execute
25Skill Claude CodeCodex
Use when the user wants to run an implementation plan. Triggers on "실행해줘", "구현해줘", "시작해줘", "execute", or after plan approval.
Skill Claude CodeCodex
Use when the user wants to run an implementation plan. Triggers on "실행해줘", "구현해줘", "시작해줘", "execute", or after plan approval.
Skill Claude CodeCodex
Use for git operations — commit, PR, branch management. Triggers on "커밋", "commit", "PR 만들어", "브랜치".
Skill Claude CodeCodex
Implement an entire Figma page as Android Jetpack Compose screens by auto-splitting into frames with visual verification. Use whenever a user shares a figma.com/design URL and wants the page built as Compose code. Triggers on '안드로이드 피그마', 'Android Figma', 'Compose 구현'. Covers any multi-frame page. Does NOT apply to…
Skill Claude CodeCodex
Implement an entire Figma page by auto-splitting into frames with visual verification. Use whenever a user shares a figma.com/design URL and wants the page built in code. Triggers on '피그마 구현', '피그마 페이지', 'Figma implement'. Covers any multi-frame page: landing, dashboard, marketing, etc. Does NOT apply to Figma asset…
Skill Claude CodeCodex
Use to convert an Obsidian note into a GitHub Issue. Triggers on "노트로 이슈 만들어", "이슈로 변환", "이거 이슈로", "노트 이슈화". Only use when user explicitly references an Obsidian note as the source. For creating issues from conversation context (without a note), use create-issue instead.
Skill Claude CodeCodex
Use to dispatch multiple independent TODOs concurrently. Invoked by execute skill when independent TODOs are detected.
Skill Claude CodeCodex
Use when a plan needs validation through a planner-critic loop. Triggers on "플랜 리뷰", "plan review", or when explicitly requested.
Skill Claude CodeCodex
Use when the user needs a structured implementation plan with concrete TODOs. Triggers on "플랜", "계획", "plan", or after brainstorming completes.
Skill Claude CodeCodex
Use to analyze plugin usage patterns and suggest improvements. Triggers on "플러그인 리뷰", "plugin review", "사용 패턴 분석".
Skill Claude CodeCodex
Persistent execution loop that continues until all work is verified complete. Auto-linked by execute skill. Triggers on "끝까지 해줘", "멈추지 마", "ralph".
Skill Claude CodeCodex
Use to generate a weekly retrospective by analyzing git log. Triggers on "회고", "retro", "이번 주 정리". Only runs on explicit user request — never auto-triggered.
Skill Claude CodeCodex
Use to save implementation summary to Obsidian vault. Auto-triggered after execute completes. Triggers on "구현 저장", "작업 정리", "save impl".
Skill Claude CodeCodex
A note-saving skill for putting design, research, meeting, or memo notes into an Obsidian vault. Obsidian is a note-taking app that stores linked notes in a local folder.
Skill Claude CodeCodex
Fallback for when the plan skill's auto-save fails or is skipped. Use to manually save a plan to Obsidian vault. Triggers on "플랜 저장", "save plan".
Skill Claude CodeCodex
Use to perform a security audit of the codebase. Triggers on "보안 점검", "security audit". Delegates to architect agent.
Skill Claude CodeCodex
Use for first-time plugin configuration. Triggers on "setup", "설정", "초기 설정".
Skill Claude CodeCodex
Use when implementing with test-driven development. Triggers on "TDD", "테스트 먼저", "test first", or when the plan specifies TDD.
Skill Claude CodeCodex
Loaded automatically at session start. Teaches the orchestrator how to route requests to skills and agents. Do not invoke manually.
Skill Claude CodeCodex
A search workflow for finding notes in an Obsidian vault, which is a personal collection of linked Markdown notes.
Skill Claude CodeCodex
Use to verify a claim of completion with evidence. Auto-triggered after each TODO in execute. Triggers on "검증", "verify", "확인해줘".
Skill Claude CodeCodex
Use to audit and repair the llm-wiki context spine by detecting stale, low-confidence, contested, or broken entries. Triggers on "위키 점검", "wiki maintain", "wiki audit", "컨텍스트 정리", "wiki 오염", "모순 감지".
Skill Claude CodeCodex
Use to record a significant learning, decision, or issue to Obsidian vault. Triggers on "기억해", "이거 기록해", "wisdom", or when a significant learning occurs during debugging/verification.