artibot

Designs A/B tests with statistical methodology, sample sizing, and significance analysis covering hypothesis formulation, variant design, and result interpretation. Use when user asks about A/B test, split test, experiment design, hypothesis, statistical significance, sample size, multivariate test, AB 테스트, 실험 설계, or 통계적 유의성.

ADR(Architecture Decision Record) 작성 스킬. 두 개 이상의 기술 선택지를 비교하고 추천안과 그 근거, 숨겨진 비용, 확장성, 2년 뒤 기술 부채까지 입체적으로 평가한다. 비개발자도 이해할 수 있도록 한국어 위주 설명 + 추천안 강조. 자동 트리거: "A vs B", "X와 Y 중 뭐가 좋아", "어떤 라이브러리 써야 해", "기술 선택", "스택 결정", "아키텍처 결정", "DB 뭐 쓰지", "프레임워크 고민", "이거 쓸까 저거 쓸까", "should I use", "which is better", "compare options", "tech choice", "library decision", "framework comparison", "architecture decision".

코드를 공격자 관점에서 리뷰하여 취약점, 엣지 케이스, 설계 결함을 탐지. Use when performing security review, attack surface analysis, pre-deploy verification, or PR security check, or mentions 적대적 리뷰, 공격적 리뷰, 취약점 리뷰.

Designs paid advertising campaigns with creative briefs, platform constraints, and bidding strategy optimization for Google Ads, Meta Ads, and LinkedIn Ads. Use when user asks about advertising, ad copy, paid media, Google Ads, Meta Ads, PPC, retargeting, creative brief, 광고, 유료 광고, or 리타겟팅.

Claude advisor tool strategy for cost-optimized quality retention. Sonnet executor + Opus advisor pattern reduces API cost by ~50% while maintaining near-Opus quality on complex judgment calls. Auto-activates when discussing model cost optimization, executor-advisor delegation, or advisor API integration. Triggers: advisor, cost optimization, model strategy, executor-advisor, advisor tool

세션 상태를 압축 스냅샷으로 변환하여 서브에이전트 프롬프트에 주입. Auto-activates when: spawning sub-agents, delegating tasks, transferring session context between agents. Triggers: snapshot, session context, agent context, sub-agent, 스냅샷, 세션 컨텍스트, 서브에이전트, 컨텍스트 주입.

AI/LLM security standards enforcing OWASP LLM Top 10 and Artibot's own threat model. Covers prompt injection (direct + indirect), excessive agency, insecure plugin design, RAG poisoning, and tool-output trust boundaries — framed for Artibot's real attack surface: hooks (arbitrary stdin), MCP servers, and native Agent Teams. Auto-activates when: authoring hooks, MCP tools, agents that read external input, RAG pipelines, or any code that feeds untrusted text into a model context. Triggers: LLM security, prompt injection, jailbreak, RAG, agent security, tool output, MCP, hook input, guardrail, excessive agency, LLM 보안, 프롬프트 인젝션, 에이전트 보안, 훅 보안

Manages brand identity including voice guidelines, visual standards, messaging consistency, and brand governance with frameworks for style guides and cross-channel consistency. Use when user asks about brand guidelines, brand voice, style guide, brand identity, tone of voice, messaging framework, visual identity, 브랜드 가이드라인, 브랜드 보이스, or 스타일 가이드.

Designs campaign architecture, timelines, cross-channel coordination, and budget allocation for marketing campaigns from planning through measurement. Use when user asks about campaign planning, launch coordination, cross-channel marketing, campaign calendar, UTM tracking, budget allocation, 캠페인 기획, 캠페인 런칭, or 멀티채널.

GitHub Actions 심화 패턴 — matrix strategy, reusable workflows, composite actions, 시크릿 관리, 캐싱 전략. Use when designing CI/CD pipelines, GitHub Actions workflows, or deployment automation. 자연어 트리거: 'CI 파이프라인 만들어줘', '깃헙 액션 워크플로 짜줘', '배포 자동화 해줘', 'GitHub Actions 설정'.

Turns ambiguous user requests into precise, actionable specifications using hypothesis-based MCQ clarification. Auto-activates when ambiguity score exceeds threshold, multi-intent conflicts are detected, or the user explicitly requests clarification on scope, target, or constraints. Triggers: clarify, unclear, what do you mean, ambiguous, vague, specify, 명확히, 구체적으로, 뭘 원하시는지, 좀 더 자세히 말해줘, 어떤 걸 원하는 건지 모르겠어, 범위가 너무 넓어

Audit CLAUDE.md files in a repository against a 6-criteria quality rubric, output a scored report and diff preview, and require user approval before any Edit. 자연어 트리거: 'CLAUDE.md 점검해줘', '메모리 품질 감사', '프로젝트 메모리 검토해줘', 'claude.md audit'.

