work-shaping

分类: 开发与编程 | 上传者: kiku-jw | 下载: 0 | 版本: v1.0（最新）

Decide how much process a task deserves before coding: tiny vs substantial, chat vs GitHub, checklist vs execution brief, architecture lane vs full bundle, operator review vs council, and private diary vs public draft. Use when the shape of the work is the blocker.

更新日志: Source: GitHub https://github.com/kiku-jw/work-shaping-skill

目录结构

当前层级: 根目录

📁 agents/
- 📄 openai.yaml 270 B
📄 LICENSE 33.3 KB
📄 README.md 1.4 KB
📄 SKILL.md 2.2 KB

SKILL.md

---
name: work-shaping
description: 'Decide how much process a task deserves before coding: tiny vs substantial, chat vs GitHub, checklist vs execution brief, architecture lane vs full bundle, operator review vs council, and private diary vs public draft. Use when the shape of the work is the blocker.'
---

# Work Shaping

## Metadata
- Trigger when: the user is not blocked by code yet, but by uncertainty about how much process, tracking, or review the work deserves.
- Do not use when: the next step is already clearly implementation and no shaping decision remains.

## Skill Purpose

Choose the smallest honest operating lane for a task so small work stays light and meaningful work gets durable structure before it disappears into chat.

## Instructions
1. Classify the work across the main axes: size, durable surface, planning depth, architecture lane, review lane, and artifact/public-writing lane.
2. Choose the smallest honest process that matches those axes. If architecture is not `none`, route toward `$spec-bundle`; if durable tracking is needed, route toward GitHub or another explicit source of truth; if the user already implied action, do not stop at classification.
3. Return a compact shaping verdict that names the recommended next lane immediately instead of leaving the process choice implied. If another skill is needed, list only the valid next skill options in the form `$skill-name` — one-line purpose.

## Non-Negotiable Acceptance Criteria
- No ceremony for tiny reversible work.
- Substantial work does not vanish into chat without a durable surface.
- Architecture risk is made explicit instead of buried inside vague planning language.
- The answer distinguishes what is already real from what is only the recommended next step.

## Output
- A compact shaping verdict across the main axes.
- The recommended next lane or skill.
- The one or two main reasons that process level is justified.
- `Next skill options` (only if needed): `$idea-validation` — validate buyer/pain/wedge truth; `$spec-bundle` — create implementation-ready contracts and gates; `$justdoit` — create plan/status/test-plan for execution; `$product-council` — pressure-test a fuzzy high-impact choice.

登录后下载/点赞/收藏 ❤ 2 | ★ 0

work-shaping

目录结构

SKILL.md

举报内容

提示