berrzebb

Create new skills, modify and improve existing skills, and validate skill structure. Use when users want to create a skill from scratch, update or optimize an existing skill, restructure SKILL.md for better triggering, or add scripts/references/assets. Do NOT use for general coding tasks, when the user just wants to use an existing skill, or when a builtin skill already covers the use case (check src/skills/ first).

워크플로우 설계 전문. 프로바이더/백엔드 구성, 에이전트 프롬프트 작성, 노드 할당, 실시간 반복 수정까지 end-to-end 워크플로우 생성·편집. Use when 사용자가 자동화 워크플로우를 만들거나 수정하려 할 때, 멀티에이전트 파이프라인을 구성할 때, closed-loop 품질 게이트가 필요할 때, 기존 워크플로우에 단계를 추가/변경/삭제할 때. Do NOT use for 단순 워크플로우 실행(이미 정의가 있으면 바로 run).

Baseline task skill example for .claude/skills/. Use as a reference template when authoring task skills. Do NOT use for builtin skills -- those go in src/skills/.

Generate map links for places, addresses, and points of interest. Use when the user mentions a location, asks for directions, or wants nearby recommendations (restaurants, cafes, etc.). Supports Google Maps, Kakao Map, and Naver Map.

Gateway direct execution 설계/구현 전문. 채널 요청을 `builtin`, `direct_tool`, `model_direct`, `workflow_compile`, `workflow_run`, `agent_required`로 분류하고, 가능한 경로는 에이전트 없이 직접 실행하도록 경계를 정리한다. Use when 게이트웨이/오케스트레이션 분리, direct tool path, model-direct path, RequestPlan, ExecutionGateway, ResultEnvelope, requested channel affinity 작업을 수행할 때. Do NOT use for phase loop 내부 구현, workflow state machine 자체 수정, 단순 프롬프트 문구 조정만 필요한 작업.

Role protocol architecture 설계/구현 전문. `src/skills/roles/*/SKILL.md`를 source of truth로 유지하면서 `RolePolicyResolver`, `ProtocolResolver`, `PromptProfileCompiler`로 역할 정책을 구조화한다. Use when 역할 기준을 런타임·워크플로우 생성기·에이전트 생성 UI가 공통으로 사용해야 할 때, `shared_protocols`를 실제 정책으로 해석해야 할 때, role baseline과 persona/task override를 분리해야 할 때. Do NOT use for 단순 문자열 prompt 조합, 한 역할의 문구만 수정하는 작업, user-facing deterministic renderer만 수정하는 작업.

디버깅 전문. 버그 추적, 근본 원인 분석(RCA), 수정 제안. Use when 버그/에러가 발생하여 원인 분석이 필요할 때. Do NOT use for 신규 기능 구현 — implementer 역할.

코드 구현 전문. 스펙 기반 파일 수정 + 셀프 검증. Use when 스펙이 확정된 코드 작성/수정 작업. Do NOT use for 설계, 리뷰, CI 실행 — 각각 PM, reviewer, validator 역할.

Run a quorum audit manually — trigger consensus review, re-run failed audits, test audit prompts, or force a specific provider. Use when the hook-based auto-trigger didn't fire, or you want explicit control.

Run any of the 20 quorum analysis tools — codebase, quality, domain checks, RTM/FVM, and audit history. Use whenever you need code analysis or domain checks.

Generate and validate design documents (Spec, Blueprint, Domain Model, Architecture) with mandatory mermaid diagrams. Use after PRD confirmation when DRM requires Design Phase. Triggers on 'generate design', 'design docs', '설계 문서', '설계 생성'.

Analyze failure scenarios for PRD requirements using Failure Driven Engineering. Generates failure tables with severity classification and derives new WBs. Use after DRM confirmation, before WB drafting. Triggers on 'failure analysis', 'FDE', '실패 분석', '장애 시나리오'.

Bootstrap a quorum-governed agent team for any project. Analyzes domain, designs team architecture, generates agents and skills with quality gates. Triggers on: 'harness 구성', 'set up harness', 'bootstrap harness', 'build a harness'.

Squash-merge a worktree branch into target with structured commit message. Use after audit consensus and retrospective.

Session orchestrator — reads handoff, picks unblocked tasks, distributes to parallel workers, tracks agent assignments, manages correction cycles.

Design tasks into tracks with work breakdowns. Writes and maintains PRDs, generates DRM-driven documents. Use whenever the user mentions planning, PRD, tracks, features, requirements, work breakdown, or 설계/기획.

Design tasks into tracks with work breakdowns and execution order. Also writes and maintains PRDs (Product Requirements Documents) — when a user requests a feature, analyzes it, adds it to the PRD, then decomposes into work breakdowns. Use for new feature planning, PRD writing, architecture changes, multi-track decomposition, requirements analysis, or adjusting existing execution plans. Trigger when user says things like 'add feature X', 'I need Y', 'plan Z', or describes any product requirement.

Run all done-criteria checks (CQ/T/CC/CL/S/I/FV/CV) and produce a pass/fail verification report. Use after implementing code, before submitting evidence to the quorum audit.

Run a consensus-loop audit manually — Codex reviews pending trigger_tag items in the watch file. Use when you want to trigger an audit without editing the watch file, re-run a failed audit, or test the audit prompt.

Squash-merge a worktree branch into the target branch with a structured commit message. Use after audit consensus and retrospective completion.

Session orchestrator for the consensus-loop — reads handoff, picks unblocked tasks, distributes to parallel workers, tracks agent assignments, manages correction cycles via SendMessage. Use when starting a work session with pending tasks, distributing implementation work, or reviewing completed worker output.

Extract learnings from audit history and conversation, manage memories, clean up stale entries. Use after completing a track, during retrospective (③ memory step), at end of session, or anytime the user wants memory maintenance. Triggers on 'what did we learn', 'memory cleanup', 'review learnings', 'retrospective', 'update memories', '회고', '메모리 정리'.

Show current consensus-loop status — pending reviews, audit state, retro marker, agent assignments. Use to check what's happening before starting work or after returning from a break.

Run consensus-loop deterministic analysis tools via CLI — code_map, dependency_graph, audit_scan, coverage_map, rtm_parse, rtm_merge, audit_history, fvm_generate, fvm_validate. Use this skill whenever you need codebase analysis (symbol index, dependency DAG, pattern scan, coverage), RTM operations (parse, merge, query), or FVM operations (generate matrix, validate against server) — even if the MCP server is not configured. This skill replaces mcp__plugin_consensus-loop_consensus-loop__* tool calls with equivalent CLI commands.

Generates numbered SQL migration files for tables, indexes, and views. Use when adding or modifying database schema.

Crawls exchange API documentation and generates structured markdown specs. Use when integrating a new exchange or updating API specifications.

Systematically diagnoses build, runtime, DB, API, and frontend errors with classification and fix workflow. Use when encountering errors or test failures.

Commits code changes with automated CHANGELOG and design document updates. Runs build/lint verification before commit. Use after completing a feature, bug fix, or refactoring.

Scaffolds a new API endpoint with router, handler, OpenAPI docs, and TS bindings. Use when adding REST endpoints to trader-api.

Scaffolds a new exchange connector with connector, provider, and trait implementations. Use when integrating a new exchange.

Adds a new trading strategy across 5 locations (mod.rs, routes, backtest, SDUI schema, frontend). Use when implementing a new strategy.

Generates SolidJS components or pages with ts-rs bindings, error handling, and reactive patterns. Use when creating new frontend UI.