Detects AI-generated code slop patterns (defensive try/catch wrapping, premature abstractions, builtin wrappers, owner-less TODOs, over-configurable params) and scores 0-100 across five tiers. Gating input for code-reviewer agent. Use when user asks about code slop, AI 코드 검토, over-engineering 탐지, defensive code, boilerplate 코드, PR 리뷰, 코드 품질 게이트, slop detection, or LLM 코드 검증.

Provides codex-plugin-cc integration workflow for cross-checking, development delegation, and idea generation with OpenAI Codex. Use when user asks about Codex, cross-check, 크로스체크, OpenAI CLI, dual-AI review, or wants to delegate tasks to Codex.

Coding standards and style guide enforcing immutability, error handling, file organization, and naming conventions. Auto-activates when: writing or modifying code, code review, creating new files or components. Triggers: code, write, edit, implement, component, function, class, style, 코드 작성, 코딩 스타일, 코드 리뷰, 구현, 컴포넌트, 함수

Dual-process cognitive routing engine inspired by Kahneman's System 1/2 theory. Routes user requests through fast intuitive pattern matching (System 1) or deep deliberative reasoning (System 2) based on complexity scoring, confidence thresholds, and adaptive learning. Auto-activates on every user request via the cognitive-router hook. Triggers: route, classify, cognitive, system1, system2, fast, deep, escalate, think, intuition, 라우팅, 분류, 인지, 빠른 사고, 깊은 사고, 에스컬레이션

컴팩션 생존 전략 — 컨텍스트 압축 시 핵심 정보 보존 방법. Auto-activates when: context is about to compact, key files must survive compression, pending work needs to persist. Triggers: compaction, context loss, context compression, 컴팩션, 컨텍스트 손실, 컨텍스트 압축, 컨텍스트 윈도우. See also strategic-compact for compaction timing.

Conducts competitive analysis and market intelligence for positioning, differentiation, and strategic advantage using SWOT analysis and competitor monitoring frameworks. Use when user asks about competitor analysis, competitive intelligence, SWOT, differentiation, market positioning, benchmark, 경쟁 분석, 시장 정보, or 포지셔닝.

Optimizes on-page content for search engines including meta tags, heading structure, content quality, topical authority, and E-E-A-T signals. Use when user asks about content SEO, on-page SEO, meta tags, content optimization, content gap analysis, topical authority, E-E-A-T, 콘텐츠 SEO, 온페이지 최적화, or 콘텐츠 갭.

LLM 컨텍스트 윈도우 실패 모드 분석 - lost-in-middle, poisoning, distraction, confusion, clash 패턴 진단 및 완화. Auto-activates when: long conversation degradation, context-related failures, agent output quality drops. Triggers: context degradation, lost in middle, context poisoning, context window, attention, 컨텍스트 열화

Pattern extraction and knowledge persistence across sessions using auto memory. Auto-activates when: recurring patterns detected, debugging insights gained, user preferences confirmed. Triggers: remember, pattern, learned, always do, never do, across sessions

Applies persuasive writing frameworks (AIDA, PAS, BAB) for headlines, CTAs, ad copy, email subjects, and landing pages with platform-specific character constraints. Use when user asks about copywriting, headline, CTA, ad copy, subject line, landing page copy, persuasive writing, tagline, 카피라이팅, 헤드라인, or 광고 문구.

Optimizes forms for lead generation, signups, and data collection with friction reduction, field optimization, validation UX, and progressive profiling strategies. Use when user asks about form optimization, form fields, lead form, signup form, form completion, progressive profiling, form UX, 폼 최적화, 양식, or 리드 폼.

Analyzes multi-step conversion funnels to identify drop-offs and optimize signup flows, onboarding, checkout processes, and lead qualification pipelines. Use when user asks about funnel analysis, conversion funnel, drop-off analysis, signup flow, checkout optimization, onboarding funnel, funnel optimization, 퍼널, 전환 퍼널, or 이탈 분석.

Optimizes landing page conversion rates covering above-the-fold analysis, value proposition clarity, CTA optimization, and trust signal placement with heuristic evaluation. Use when user asks about landing page optimization, page CRO, conversion rate, above the fold, CTA optimization, trust signals, hero section, 랜딩페이지, 페이지 CRO, or 전환율 최적화.

