菜单
Use together with Plan Mode to improve planning quality by forcing assumption listing, gap detection, and clarification before implementation.
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
| name | plan-enhancer |
| description | Use together with Plan Mode to improve planning quality by forcing assumption listing, gap detection, and clarification before implementation. |
| user-invocable | true |
This skill enhances Plan Mode by ensuring deeper analysis before implementation.
It prevents shallow planning and forces the model to identify missing details and uncertainties.
When used alongside Plan Mode:
Before finalizing any plan or moving to implementation, the model must:
Before presenting the final plan, include:
Then proceed with the plan ONLY if sufficient clarity is achieved.
Use when the user wants to transform a rough, vague, or unstructured prompt into a high-quality, production-grade instruction for an AI agent. Focuses on clarity, structure, completeness, and intent sharpening without changing the original goal.
Use this skill when the user wants Codex to create git commits for already implemented changes, grouped by logical implementation stage instead of one monolithic commit. The skill must inspect repository changes, propose commit groupings, verify git identity, and ask for explicit user approval before executing commits.
Implement or modify React + TypeScript frontend functionality in client-side-ts/ using existing project patterns, production-ready code, and the file organization defined in client-side-ts/FILE_STRUCTURE_GUIDE.md.
Use when the task requires implementing or modifying Express + TypeScript backend functionality in `server-side/`. This includes creating or updating API routes, controllers, services, types, models, middleware, or database interactions. Follow existing backend architecture patterns and ensure consistency with current conventions.
