wogus-plugin

This skill should be used when the user asks to "run ralph", "랄프 실행", "랄프 돌려", "ralph loop 실행", "ralph로 해줘", "랄프로 돌려", "run-ralph", "choo choo", "랄프 출발", or wants to execute ANY iterative work — code changes, design documents, skill/workflow integrations, architectural decisions, documentation, or refactoring — through a multi-agent team. Choo-choo is a general iterative framework, not a code-only tool: anything expressible as "work + acceptance criteria + per-iteration verification" fits. Transforms the request into a Ralph Loop prompt with a multi-agent team (mandatory Reviewer + QA), 3-level acceptance criteria, and an iteration workflow that gates completion-promise emission on independent verdicts. Anchors all sentinel/prompt files to the project root so a Worker that `cd`s into a sub-directory mid-loop does not break the gate. Then invokes /ralph-loop.

arkraft-wiki repo에 지식 문서를 생성/업데이트하는 thin wrapper skill. "지식화하자", "지식화", "wiki에 추가해줘", "wiki 문서로 남겨줘", "wikify", "wiki 정리", "wiki에 정리" 같은 요청에 사용. 사용자가 wiki_root를 settings로 지정하면 그 경로로 작업. 컨텍스트 수집 + 섹션 결정 + Acceptance Criteria 도출은 직접 처리하지 않고 run-ralph:choo-choo에 위임 — 이 skill은 wiki-specific 컨텍스트(섹션 구조, lifecycle, harness 검사 항목, scan 대상 repo 목록)를 묶어 choo-choo에 깨끗한 task 명세로 전달하는 것이 전부.

Execute approved implementation plans with TaskList tracking, test verification, and success criteria validation. Use when you have an approved *_PLAN.md file and need step-by-step implementation with comprehensive tracking. Outputs implemented code, test results, and execution summary. Documentation handled by record skill. Korean triggers: 계획 실행, 플랜 실행, 구현 시작, 개발 시작, 코딩 시작, 작업 시작, 실행해줘, 구현해줘, 만들어줘, 개발해줘, 코드 작성, 기능 구현, 태스크 실행.

Create high-quality implementation plans through iterative refinement until all quality standards are met. Use when creating implementation plans from analysis reports or requirements, especially for complex features or critical bug fixes. Generates [FEATURE]_PLAN.md with task breakdown, dependencies, and success criteria. Korean triggers: 구현 계획, 실행 계획, 개발 계획, 플랜 작성, 계획 수립, 작업 계획, 태스크 분해, 계획 세워줘, 플랜 만들어줘, 어떻게 구현할지, 작업 분해해줘.

Independent acceptance verification skill. Run after execute completes to validate the implementation against the original REPORT's reproduction scenarios and the PLAN's success criteria — using actual environment verification (test runs, API calls, UI checks via agent-browser, DB state inspection), not just self-reported "tests pass". Generates [ISSUE_ID]_QA.md with PASS/FAIL verdict and evidence. Korean triggers: QA 검증, 품질 검증, qa 실행, 인수 검증, 시나리오 검증, qa 돌려, 검증해줘, 인수 테스트, 수용 검증.

Consolidate workflow artifacts (analysis reports, plans, implementation results) into comprehensive project documentation. Updates README, CHANGELOG, CLAUDE docs and stores technical insights in Serena memory. Use after completing implementation to finalize and document completed work with optional git commit/push. Korean triggers: 문서화, 문서 작성, 문서 업데이트, README 작성, CHANGELOG 작성, 변경사항 기록, 릴리즈 노트, 정리해줘, 문서 정리, 커밋해줘, 푸시해줘, 마무리해줘, 완료 처리.

Ask YouTube's built-in Gemini AI about video content using CDP automation. Use when user provides a YouTube URL and wants to ask questions about the video. Supports multi-turn questioning: open the panel once, ask multiple questions. Triggers on: "YouTube 영상에 대해 질문해줘" + URL, "이 유튜브 영상 요약해줘" + URL, "/ask-yt [URL] [question]", User wants to query YouTube's Ask/질문하기 built-in AI feature

Systematically analyze the root cause of bugs and issues using multi-perspective investigation with First Principles Thinking. Use when analyzing JIRA issues, Sentry errors, or investigating bug reports. Generates [ISSUE_ID]_REPORT.md with root cause analysis, code locations, reproduction steps, and fix recommendations.

Fix GitLab MR discussions by modifying code, posting a reply explaining the fix, and resolving the discussion. Supports single or batch processing. Korean triggers: discussion 수정, 디스커션 해결, 코멘트 반영, 리뷰 반영, discussion resolve, 피드백 반영.

Perform context-aware comprehensive code review on GitLab MR changes. Validates architecture consistency, business logic accuracy, historical issue patterns, JIRA requirements, security, and test coverage. Generates INLINE_DISCUSSION.json for GitLab inline comments and SUMMARY_COMMENT.md for overall summary. Korean triggers: MR 리뷰, 코드 리뷰, 머지 리퀘스트, PR 리뷰, 코드 검토, 리뷰해줘, 코드 봐줘, MR 확인해줘.