my-cc-harness

현재 브랜치의 변경사항 코드 리뷰. Codex + Claude 듀얼 리뷰. 트리거: "리뷰", "review", "검토", "코드 리뷰", "PR 리뷰", "변경사항 확인" 안티-트리거: "구현", "코드 작성", "빌드"

코드 단순화 및 리팩토링 — 변경된 코드를 리뷰하고 불필요한 추상화, 중복, 복잡성을 제거합니다. Triggers on: 단순화, simplify, 리팩토링, 코드 정리, 코드 개선. NOT for: 새 기능 추가, 버그 수정.

작업 완료 후 코드 검증 (타입체크, 린트, 테스트, 빌드). Triggers on: 검증, verify, 테스트 돌려, 빌드 확인, 타입체크. NOT for: E2E 테스트, 코드 작성, 구현.

Ralph Loop 병렬 버전. 여러 에이전트가 worktree에서 동시에 PRD 항목을 처리한다. 트리거: "병렬 실험", "autodev parallel", "동시에 실험", "워크트리 실험", "병렬 랄프" 안티-트리거: "순차 실험", "하나씩"

Ralph Loop 기반 자율 개발 루프. Stop Hook이 세션 종료를 가로채어 PRD 항목을 하나씩 완료하며 자동 커밋한다. 트리거: "autodev", "자율 개발", "밤새 돌려", "랄프 루프", "ralph loop", "자동 개발" 안티-트리거: "직접 구현해", "한번만 해", "수동"

변경사항 커밋, 푸시, PR 생성을 한 번에 수행하는 워크플로우. 트리거: "커밋", "PR 만들어", "푸시해", "PR 생성", "commit and push", "커밋하고 푸시", "PR 올려" 안티-트리거: "코드 리뷰", "git log 확인", "diff 보여줘", "브랜치 목록"

개발 완료 후 피처 기반 E2E 테스트 작성 및 실행. /verify 이후 실제 사용자 플로우를 검증합니다. Triggers on: e2e 검증, e2e-verify, E2E 테스트, 브라우저 테스트 실행. NOT for: 유닛 테스트, 타입체크, 빌드 검증.

코드 산출물을 4축(기능/품질/독창성/보안)으로 평가하고 점수 산출. Evaluator 에이전트를 스폰하여 독립 평가 실행. Triggers on: eval, 평가, 품질 점수, 코드 평가, quality score. NOT for: 코드 작성, 구현, 리뷰.

하네스(hooks, skills, agents, rules) 전체 건강도를 진단하고 점수 산출. Triggers on: harness audit, 하네스 진단, 설정 점검, 하네스 점검. NOT for: 코드 작성, 구현.

REQUIRED for all image generation requests. Generate and edit images using Gemini CLI with image models. Handles blog featured images, YouTube thumbnails, icons, diagrams, patterns, illustrations, photos, visual assets, graphics, artwork, pictures. Use this skill whenever the user asks to create, generate, make, draw, design, or edit any image or visual content.

복잡한 작업 전 계획 수립. Plan 모드에서 사용하거나 자동 활성화. 트리거: "계획", "플랜", "어떻게 구현", "설계", "아키텍처 결정" 안티-트리거: "바로 구현해", "간단한 수정", "마이크로 수정"

테스트 주도 개발 - 테스트 먼저 작성 후 구현. 트리거: "TDD", "테스트 먼저", "테스트 주도", "RED-GREEN-REFACTOR". 안티-트리거: "구현 먼저", "테스트 나중에".

UI/UX design intelligence. 50 styles, 21 palettes, 50 font pairings, 20 charts, 9 stacks (React, Next.js, Vue, Svelte, SwiftUI, React Native, Flutter, Tailwind, shadcn/ui). Actions: plan, build, create, design, implement, review, fix, improve, optimize, enhance, refactor, check UI/UX code. Projects: website, landing page, dashboard, admin panel, e-commerce, SaaS, portfolio, blog, mobile app, .html, .tsx, .vue, .svelte. Elements: button, modal, navbar, sidebar, card, table, form, chart. Styles: glassmorphism, claymorphism, minimalism, brutalism, neumorphism, bento grid, dark mode, responsive, skeuomorphism, flat design. Topics: color palette, accessibility, animation, layout, typography, font pairing, spacing, hover, shadow, gradient. Integrations: shadcn/ui MCP for component search and examples.

React and Next.js performance optimization guidelines from Vercel Engineering. This skill should be used when writing, reviewing, or refactoring React/Next.js code to ensure optimal performance patterns. Triggers on tasks involving React components, Next.js pages, data fetching, bundle optimization, or performance improvements. Anti-triggers: "Vue", "Angular", "백엔드 최적화".

REST and GraphQL API design principles guide. Use when designing new APIs, reviewing API specs, or establishing API design standards. Triggers: "API design", "REST", "GraphQL", "endpoint design", "API versioning", "API patterns" Anti-triggers: "frontend UI", "database schema only", "CSS/styling", "deployment/infra"

Python asyncio and async/await patterns guide. Use when building async APIs, concurrency systems, or I/O-bound applications. Triggers: "asyncio", "async/await", "async Python", "concurrent", "aiohttp", "async patterns" Anti-triggers: "synchronous Python", "JavaScript async", "Go goroutine", "multiprocessing only"

AutoDeveloper orchestrator that runs multiple experiments in parallel using worktrees. Triggers: "parallel experiments", "autodev parallel", "simultaneous experiments", "worktree experiments" Anti-triggers: "sequential experiments", "one at a time"

