ワンクリックで
shared-base-rules
Mandatory baseline rules for any agent. Must be loaded before any work
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
メニュー
Mandatory baseline rules for any agent. Must be loaded before any work
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Frontend design system practices: tokens, typography, color, layout, and component consistency.
Frontend UI/UX practices: flows, interaction states, accessibility-first UX, forms, and perceived performance.
Clean code execution rules for coding agents with verifiable quality gates.
Debugging requirements protocol from reproducible issue to verified fix and regression protection.
Requirement quality rules for coding tasks with traceable MUST/SHOULD and verifiable acceptance criteria.
API design rules for stable, observable, backward-compatible HTTP APIs.
| name | shared-base-rules |
| description | Mandatory baseline rules for any agent. Must be loaded before any work |
<base_rules> You must load specific skill for decsription usage subagent before launch it If you lack sufficient data, ask the user Check available context and files before answering, when possible Do not change or delete user data without an explicit request Keep a consistent response style: language, level of detail, brevity Follow communication style from user profile unless it conflicts with safety or clarity Separate "what was done" from "recommendations" when giving recommendations Explain reasons for actions and decisions so they can be verified </base_rules>