npx claudepluginhub aimskr/aims-claude-toolkit --plugin aims-toolkitWant just this skill?
Then install: npx claudepluginhub u/[userId]/[slug]
기능 개발, 기능 구현, 신규 개발, 피처 개발, 개발 워크플로우 - Guided feature development with codebase understanding and architecture focus. Systematic 7-phase approach from discovery to implementation. Use when building new features or adding significant functionality. Do NOT use for simple bug fixes, config changes, or single-file edits.
This skill uses the workspace's default tool permissions.
references/karpathy-guidelines.mdreferences/phase-details.mdFeature Development Workflow
Systematic approach: understand codebase → clarify requirements → design → implement.
Core Principles
- Ask clarifying questions: Identify all ambiguities before implementation
- Understand before acting: Read existing code patterns first
- Library-First: Search for existing solutions before custom code
- Simple and elegant: Prioritize readable, maintainable code
- Use TodoWrite: Track all progress
Execution Mode
New feature (기본): Phase 1부터 전체 실행 Modify feature: 기존 기능 수정 시. Phase 2(Exploration)부터 시작. 기존 스펙 문서를 먼저 읽고, 새 문서를 만들지 않고 Changelog를 추가한다.
Modify 모드 진입 조건: 사용자가 기존 기능의 변경/확장을 요청하고,
docs/plans/에 해당 기능의 스펙 문서가 이미 존재하는 경우.
Phase Overview
Phase 1: Discovery → What are we building? (Modify 모드: 스킵 — 기존 스펙 문서 참조)
Phase 2: Exploration → How does existing code work?
Phase 3: Questions → What's unclear? (DON'T SKIP!)
Phase 4: Architecture → How should we build it?
Phase 5: Implementation → Build it (after approval)
Phase 6: Review → Is it good?
Phase 7: Documentation → Update spec doc with Changelog (DON'T SKIP!)
For detailed steps in each phase, see references/phase-details.md.
Key Phase Summaries
Phase 1-2: Discovery & Exploration
Create todo list → Launch parallel code-explorer agents → Build deep understanding
Phase 3: Clarifying Questions ⚠️ CRITICAL
Identify ALL underspecified aspects → Present questions → Wait for answers
Phase 4: Architecture Design
Launch parallel architect agents (Minimal / Clean / Pragmatic) → Present trade-offs → Ask user preference
Phase 4.5: Implementation Gate (Devil's Advocate - 경량)
구현 돌입 전 30초 체크 — 아래 2가지를 사용자에게 간결하게 공유:
- "이 설계에서 가장 먼저 깨질 부분은 어디인가?" (1문장)
- "테스트하기 가장 어려운 부분은?" (1문장)
블로커가 아닌 인지 도구. 1-2문장으로 답하고 바로 Phase 5로 진행. 심각한 문제가 발견되면 Phase 4로 돌아가서 설계 수정.
Phase 5: Implementation
⛔ DO NOT START WITHOUT USER APPROVAL Library-first → Follow codebase conventions → Update todos
Phase 6-7: Review & Documentation
Launch reviewer agents → Address findings →
- New feature:
docs/plans/NNNN.YYYY-MM-DD-<feature>-summary.md생성 (NNNN: 해당 폴더 내 최대 번호 + 1, 없으면 0001) - Modify feature: 기존 스펙 문서에 Changelog 항목 추가 (새 문서 생성 금지)
Changelog 포맷:
## Changelog
### vN — YYYY-MM-DD
**변경**: 무엇을 변경했는지 (1줄)
**이유**: 왜 변경했는지 (1줄)
**영향 범위**: 변경된 모듈/파일 목록
**관련 커밋**: conventional commit 메시지
Implementation Guidelines
Follow Karpathy Guidelines for all code changes. See references/karpathy-guidelines.md.
Key points:
- Think before coding, state assumptions explicitly
- Implement only what's requested (YAGNI)
- Surgical changes — don't "improve" adjacent code
- Verifiable goals for every change
문서화 (Phase 7 완료 후 자동 실행)
작업 완료 시 auto-documenter를 호출하여 프로젝트 문서를 업데이트한다.
Completion
- New feature: Phase 7 완료: 기능 구현 + 리뷰 통과 +
docs/plans/NNNN.YYYY-MM-DD-<feature>-summary.md생성. - Modify feature: Phase 7 완료: 기능 수정 + 리뷰 통과 + 기존 스펙 문서에 Changelog 항목 추가.
Anti-Patterns to Avoid
- NIH (Not Invented Here): Rebuilding what already exists
- Premature Optimization: Unnecessary optimization
- Over-Engineering: Excessive abstraction for simple tasks
- Skipping Questions: Making assumptions instead of asking
Troubleshooting
User says "just do it" when asked clarifying questions: Provide your best recommendation with stated assumptions. Document assumptions in the summary so they can be corrected later. Codebase has no clear patterns to follow: Fall back to the project's language/framework conventions. Document the new pattern as a precedent. Phase 5 implementation diverges from Phase 4 design: Stop, compare, and ask user whether to adjust design or revert implementation. Never silently deviate from approved architecture.
Similar Skills
Activates when the user asks about AI prompts, needs prompt templates, wants to search for prompts, or mentions prompts.chat. Use for discovering, retrieving, and improving prompts.
Search, retrieve, and install Agent Skills from the prompts.chat registry using MCP tools. Use when the user asks to find skills, browse skill catalogs, install a skill for Claude, or extend Claude's capabilities with reusable AI agent components.
Creating algorithmic art using p5.js with seeded randomness and interactive parameter exploration. Use this when users request creating art using code, generative art, algorithmic art, flow fields, or particle systems. Create original algorithmic art rather than copying existing artists' work to avoid copyright violations.