Maps customer journeys with touchpoint analysis, lifecycle stage management, and experience optimization from awareness through advocacy. Use when user asks about customer journey, journey map, touchpoint analysis, lifecycle management, customer experience, onboarding, retention, CX, 고객 여정, 터치포인트, or 라이프사이클.

End-of-day retrospective dashboard of today's commits, quality, work status, and next steps. Use when the user wants a daily recap or retrospective. Triggers: daily, recap, 회고, 일일 보고, 오늘 작업, 오늘 뭐 했지, retrospective, daily report, 오늘 뭐 했는지 정리해줘, 일일 리뷰 해줘, 오늘 작업 요약해줘

Performs statistical data analysis, data cleaning, and insight extraction covering KPI calculation, trend analysis, cohort analysis, funnel analysis, and attribution modeling. Use when user asks about data analysis, statistics, KPI, metrics, trend analysis, cohort, funnel analysis, attribution, forecast, 데이터 분석, 지표, 통계, 코호트, or 퍼널.

Guides data visualization with chart selection, dashboard layout design, visual encoding, and data storytelling best practices. Use when user asks about data visualization, chart selection, graph design, dashboard layout, infographic, data storytelling, 데이터 시각화, 차트, 대시보드, or 그래프.

DDD 전략적 설계 - 바운디드 컨텍스트, 유비쿼터스 언어, 서브도메인 분류, 팀 소유권 정렬. Auto-activates when: domain boundary definition, bounded context design, ubiquitous language creation. Triggers: DDD strategic, bounded context, ubiquitous language, subdomain, 바운디드 컨텍스트, DDD 전략

DDD 전술적 설계 - Entity, Value Object, Aggregate, Repository, Domain Service, Domain Event 패턴. Auto-activates when: implementing domain models, aggregate design, invariant enforcement. Triggers: DDD tactical, aggregate, value object, entity, domain event, DDD 전술, 도메인 모델

Delegation strategies for parallel and complex multi-file operations using Sub-Agent or Team Mode. Sub-Agent Mode: Task tool for focused, single-session tasks. Team Mode: Agent Teams API for complex multi-domain coordination with peer communication. Auto-activates when: >7 directories, >50 files, multi-domain operations, high complexity tasks, team coordination needed. Triggers: delegate, parallel, sub-agent, team, concurrent, large-scale, orchestrate, coordinate, 위임, 병렬, 대규모, 팀

Provides a structured DESIGN.md schema and curated brand references for AI-native visual design specification. 자연어 트리거: '디자인 시스템 만들어줘', 'DESIGN.md 작성', '디자인 토큰 정의해줘', '비주얼 디자인 가이드', '컬러 팔레트 정해줘'.

Creates email campaigns, automation sequences, and deliverability optimization covering drip campaigns, newsletters, onboarding, and cold outreach with compliance best practices. Use when user asks about email campaign, newsletter, drip sequence, email automation, deliverability, cold outreach, 이메일 마케팅, 이메일 캠페인, 뉴스레터, or 자동화 시퀀스.

Auto-captures user bug reports and feature requests, conducts a structured interview, and creates a GitHub issue on the Artibot repository — all without the user typing any slash-command. Triggered by natural complaint/request language. MUST trigger when user says any of: - "버그", "버그 신고", "이상해", "안 돼", "안 됨", "작동 안 해", "에러 나", "오류 나" - "기능 요청", "이런 기능 있으면", "추가해주면", "~되게 해줘" - "개선해줘", "불편해", "느려", "느린 것 같아" - "bug", "broken", "doesn't work", "feature request", "please add" DO NOT trigger for: - Casual frustration about unrelated things ("오늘 날씨 이상해") - Questions about how something works ("이거 어떻게 써?") - Rhetorical complaints inside code ("이 코드 이상하네" during review) The user never needs to type `/feedback` — Claude detects intent and starts the flow.

명령형 코드를 함수형 패턴으로 전환하는 마이그레이션 가이드. Auto-activates when: fp-ts 도입, 함수형 에러 핸들링 전환, 또는 Either/Option/Task 패턴 적용 요청 시. Triggers: fp refactor, Either, Option, TaskEither, fp-ts, functional programming, 함수형 리팩토링, 함수형 프로그래밍, try-catch 없애고 싶어, fp-ts 써보고 싶어, 함수형으로 바꿔줘

Input/output and per-tool guardrails with tripwire semantics. Auto-activates when: an agent processes untrusted input, calls a sensitive tool, or must short-circuit on a policy violation. Triggers: guardrail, input validation, tripwire, tool guardrail, policy check, refusal, agent safety.

