一键导入
plan
Structured design intake for non-trivial tasks. Asks clarifying questions and produces a design proposal.
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。
菜单
Structured design intake for non-trivial tasks. Asks clarifying questions and produces a design proposal.
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 |
| description | Structured design intake for non-trivial tasks. Asks clarifying questions and produces a design proposal. |
| allowed-tools | Bash, Read, Glob, Grep, AskUserQuestion, Agent |
Structured intake for complex tasks. The plan lives in a file, not in conversation output. Stay at the architecture level — define boundaries, interfaces, and data flow. Leave implementation details to the implementer.
AskUserQuestion — this is mandatory, never skip it. Use batches of ~4 questions, multiple-choice with 2-4 options + "Other / Not sure" when possible. Iterate until the reasoning framework is filled in — Problem, Goal, Background, Constraints. Do not write code or produce a plan until the user has answered at least one round of questions./tmp/plan-<timestamp>.md). Print a markdown link to the file (e.g., [plan-name.md](/tmp/plan-name.md)) so the user can click to open it — do not dump the plan into conversation output. Use the reasoning framework from workflow.md, structured top-down:
Amend the last commit with current changes and/or an updated message.
Stage changes and create a commit with a conventional commit message.
Full autopilot — plan, implement, and polish a task end-to-end.
Fetch and display diffs for various scopes (PR, commit, branch, staged, working tree). Used by other skills that need to analyze changes.
Execute a plan file — parse phases, spawn parallel subagents, handle failures.
Start or end a forked session for exploring rabbit holes without losing the main conversation's context.