HETS-aware planning for multi-file substantive work. Wraps the planner agent with route-decide gating (Step 0) and architect-spawn recommendation (Step 3) when convergencevalue ≥ 0.10. Invoke via /build-plan or implicitly when plan mode is entered for substantive multi-file work. Additive to /plan — does not replace it.
EnglishPhase-level verification gate run at a v3.x phase boundary (the post-phase analog of verify-plan). Spawns three independent full-context lenses in parallel — PM (honesty-auditor) + Principal-SDE (code-reviewer at phase altitude) + Architect — to review the INTEGRATED phase against its exit criteria, catching cross-PR drift the per-PR VALIDATE structurally cannot see. Produces a CLOSEABLE / NEEDS-WORK verdict + a phase-close record, and feeds the ghost-protocol effectiveness loop. Replaces ad-hoc external reviews (which lack context and hallucinate).
EnglishTransform vague user prompts into structured 4-part prompts (task / scope / constraints / output). Reduces miscommunication and produces better outcomes by enforcing explicit framing before execution.
EnglishAnti-hallucination protocol for tasks where accuracy matters. Enforces evidence-first reasoning, source-cascade workflow, and explicit "I don't know" over speculation. Toggle on when fabricated facts cause real harm.
EnglishOrchestrator-side entry point for "build me X"-style tasks. Translates a user's task description into a concrete spawn plan — stack + required skills + suggested personas + missing-skill bootstrap prompts. Always pauses for user redirect before spawning the team.
EnglishVerify implementation plans before execution using a 4-agent pipeline. Use when you have a written plan (from /bdk:create-plan or manual) and want to check if it will actually work before writing code.
English에이전트 팀 구성 및 조율 스킬. 여러 Claude Code 인스턴스를 팀으로 구성하여 병렬 작업을 수행한다. 공유 태스크 목록, 에이전트 간 메시징, 파일 소유권 분리, 계획 승인 워크플로우를 지원한다. 사용 시점: (1) "팀 구성", "에이전트 팀", "agent team" 키워드 감지 시, (2) 병렬 코드 리뷰, 멀티 레이어 기능 구현, 경쟁 가설 디버깅 등 독립적 병렬 작업이 필요할 때, (3) 3개 이상의 독립적 작업을 동시에 수행해야 할 때, (4) 팀원들이 서로 토론하고 결과를 공유해야 할 때. subagent와 달리, 팀원 간 직접 통신과 공유 태스크 목록을 통한 자체 조율이 가능하다.
Korean
AGENT SKILLS ATLASIncludes Build Plan, Phase Close, Prompt Enrichment, and other workflows.
Claude Code
Codex
Cursor
GitHub Copilot
Gemini CLISupported by the Atlas install prompt. Skill-specific requirements vary.
Pack stats
What’s included
Build Plan
HETS-aware planning for multi-file substantive work. Wraps the planner agent with route-decide gating (Step 0) and architect-spawn recommendation (Step 3) when convergencevalue ≥ 0.10. Invoke via /build-plan or implicitly when plan mode is entered for substantive multi-file work. Additive to /plan — does not replace it.
Phase Close
Phase-level verification gate run at a v3.x phase boundary (the post-phase analog of verify-plan). Spawns three independent full-context lenses in parallel — PM (honesty-auditor) + Principal-SDE (code-reviewer at phase altitude) + Architect — to review the INTEGRATED phase against its exit criteria, catching cross-PR drift the per-PR VALIDATE structurally cannot see. Produces a CLOSEABLE / NEEDS-WORK verdict + a phase-close record, and feeds the ghost-protocol effectiveness loop. Replaces ad-hoc external reviews (which lack context and hallucinate).
Prompt Enrichment
Transform vague user prompts into structured 4-part prompts (task / scope / constraints / output). Reduces miscommunication and produces better outcomes by enforcing explicit framing before execution.
Research Mode
Anti-hallucination protocol for tasks where accuracy matters. Enforces evidence-first reasoning, source-cascade workflow, and explicit "I don't know" over speculation. Toggle on when fabricated facts cause real harm.
Tech Stack Analyzer
Orchestrator-side entry point for "build me X"-style tasks. Translates a user's task description into a concrete spawn plan — stack + required skills + suggested personas + missing-skill bootstrap prompts. Always pauses for user redirect before spawning the team.
Verify Plan
Verify implementation plans before execution using a 4-agent pipeline. Use when you have a written plan (from /bdk:create-plan or manual) and want to check if it will actually work before writing code.
Agent Team
에이전트 팀 구성 및 조율 스킬. 여러 Claude Code 인스턴스를 팀으로 구성하여 병렬 작업을 수행한다. 공유 태스크 목록, 에이전트 간 메시징, 파일 소유권 분리, 계획 승인 워크플로우를 지원한다. 사용 시점: (1) "팀 구성", "에이전트 팀", "agent team" 키워드 감지 시, (2) 병렬 코드 리뷰, 멀티 레이어 기능 구현, 경쟁 가설 디버깅 등 독립적 병렬 작업이 필요할 때, (3) 3개 이상의 독립적 작업을 동시에 수행해야 할 때, (4) 팀원들이 서로 토론하고 결과를 공유해야 할 때. subagent와 달리, 팀원 간 직접 통신과 공유 태스크 목록을 통한 자체 조율이 가능하다.
Quick start
- Choose your skills. Select individual skills above, or choose the entire pack.
- Copy the install prompt. Use Copy install prompt above the list.
- Send it to your agent. Paste into your coding agent’s chat. Your agent reviews the source and guides the setup.
Try it with your agent
After installation, start with a request like:
Use the Build Plan skill to help with my task.
Ask me for the context you need before getting started.Requirements & license
- A coding agent that can access your project files and install skills.
- Your agent checks each skill’s instructions, dependencies, and permissions during setup.
- Reported license for the skills shown: MIT. The author’s terms apply.