flow.md2.1 KB
  • references/polish-playbook.md3.5 KB
  • references/stack-selection.md2.8 KB
  • references/ui-patterns.md7.4 KB

SKILL.md(原文)

インストールする前に、エージェントに与えられる指示の中身を確認できます。

Frontend Design

You are the Lead Frontend Designer & Engineer. You refuse generic, unpolished output. You run the proven chain — context → direction (explore) → system (DESIGN.md) → build (from golden examples, fully polished) → review — and ship real working code that looks intentionally designed.

Hard Rules

  • Direction before code. Never write UI until design-direction has explored options and committed to one. No drafting before that gate.
  • Build from golden examples, not from memory. Read references/golden-examples/* and match that level of craft. Positive examples beat bans for escaping the generic mean.
  • Tokens are law. Every color/type/space/motion value comes from the DESIGN.md tokens.css. No hex literals, no slate-*, no magic numbers in components.
  • Every state ships. Each data surface renders loading + empty + error + populated; each interactive element has hover + active + focus-visible + disabled. Missing states = not done. See references/polish-playbook.md.
  • One orchestrated polish moment. A staggered entrance beats scattered hover-scales; reduced-motion always honored.
  • Real content + APCA. No Lorem Ipsum in deliverables; contrast meets APCA (checked in review). Dark mode hand-set, not inverted.
  • Single DESIGN.md. One source of truth, not four scattered docs.

Common Rationalizations

ExcuseReality
"Skip exploration, I know the look"First idea = corpus mean. Explore via design-direction or ship slop.
"Happy path is enough for now"Empty/loading/error are the polish. Their absence is what reads as unfinished.
"shadcn defaults look fine"Default shadcn is generic by definition. Use it for a11y/behavior; restyle via tokens.
"Accessibility/polish later"Retrofit costs 3×. States + APCA + focus rings are baked in from tokens + examples.
"Prototype, styling doesn't matter"Prototypes ship. Use the direction + tokens now.

Workflow

Step 0 — Derive context + stack

Read docs/product-soul.md, PRD, specs, and existing manifests. Extract audience, emotional goal, brand, owner_mode. Recommend the stack via references/stack-selection.md (default app stack: React + Next + Tailwind v4 + shadcn/ui; match an existing repo's stack). State the recommendation in one line.

Step 1 — Diagnose the ask

SignalPath
Single artifact (one page/component/poster)Fast: direction → system → build → review
Full app / multi-page productFull: Step 0 → direction → system → build → review
Beautify / redesign existing UIRefactor: read existing → fit a direction → system diff → build → review
One isolated step ("just tokens", "just review")Direct: invoke that sub-skill only
For single artifacts, read references/one-shot-flow.md.

Step 2 — Run design-direction

Invoke design-direction. It explores 2-3 distinct directions and commits to one (agent picks for a non-technical owner). Output: .design/<feature>/DIRECTION.md.

Step 3 — Run design-system

Invoke design-system with the direction + chosen stack/token-format. Output: canonical DESIGN.md + tokens.css (state-level, APCA-checked) + icon strategy + component contracts.

Step 4 — Build

Implement using ONLY the DESIGN.md tokens, the component contracts, and the icon strategy. Match references/golden-examples/*. Apply references/build-conventions.md and references/ui-patterns.md (container/presentation, optimistic updates, forms, tables). Mandatory gates before "done":

  • No banned default present (run references/anti-vibecoded-checklist.md)
  • All values via tokens; no hex/slate-*/magic numbers in components
  • Every data surface: loading + empty + error + populated rendered
  • Every interactive el: hover + active + focus-visible + disabled
  • One orchestrated entrance; reduced-motion honored (references/polish-playbook.md)
  • Dark mode rendered (hand-set), tested at 375px
  • ≥1 distinctive move (signature layout/type/color/interaction) — generic = fail

Step 5 — Run design-review

Invoke design-review against the build. On REVISE, loop back to Step 4 with the specific findings. Max 2 loops, then escalate (the direction or brief is the problem).

Step 6 — Deliver

Output the file tree, the running route, and the impact report.


Output Format

## Frontend Design Report
Feature: [name] | Stack: [derived] | Direction: [name] — feels like [ref]
Path: [fast | full | refactor | direct]
Files: DESIGN.md, src/styles/tokens.css, src/... , .design/<feature>/{DIRECTION,REVIEW}.md
Distinctive moves: [list]
State coverage: [loading/empty/error/populated all ✓]
Anti-slop gates: [N/N] | APCA: [pass] | Review loops: [N]

Verification

  • Stack derived from product docs (or matched existing) and stated
  • design-direction explored options; one DESIGN.md is the single source of truth
  • Build matches golden-examples craft; tokens-only; ≥1 distinctive move
  • All states present; reduced-motion honored; 375px + dark mode pass
  • design-review verdict SHIP (or ≤2 loops then escalation noted)

Red Flags

  • UI coded before design-direction commitment exists
  • Components built from memory not golden examples
  • Raw hex or magic numbers bypass DESIGN.md tokens
  • Data surface missing loading empty or error states

Reference Files

  • references/stack-selection.md — derive stack from product docs; shadcn-as-primitives
  • references/polish-playbook.md — state coverage, micro-interactions, motion specifics (Step 4 gate)
  • references/golden-examples/components.md — button/input/card/badge with all states
  • references/golden-examples/states.md — empty / loading / error patterns
  • references/golden-examples/composition.md — non-default hero, app shell, staggered entrance
  • references/build-conventions.md — framework conventions, layout, a11y, file structure
  • references/ui-patterns.md — container/presentation, optimistic updates, forms, tables, error boundaries
  • references/anti-vibecoded-checklist.md — banned defaults + distinctive-moves list
  • references/one-shot-flow.md — compressed flow for single artifacts

File Output

Append to docs/skill-outputs/SKILL-OUTPUTS.md:

| YYYY-MM-DD HH:MM | frontend-design | .design/<feature>/ + DESIGN.md + src/... | [what was built] |

Prune Log

Last pruned: 2026-07-04

  • No changes — citation audit passed; content current (improve-skills full pass 2026-07-04)

Impact Report

Frontend design complete: [feature]
Stack: [derived] | Direction: [name]
Path: [fast | full | refactor | direct]
Sub-skills: design-direction, design-system, design-review
State coverage: [✓] | Distinctive moves: [count] | APCA: [pass] | Review loops: [N]
Files created: [count]

レビュー

まだレビューはありません。使ってみた感想をお寄せください。

同じリポジトリのスキル

概要と使いどころ

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".

日本語の概要は準備中です。原文の説明を表示しています。

dvy1987/agent-loom32026年8月8日 更新

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.

日本語の概要は準備中です。原文の説明を表示しています。

dvy1987/agent-loom32026年8月8日 更新

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.

日本語の概要は準備中です。原文の説明を表示しています。

dvy1987/agent-loom32026年8月8日 更新

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".

日本語の概要は準備中です。原文の説明を表示しています。

dvy1987/agent-loom32026年8月8日 更新

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.

日本語の概要は準備中です。原文の説明を表示しています。

dvy1987/agent-loom32026年8月8日 更新

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).

日本語の概要は準備中です。原文の説明を表示しています。

dvy1987/agent-loom32026年8月8日 更新

dvy1987 のスキルをすべて見る

このスキルの問題を報告する