菜单
Combined safety workflow that applies destructive-command warnings and scoped edit boundaries together. Use for high-risk tasks where both command safety and edit containment matter.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
API design guidance for resources, validation, error envelopes, pagination, and compatibility. Use when defining or reviewing HTTP APIs and service interfaces.
Backend architecture and API design guidance for service boundaries, validation, persistence, and caching. Use when building or refactoring backend flows and server-side systems.
Destructive-action warning workflow. Use when operations may delete files, rewrite git history, discard changes, or modify high-risk infrastructure and data paths.
Checkpoint workflow for capturing current state, decisions, and remaining work. Use before context switching, handing off work, or pausing a long-running task.
Shared implementation standards for maintainable, safe, and readable code. Use when writing or reviewing code to keep naming, structure, validation, and error-handling expectations consistent.
Multi-source research workflow for current topics that require citations, synthesis, and explicit confidence. Use for technical, product, market, or ecosystem investigations with evidence requirements.
| name | guard |
| description | Combined safety workflow that applies destructive-command warnings and scoped edit boundaries together. Use for high-risk tasks where both command safety and edit containment matter. |
| origin | unified |
| user-invocable | true |
carefulfreeze