Autonomous code experimentation loop. An AI agent modifies code, validates, and iteratively keeps or discards changes for autonomous improvement. Applies Karpathy's autoresearch pattern to general software development. Triggers: "autodev", "autonomous development", "autonomous experimentation", "run overnight", "experiment loop", "auto optimization" Anti-triggers: "implement it yourself", "do it once", "manual"

Workflow for incrementally fixing TypeScript and build errors. Triggers: "build error", "build error", "compile error", "tsc error", "type error", "build failure", "build fix" Anti-triggers: "runtime error", "test failure", "lint errors only", "deployment error"

Workflow for committing changes, pushing, and creating PRs in one step. Triggers: "commit", "create PR", "push", "PR creation", "commit and push", "commit and push", "submit PR" Anti-triggers: "code review", "check git log", "show diff", "branch list"

Context window management and token optimization guide. Triggers on: context, tokens, compact, memory management. NOT for: code writing, debugging.

E2E test automation using agent-browser CLI. Performs stable browser tests with snapshot-based accessibility trees and the ref system. Triggers: "E2E", "browser testing", "agent-browser", "write e2e tests", "E2E testing", "browser automation" Anti-triggers: "unit tests", "API tests", "pytest", "Jest tests only", "use Playwright directly"

Feature-based E2E test writing and execution after development. Verifies actual user flows after /verify. Triggers on: e2e verification, e2e-verify, E2E testing, browser test execution. NOT for: unit tests, type checking, build verification.

Production-grade FastAPI project creation and setup guide. Includes async patterns, DI, and error handling. Triggers: "FastAPI", "Python API project", "FastAPI project creation", "FastAPI template", "FastAPI setup" Anti-triggers: "Django", "Flask", "Node.js API", "Express", "modifying existing FastAPI only"

Big-tech style (Stripe, Vercel, Apple) UI development. From planning to implementation in one go. Triggers: "frontend", "create UI", "create page", "create component", "landing page", "dashboard UI" Anti-triggers: "backend API", "database", "CLI tools"

Generate HANDOFF.md work handoff document before ending session. Triggers on: handoff, handoff, session cleanup, HANDOFF.md, pass work. NOT for: code writing, review.

Self-diagnostics and improvement suggestions based on 12-principle agent harness. Triggers on: harness diagnostics, audit, environment check, setup, maintenance, drift check. NOT for: code implementation, bug fixes, writing tests.

REQUIRED for all image generation requests. Generate and edit images using Nano Banana 2 (Gemini 3.1 Flash Image). Handles blog featured images, YouTube thumbnails, icons, diagrams, patterns, illustrations, photos, visual assets, graphics, artwork, pictures. Use this skill whenever the user asks to create, generate, make, draw, design, or edit any image or visual content.

Plan before tackling complex tasks. Used in Plan mode or auto-activated. Triggers: "plan", "planning", "how to implement", "design", "architecture decision" Anti-triggers: "just implement it", "simple fix"

Implement comprehensive testing strategies with pytest, fixtures, mocking, and test-driven development. Triggers: "pytest", "Python testing", "fixture", "mocking", "writing tests" Anti-triggers: "JavaScript tests", "Jest", "E2E tests"

Comprehensive React 19 patterns expert covering Server Components, Actions, use() hook, useOptimistic, useFormStatus, useFormState, React Compiler, concurrent features, Suspense, and modern TypeScript development. Triggers: "React", "components", "hooks", "JSX/TSX", "state management" Anti-triggers: "Vue", "Svelte", "Angular", "backend"

Code review of changes on the current branch. Triggers: "review", "review", "inspection", "code review", "PR review", "check changes" Anti-triggers: "implementation", "code writing", "build"

Complete shadcn/ui component library guide including installation, configuration, and implementation of accessible React components. Use when setting up shadcn/ui, installing components, building forms with React Hook Form and Zod, customizing themes with Tailwind CSS, or implementing UI patterns like buttons, dialogs, dropdowns, tables, and complex form layouts. Anti-triggers: "Material UI", "Chakra UI", "pure CSS components"

Code simplification and refactoring -- reviews changed code and removes unnecessary abstractions, duplication, and complexity. Triggers on: simplify, simplify, refactoring, code cleanup, code improvement. NOT for: adding new features, bug fixes.

SPEC-based development interview -- generates a detailed SPEC.md specification through in-depth questioning. Triggers on: spec, spec, specification, interview, feature design. NOT for: immediate implementation, simple fixes, code writing.

Verifies implementation completeness against SPEC.md specification. Triggers on: spec verify, spec-verify, specification verification, implementation verification. NOT for: code review, unit tests, build verification.

Analyzes a Stitch project to generate a DESIGN.md file -- documents the design system in semantic, natural language to support consistent UI generation. Triggers on: Stitch, design system, DESIGN.md, Stitch design analysis. NOT for: React component conversion, code implementation.

Transforms vague UI ideas into detailed prompts optimized for Stitch -- enhances specificity, adds UI/UX keywords, injects design system context. Triggers on: prompt enhancement, Stitch prompt, UI idea, prompt improvement. NOT for: direct coding, React implementation.

Iterative build loop pattern for autonomously generating multi-page websites using Stitch -- continuous website development via baton system. Triggers on: Stitch loop, website generation, multi-page, build loop. NOT for: single page, React apps.

Converts Stitch screens into a React component system -- includes design token extraction, component decomposition, TypeScript type generation, and automated verification. Triggers on: Stitch React, component conversion, React conversion, HTML to React. NOT for: creating new React apps, API implementation.