357건 · 관련도순
Use when creating, improving, or troubleshooting Claude Code subagents. Expert guidance on agent design, system prompts, tool access, model selection, and best practices for building specialized AI assistants.
한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.
Build custom AI agents in Claude Code from a user's problem statement. This skill analyzes the user's use case, asks smart clarifying questions, researches the internet for similar agents (GitHub repos, blogs, Claude Code community patterns), and then architects and builds production-ready Claude Code agents — including subagents, skills, hooks, slash commands, MCP integrations, and CLAUDE.md configuration. Use this skill whenever the user wants to create an agent, build an automation workflow, set up a Claude Code subagent, design a multi-agent system, or says things like 'build me an agent for...', 'automate this with Claude Code', 'I want a subagent that...', 'help me create a workflow', 'set up an agent pipeline', or any variation of wanting Claude Code to do something specialized. Also trigger when the user mentions agent architecture, agent SDK, agentic workflows, or task delegation in Claude Code — even if they don't use the word 'agent' explicitly.
한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.
Split independent work across parallel agents with explicit boundaries and a parent integration review. Use only when user allows subagents and tasks do not share mutable files.
한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.
Analyze an existing codebase in parallel via 7 specialized mapper subagents. Outputs structured findings to docs/codebase/ for persistent cross-session context. Use when onboarding to a brownfield project or when reactive on-demand file reading is too slow.
한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.
Build reusable agent workflow patterns — subagents with scoped briefs, event hooks, slash commands, and session management. Use when turning one-off agent tasks into repeatable, composable workflows.
한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.
Use ONLY when the user explicitly asks to orchestrate parallel external agents (Codex or Claude Code) in tmux sessions via tp (tmux-pilot). Or when a user tells you that you're the boss or orchestrator. You can also use this if you are an operclaw agent. NEVER auto-invoke for a generic "do this in parallel" request. For in-process subagents, use superpowers:dispatching-parallel-agents instead.
한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.
Decision guide for delegating to caveman-style subagents. Tells the main thread WHEN to spawn `cavecrew-investigator` (locate code), `cavecrew-builder` (1-2 file edit), or `cavecrew-reviewer` (diff review) instead of doing the work inline or using vanilla `Explore`. Subagent output is caveman-compressed so the tool-result injected back into main context is ~60% smaller — main context lasts longer across long sessions. Trigger: "delegate to subagent", "use cavecrew", "spawn investigator/builder/reviewer", "save context", "compressed agent output".
한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.
Run the speckit implementation phase chunk-by-chunk in clean-context subagents, then review, then produce a final report. Picks up from an existing tasks.md.
한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.
Produce a researched long-form article from a topic prompt via an orchestrated pipeline - research agent (first-person sources, working-definition gate), narrative-architecture outline, writer/cold-reviewer loop with an explicit ACCEPT/REVISE verdict contract, then a catalog-deslop pass with a regression gate. The orchestrator dispatches subagents only; the writer never judges its own draft. Use when the user says "article factory", "write an article about X", "run the article pipeline", or asks for a researched long-form piece produced end-to-end. For essays and micro posts in the user's own voice without a research stage, use the prose skill instead.
한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.
Run work as an explicit event-driven state machine executed by subagents. The orchestrator is the runtime from the canonical loop — `state, effects = init(machine); execute(effects); while state is active: event = nextEvent(); state, effects = transition(machine, state, event); execute(effects); return state.output`. All routing judgment is authored ONCE, at plan time, into a machine file (states, event alphabet, transition table, budgets, final states); at runtime the orchestrator does pure table lookups and dispatches effects — it never decides, never authors, never ingests artifacts. Subagent/sync protocol is the timeboxed-iterating one: role prompts scaffolded once to disk and dispatched BY PATH, an initialiser preamble every subagent reads first, a self-populating cheatsheet, per-unit files, tiny structured returns, independent measurement gate. Use when the user says "workmachine", "run this as a state machine", "state-machine this", or when a goal has a small number of clearly named phases with explicit outcomes (built / passed / failed / blocked) and you want the routing to be auditable rather than judged on the fly. For clock-driven loops use timeboxed-iterating; for wide graphs with judged edge criteria use workgraph.
한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.
Build a HyperFrames video composition from a brief/script autonomously. Uses separated builder and critic subagents — builders never review their own work. Quality decisions are grounded in project context (identity, audience, goals) and established visual/motion design principles. Measurable criteria (text size, contrast, overflow) are verified with tooling, not visual judgment. Use when the user says "make this video", "implement this brief", "build this composition", or provides a video script/brief to implement.
한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.