로컬 관측성(observability) 스킬 — Claude Code hook 이벤트를 JSONL 큐로 fan-out하고 localhost WebSocket 브로드캐스트로 대시보드에 노출. 외부 전송 0건, Artibot 내부 전용. Auto-activates when: building a local hook monitor, an on-device agent dashboard, or an observability surface. Triggers: hook monitor, agent dashboard, event tap, local telemetry, claude code hooks, 훅 모니터, 이벤트 대시보드, 관측성.

Hook feedback to tool result merge pattern. Documents how pre/post hook feedback (warnings, blocks, suggestions) gets merged into tool results so the model sees quality signals inline with tool output. Auto-activates when: writing hooks that produce feedback, reviewing hook output flow. Triggers: hook feedback, tool result, merge, pre-write, post-edit, quality gate

AI image prompt engineering and generation workflow. Use when user asks to generate, create, edit, or refine images. Triggers on: 이미지 만들어줘, 그림 생성해줘, 이미지 그려줘, 썸네일 만들어, 로고 만들어줘, 이미지 수정해줘, 사진 편집해줘, 아까 만든 거 수정, 색감 바꿔줘, image generate, create image, edit image, modify image, refine image.

Manages lead lifecycle including lead scoring, qualification, nurture sequences, CRM workflows, MQL/SQL handoff, and pipeline optimization. Use when user asks about lead management, lead scoring, lead qualification, MQL, SQL, lead nurture, pipeline, CRM workflow, 리드 관리, 리드 스코어링, or 리드 육성.

Provides Mermaid diagramming patterns for technical documentation including flowcharts, sequence diagrams, ERDs, class diagrams, state diagrams, Gantt charts, and mindmaps. Use when creating architecture diagrams or documenting API sequences. 자연어 트리거: '다이어그램 그려줘', '머메이드로 그려줘', '시퀀스 다이어그램 만들어', '플로우차트 그려줘'.

Provides shadcn/ui component patterns for React including installation, theming, dark mode, CVA variants, and accessibility with Radix UI primitives. Use when the user asks about building UIs with shadcn/ui, setting up component libraries, customizing themes or dark mode, or composing form patterns. Triggers: shadcn, shadcn/ui, radix, tailwind components, ui library, component library, dark mode.

Continuous learning pipeline that captures session experiences, performs batch learning via GRPO, and transfers validated knowledge between System 1 and System 2 caches. Auto-activates when: session end, pattern discovery during routing, knowledge transfer triggers, manual /learn command. Triggers: learn, experience, knowledge, transfer, promote, demote, grpo, batch, pattern, skill development, growth, continuous improvement

k6 기반 성능·부하 테스트 가이드 — 시나리오 설계, 임계값 설정, CI/CD 통합. Use when designing load tests, stress tests, or performance benchmarks. 자연어 트리거: '부하 테스트 해줘', '성능 테스트 만들어줘', 'k6 스크립트 작성', '스트레스 테스트'.

Analyzes marketing performance with KPI dashboards, channel attribution, campaign ROI measurement, forecasting, and industry benchmarking. Use when user asks about marketing analytics, KPI dashboard, attribution, ROAS, CAC, LTV, channel performance, campaign ROI, 마케팅 분석, KPI 대시보드, 어트리뷰션, or ROAS.

Provides marketing strategy frameworks for market analysis, positioning, segmentation, and go-to-market planning using TAM/SAM/SOM and Porter's Five Forces. Use when user asks about marketing strategy, market analysis, go-to-market, GTM, positioning, growth strategy, value proposition, channel strategy, 마케팅 전략, 시장 분석, or GTM 전략.

Provides Context7 MCP server workflow for looking up library documentation, framework patterns, and API references with caching and fallback strategies. Use when user asks about documentation lookup, library usage, framework patterns, import questions, package API, Context7, 라이브러리, or 프레임워크.

Coordinates MCP server selection and multi-server orchestration with parallel/sequential modes and fallback strategies for complex multi-domain tasks. Use when user asks about MCP coordination, server selection, multi-server orchestration, tool coordination, fallback strategy, or MCP 조율.

Provides Playwright MCP server workflow for E2E testing, visual validation, cross-browser automation, and performance measurement with error recovery strategies. Use when user asks about Playwright, E2E testing, browser automation, screenshot testing, visual regression, cross-browser testing, E2E 테스트, or 브라우저.