Loading...
Loading...
Found 335 Skills
Use when asked to do a full build, complete build, rebuild, or build everything in a Quarkus project. Runs the full Quarkus build with optimized flags via a subagent to keep the main conversation responsive.
This skill should be run only when the user explicitly invokes it. Orchestrates end-to-end task implementation — understands the task, assesses complexity, implements directly or via a team of subagents for complex work, and always finishes with a code-polish pass.
Three-phase design review. Chain architect → refiner → critique subagents. Triggers on: 'design review', 'architecture review', '/arc', system design proposals, significant refactoring decisions, new service or module design.
Guides the design and structuring of workflow-based Claude Code skills with multi-step phases, decision trees, subagent delegation, and progressive disclosure. Use when creating skills that involve sequential pipelines, routing patterns, safety gates, task tracking, phased execution, or any multi-step workflow. Also applies when reviewing or refactoring existing workflow skills for quality.
Design and build AI agents for any domain. Use when users: (1) ask to "create an agent", "build an assistant", or "design an AI system" (2) want to understand agent architecture, agentic patterns, or autonomous AI (3) need help with capabilities, subagents, planning, or skill mechanisms (4) ask about Claude Code, Cursor, or similar agent internals (5) want to build agents for business, research, creative, or operational tasks Keywords: agent, assistant, autonomous, workflow, tool use, multi-step, orchestration
Use when about to ask the user a factual question, propose a solution, diagnose an error, or choose between approaches. Triggers on: 'Do you have X installed?', 'What version?', 'Is X configured?', 'We should...', 'The fix is...', 'Options: 1...', 'Based on my understanding...', 'I believe X supports...'. Before deciding anything, spin up parallel subagents to WebSearch for current docs, community solutions, framework best practices, and GitHub issues. Your memory is stale — verify everything.
Code review practices with technical rigor and verification gates. Practices: receiving feedback, requesting reviews, verification gates. Capabilities: technical evaluation, evidence-based claims, PR review, subagent-driven review, completion verification. Actions: review, evaluate, verify, validate code changes. Keywords: code review, PR review, pull request, technical feedback, review feedback, completion claim, verification, evidence-based, code quality, review request, technical rigor, subagent review, code-reviewer, review gate, merge criteria. Use when: receiving code review feedback, completing major features, making completion claims, requesting systematic reviews, validating before merge, preventing false completion claims.
RED-GREEN-REFACTOR testing for agents: dispatch subagents with known inputs, capture verbatim outputs, verify against expectations. Use when creating, modifying, or validating agents and skills. Use for "test agent", "validate agent", "verify agent works", or pre-deployment checks. Do NOT use for feature requests, simple prompt edits without behavioral impact, or agents with no structured output to verify.
Orchestrate subagent workflows for complex tasks that benefit from decomposition, role-based delegation, and parallel execution. Use when Codex should assemble a temporary team of subagents, choose roles from a reusable role library, create a controlled fallback role when no preset role fits, coordinate read-heavy work in parallel, or handle write-heavy work with ownership boundaries, staged execution, and an integrator-led merge path.
Runs a second-pass cleanup over AI-written code using the repo's style guide in style.md. Prefers parallel subagents to simplify recently modified files without changing behavior. Use when the user says "deslop", "clean this up", "make this less AI", "apply my style guide", "second pass", or asks to simplify generated code after implementation.
Aggressively clean up a codebase by removing AI slop, dead code, weak types, defensive over-engineering, duplication, and legacy cruft. Orchestrates 8 specialized subagents in parallel to deduplicate code, consolidate types, kill unused code, untangle circular dependencies, strengthen weak types, remove unnecessary try/catch, delete deprecated/legacy paths, and strip unhelpful comments. Use when the user asks to 'clean up the codebase', 'remove slop', 'improve code quality', 'remove dead code', 'kill AI slop', 'tighten types', 'remove legacy code', 'deduplicate code', 'DRY this up', 'untangle dependencies', or wants a thorough code quality pass. Also use when the user mentions code smells, technical debt cleanup, or refactoring for clarity — even if they don't use the word 'slop'.
原始人スタイル subagent への委譲判断ガイド。`genshijin-investigator` (コード位置特定)、 `genshijin-builder` (1-2ファイル編集)、`genshijin-reviewer` (diff レビュー) を inline作業 or vanilla `Explore` の代わりにスポーンするタイミングを示す。subagent 出力は原始人圧縮 → 主コンテキストに戻る tool-result が約60%縮小 → 長セッション持続。 Trigger: 「subagent 委譲」「genshijin-crew 使用」「investigator/builder/reviewer 起動」「コンテキスト節約」「圧縮 agent 出力」。