Software Development Skills

Homer L3 — Code Archaeology Muse. Maps existing codebases, finds call sites, builds dependency graphs, spots dead code. Zeus dispatches Clio when the plan needs "what's already in the codebase" before any new work. Read-only. Every claim cites file:line.

Create high-quality git commits: review/stage intended changes, split into logical commits, and write clear commit messages (including Conventional Commits).

Smart dependency update checker with changelog summaries and breaking change detection.

Create and maintain persistent design plan files for non-ticket architecture discussions. Auto-triggers when user starts a design discussion; accumulates decisions as file not memory; locks when user approves; implementation reads plan as spec. Fills the gap between refinement/sasd-review (ticket-scoped) and informal conversation (ephemeral). Trigger: '想討論', '怎麼設計', '重構', '重新設計', '要怎麼改', '要怎麼重做', 'design plan', 'ADR'.

Verify plan consistency, generate summaries, maintain plan health. Use for review, verification, summaries, or plan maintenance. Mostly read-only with maintenance operations.

Find 10x product opportunities and high-leverage improvements. Use when user wants strategic product thinking, mentions '10x', wants to find high-impact features, or says 'what would make this 10x better', 'product strategy', or 'what should we build next'.

Generate memes using the memegen.link API. Use when users request memes, want to add humor to content, or need visual aids for social media. Supports 100+ popular templates with custom text and styling.

Convert OpenAPI 3 JSON/YAML schemas into TypeScript interfaces, type guards, and API types.

Create and manage Claude Code plugins with proper structure, manifests, and marketplace integration. Use when creating plugins for a marketplace, adding plugin components (commands, agents, hooks), bumping plugin versions, or working with plugin.json/marketplace.json manifests.

Unified Orchestrator worker entrypoint. Use when the user says "pmworker" and wants a worker to claim tasks, lock files, implement changes, and report completion.

Build type-safe React/TypeScript components, hooks, events, useEffect, React 18-19/Server Components, generic components, and TanStack/React Router integration.

Manual-only skill for minimizing total codebase size. Only activate when explicitly requested by user. Measures success by final code amount, not effort. Bias toward deletion.

Transform learning content (like YouTube transcripts, articles, tutorials) into actionable implementation plans using the Ship-Learn-Next framework. Use when user wants to turn advice, lessons, or educational content into concrete action steps, reps, or a learning quest.

Evaluate Agent Skill design quality against official specifications and best practices. Use when reviewing, auditing, or improving SKILL.md files and skill packages. Provides multi-dimensional scoring and actionable improvement suggestions.

Social Login Implementation Guide. 소셜 로그인(Google, Apple, Kakao, Naver) 구현 가이드. React Native(Expo), Swift, Kotlin, Web 환경. 실전 gotcha + 체크리스트.

Google Stitch MCP - generazione e modifica UI design da prompt, output React/TypeScript/HTML/CSS, estrazione design tokens. Use when generating UI from text prompts via Stitch, converting Stitch HTML/CSS to Tailwind/shadcn, or using Stitch as a Figma alternative.

Only invoke when explicitly requested via "UI审查"、"设计审计"、"@ui-ux-auditor" or "design audit". Do NOT auto-trigger.

Navigate difficult workplace conversations and deliver constructive feedback using Preparation-Delivery-Followup framework and SBI (Situation-Behavior-Impact) model. Use when preparing for tough conversations, giving feedback, managing conflicts, or handling sensitive discussions.

Apply Strunk's timeless writing rules to ANY prose humans will read—documentation, commit messages, error messages, explanations, reports, or UI text.

Homer L2 — Orchestrator skill. Decomposes a task, plans via extended thinking, delegates to MUSES workers in parallel (Anthropic MARS pattern), synthesizes, hands to Oracle for eval, writes checkpoint to VAULT. Use when a task would benefit from multi-agent decomposition (3+ subtasks, research burst, architecture decisions, multi-file work). Do NOT use for trivial single-tool tasks — route those through Brain L1 directly.

Codex CLI로 병렬 구현할 때 사용. multiagent 기능으로 default/explorer/worker를 조합해 실행. zephermine 섹션 모드와 자유 모드 지원. /agent-team, /poseidon, /agent-team-codex로 실행.

준공검사 감리 스킬. 설계 산출물(spec, api-spec, qa-scenarios, flow-diagrams) 대비 구현 검증. 정적 분석 + 런타임 검증 + API 일치 + QA 시나리오 + 도면 대조. /argos로 실행.

끊김 없는 자율 루프. Claude는 Stop 훅, Codex는 notify 자동 재개로 AI가 작업을 끝까지 완수하도록 강제합니다. FIND → FIX → VERIFY 사이클을 반복하며, --max-iterations와 --completion-promise로 제어합니다. /chronos 또는 /loop로 실행. Also known as 크로노스.

사업성 검토 통합 스킬. 수요 검증 + 비즈니스 모델 캔버스, 수익/가격 전략, 시장 분석(TAM/SAM/SOM), GTM 전략, 북극성 지표, 코호트 분석을 한 번에 수행. /hermes로 실행.