Menu
Use when starting any conversation — establishes how to find and use skills, requiring skill invocation before ANY response including clarifying questions. 啟話即用：定技能發現與調用之法，凡答（含反問）前必先察技能。
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
Record + replay an interactive guided demo overlay on the live app — self-advancing walkthrough, labelled gesture affordances. 瀏覽器互動導覽。 Use when: demo a feature, create walkthrough, run walkthrough tool, guided product tour, replay a flow, show what was built, onboarding tour, step-by-step demo overlay.
Adversarial cooperation loop for code quality verification with plan adjustment at each phase. 對抗協作循環：實施者與驗證者互相制衡，逐階調整計劃，確保代碼品質。 Use when: execute dart task, run quality pipeline, adversarial review, implement with verification
Search + synthesize prior agent session history. Claude Code primary (~/.claude/projects/); Codex/Cursor secondary. Discover -> branch/keyword filter -> extract skeleton/errors -> session-historian subagent. Use when: what was tried, prior attempts, past decisions, recent work, why approach abandoned, cross-session context, even without saying 'sessions'.
Canonical verdict-only return schema for adversarial reviewer subagents. Reviewers emit compact ≤30-line verdict block; main loop parses verdict + blockers only. Use when: authoring or updating reviewer agent, building verdict consumer, debugging reviewer report parsing.
Diagnose + configure compound-engineering env — CLI deps, plugin version, repo-local config. Guided install for missing tools, gitignore + config bootstrap. Use when: troubleshoot missing tools, verify CE setup, onboard repo, refresh config template, /ce-setup re-check. Skip: unrelated to compound-engineering tooling.
Resolve review feedback by evaluating validity and fixing issues in parallel. Pre-commit-first. Use when: applying review feedback from local reviewer agent output, resolving pre-commit review findings, working through code-review agent suggestions, addressing reviewer-flagged issues before commit. Also: PR comments via gh as secondary path. 解決審查回饋：本地審查代理輸出優先，PR次之。
| name | using-superpowers |
| description | Use when starting any conversation — establishes how to find and use skills, requiring skill invocation before ANY response including clarifying questions. 啟話即用：定技能發現與調用之法，凡答（含反問）前必先察技能。 |
技能適任時，汝無選擇餘地 —— 必用之。此不可議，不可自圓其說繞過。
於任何回應或動作前，調相關/請求之 skill —— 含反問、探 codebase、查檔。若後證不合，不必續用。
入 plan mode 前： 若任務仍需構想、設計收斂或策略選擇，先調 ideation skill。
then 宣告 "Using [skill] to [purpose]"，遵 skill exact。有 checklist 則每項一 todo。
多 skill 適時，process skills 先 —— 定法，後由 implementation skills 執。ideation 與 systematic-debugging 為最常之 process skill，然此則通用於任一。
ideation 先收斂目的、策略、principles/decision tree/plan draft，後 implementation skills。systematic-debugging 先，後 domain skills。test-driven-development（red→green→refactor）。規劃、執行、branch-finishing 委予 worktrack-loop plugin，非 superpowers：
/worktrack:plan（worktrack-scope-tasks skill）—— 上下文預算切分、最輕 model+effort、測試先行、fresh-context 對抗審。/worktrack:start（或 /loop /worktrack:start --max-tasks 1 乾隊）。/worktrack:task。ideation 之 terminal handoff 可指 /worktrack:plan，代舊 writing-plans → executing-plans 鏈。
下列由 harness 原生具，用之，勿另立 skill：
| 需 | 原生 |
|---|---|
| 驗證完成 | /verify skill |
| Code review | /code-review（含 ultra cloud）、/review、/security-review |
| 平行 subagent | Agent tool（run_in_background） |
| Git worktree 隔離 | EnterWorktree/ExitWorktree、Agent isolation:"worktree" |
| Plan mode | EnterPlanMode/ExitPlanMode |
| Skill 發現 | Skill tool、find-skills |
| 建/審 skill | plugin-dev:skill-development、dev-standards:add-skill |
下列念頭 = STOP，汝在自圓：
| 念頭 | 實 |
|---|---|
| "此僅簡單問題" | 問即任。察 skill。 |
| "須先取 context" | Skill 察先於反問。 |
| "先探 codebase" | Skill 教汝如何探。先察。 |
| "快查 git/檔" | 檔缺對話 context。察 skill。 |
| "先蒐資訊" | Skill 教汝如何蒐。 |
| "此不需正式 skill" | Skill 存則用。 |
| "我記得此 skill" | Skill 會演進。讀現版。 |
| "此不算任務" | 動作即任。察 skill。 |
| "Skill 過度" | 簡事會轉繁。用之。 |
| "先做這一件" | 動前先察。 |
User instructions（CLAUDE.md、AGENTS.md、直接請求）勝 skills，skills 勝 default behavior。唯 human partner 明令方可略 skill workflow。