AI 코딩 세션은 시간이 지나면서 중요한 프로젝트 컨텍스트를 잃습니다. 이 스킬은 결정 사항, 현재 작업, 진행 상황을 구조화된 Markdown 파일에 유지합니다.

계획, 요구사항, 검증 단계가 흩어져 있으면 팀은 작업 맥락을 잃기 쉽습니다. 이 스킬은 저장소 맥락과 템플릿을 바탕으로 일관된 작업 문서를 만듭니다.

프로젝트 지침은 여러 어시스턴트 및 IDE 파일에 흩어져 있는 경우가 많습니다. 이 스킬은 해당 파일을 찾아 공통 규칙을 요약하고 하나로 통합된 AGENTS.md를 제안합니다.

Zig API와 언어 세부 사항은 버전마다 빠르게 변경됩니다. 이 스킬은 에이전트가 Zig 소스, 참조 문서, 가이드에서 최신 문서를 가져오도록 안내합니다.

Renovation work involves sequencing decisions, hidden defects, and unclear contractor responsibilities. This Chinese-language handbook organizes construction references into practical planning advice and inspection checklists.

Canton, DAML, Splice 버전 불일치는 빌드와 패키지 업그레이드를 망가뜨릴 수 있습니다. 이 스킬은 버전 매핑, 파일 위치, 문제 해결 단계를 제공합니다.

Waterproofing decisions can be difficult when contractor quotes, drainage layouts, and inspection requirements are unclear. This skill organizes Chinese renovation references into practical questions, checklists, and troubleshooting guidance.

代码改动后，README 与版本记录常被遗漏，文档容易与实现脱节。本技能读取真实代码与现有文档，按固定清单更新 README 和版本条目，并如实标注验证状态。

Technical plans often drift from product intent or the current codebase. This skill researches repository structure and turns constraints into an actionable TECH.md implementation plan.

Spec-triggered documentation created drafts for features that had not shipped. This retired skill explains the failure and directs users to release-gated alternatives.

Volcengine VoiceChat integrations involve changing APIs, models, audio settings, and runtime stages. This skill provides safe configuration guidance, official documentation routing, and evidence-based troubleshooting.

Ponytail users need a clear reference for modes, commands, and configuration. This skill displays that reference once without changing modes or saving state.

Ambiguous requirements create conflicting implementations and unverifiable reviews. This skill guides linked specifications, risk assessments, and verification strategies through explicit human approval.

Vague architecture leaves teams with incompatible interfaces and undocumented decisions. This skill structures architecture, decisions, contracts, and component designs for human review before specification.

Teams often declare work complete without proving every requirement, path, review, and gap. This skill creates a traceable chain from intent to evidence and a defensible delivery decision.

OpenSpec tasks and archive states do not prove evidence-backed completion. This skill maps requirements, acceptance items, journeys, and gates into an auditable RVTF closure process.

BMAD work can lose traceability between requirements, reviews, tests, and closure decisions. This skill maps BMAD artifacts to RVTF records while preserving host ownership, evidence quality, and review boundaries.

Project stack documentation often becomes incomplete after dependency changes. This skill detects supported technologies and updates .specsfy/STACK.md while preserving human-authored content.

Structural changes can leave project documentation incomplete or contradictory. This skill identifies affected documents and keeps durable decisions aligned with the code.

Stale context files mislead agents and increase maintenance risk. Gardener verifies documentation against code, repairs references, and keeps risk baselines controlled.