**WORKFLOW SKILL — MANDATORY for ANY hook work.** Create, configure, edit, debug, or review lifecycle hooks for VS Code Copilot (default) and Claude Code. USE FOR: criar/configurar/editar hook, hook do Copilot/Claude/agente, lifecycle hook, PreToolUse, PostToolUse, SessionStart, SubagentStop, Stop hook, UserPromptSubmit, PreCompact, bloquear comando, validar tool call, injetar contexto, guard, .agent.md frontmatter hooks, hooks.json, .github/hooks, .claude/settings.json, hook Node.js, cross-platform, hook não dispara/silencioso/ignorado, permissionDecision, hookSpecificOutput, additionalContext. **DEFAULT = VS Code GitHub Copilot.** Switch to Claude Code only when user explicitly says "Claude" OR workspace has `.claude/`. NEVER assume Claude. DO NOT USE FOR: general coding, creating agents (use agent-creator), creating skills (use skill-creator).
한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.
Run the current session as a senior developer (Opus) leading a configurable junior fleet. Decomposes requests into scoped tasks, dispatches junior-developer subagents in parallel, optionally runs an architect plan-review gate before dispatch, autonomously resolves escalations, optionally dispatches a doc-writer second wave after juniors complete, then integrates and gates with /test + /governance. Use when the user says "/dev-team", "run the dev team", "delegate this to the juniors", "act as the senior dev", or wants a task built by a senior-led fleet.
한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.
Use when designing or reviewing Agentforce topic (renamed subagent in April 2026) structure, including topic boundaries, instruction quality, handoff rules, out-of-scope behavior, and topic-selector strategy. Triggers: 'agentforce subagents (formerly topics)', 'topic design', 'topic selector', 'agent scope boundary', 'handoff conditions'. NOT for restricting what the agent may do or say (topic Scope, action filters, abuse prevention) — use agentforce/agentforce-guardrails. NOT for the mechanics of transferring a live conversation out — Omni-Channel routing and the context package — use agentforce/agentforce-agent-handoff-patterns.
한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.
Authoring and managing Agentforce agent definitions using the declarative Agent Script DSL (.agent files) and associated metadata types, including the Agent Script language for the new Agentforce Builder (hybrid-reasoning -> logic vs | prompt instructions, @actions/@subagent/@variables references, subagents). Use when creating agents in source control, writing or debugging Agent Script, debugging agent metadata, or understanding the metadata lifecycle of GenAiPlugin/GenAiPlanner/BotVersion types. Triggers: 'write or debug Agent Script', 'how do I deploy an Agentforce agent using source control', 'convert an agent topic to a subagent', 'what metadata types make up an Agentforce agent', 'agent test run command failing in CI pipeline', 'GenAiPlugin vs GenAiPlanner metadata relationship', 'Agent Script logic vs prompt instructions'. NOT for Apex-based agent actions — use agentforce/custom-agent-actions-apex. NOT for clicking the agent together in the Builder UI — use agentforce/agentforce-agent-creation.
한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.
Execute a Prometheus work plan in Codex with Boulder state, evidence ledger updates, worktree discipline, parallel subagents, and Stop-hook continuation. Use after planning when the user says start work, execute plan, continue plan, resume plan, or asks to run a .omo/plans plan.
한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.
Resolve CodeRabbit review feedback on a GitHub pull request. Collects every open CodeRabbit finding (inline threads, plus the nitpick and outside-diff findings packed into review bodies), judges and fixes each one against the real code with subagents, unifies fixes that share a pattern, commits locally, asks before pushing, then shows every reply draft linked to its finding and posts only on approval. Runs only on top-tier models. Use on "resolve the CodeRabbit feedback", "address coderabbit comments", "answer the rabbit", "fix what CodeRabbit flagged on
한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.
Use when working with *.excalidraw or *.excalidraw.json files, user mentions diagrams/flowcharts, or requests architecture visualization - delegates all Excalidraw operations to subagents to prevent context exhaustion from verbose JSON (single files: 4k-22k tokens, can exceed read limits)
한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.
Create subagent definitions for Claude Code and OpenCode that delegate to skills. Use when creating new subagents or refactoring existing ones to follow the delegation pattern.
한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.
Coordinate staged or parallel jobs across coding-agent runtimes and subagents: routes by capability/risk, isolates writers in worktrees, arbitrates results. NOT for a single-agent task (mk:cook).
한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.
Fast parallel codebase scouting. Spawns multiple Explore subagents to search directories simultaneously, returning a consolidated file map with architecture fingerprint, complexity estimates, and routing suggestions. Use before planning, debugging, or any task spanning multiple directories. NOT for reading a single known file in depth (use Read directly); NOT for semantic find-usages or go-to-definition.
한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.
Prevent session idle during background agent work via SubagentStop prompt hook auto-continuation
한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.
Arm proactive 5-hour usage protection for long Claude Code sessions (batch, subagents, high window % at start). Daemon reads account-level 5-hour window usage (not per-session). User types /usage-guard at session start. Re-arm every sitting. If user mentions usage-guard by name, read this file and run arm.sh.
한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.
Look up how Claude Code itself works — hooks, skills, subagents, MCP, plugins, settings — from the official docs, so answers come from the current spec instead of memory.
한국어 요약은 준비 중입니다. 원문 설명을 표시하고 있습니다.