requirement-convergence SKILL.md skillA

requirement-convergence from shinpr/ai-coding-project-boilerplate is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.

mdr add shinpr/ai-coding-project-boilerplate/requirement-convergence@git:20260923.558b381

Pins git:20260923.558b381 in mdr.lock. Pin the fingerprint instead.

Audit grade

A, 17 of 17 checks passed

Kind

SKILL.md skill

Repository

shinpr/ai-coding-project-boilerplate

Path

.claude/skills-zh-CN/requirement-convergence/SKILL.md

Versions recorded

5

Latest version

git:20260923.558b381, committed Sep 23, 2026

First version recorded

Aug 31, 2026

Size

4,774 bytes

Repository stars

234

License

MIT

将变更必须产生的结果与为达成该结果而提出的需求区分开来，记录用户明确排除的内容，并根据结构对成本进行分级评估。当需求进入某个工作流时、设计开始之前，或提到“做到什么程度/范围外是什么/是否值得做”时使用。.

Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.

What the file says

# 需求收敛

## 目的

收到的需求往往过于臃肿、含糊不清，或指向错误的目标。能力足够的模型会将这三者调和为一个连贯的计划，并忠实地构建它——当被要求的内容本身就是错的时，也精确地交付被要求的内容。

本技能收敛的是**要构建什么**。如何构建，以及该变更需要哪些文档，都在“要构建什么”确定之后再解决。

## 收敛字段

| 字段 | 通过条件 |
|-------|----------------|
| `outcome` | 一个可观测的结果。不服务于该结果的需求即为多余。 |
| `requirements[]` | 每个与构建相关的项都标注为 `current-state` 或 `desired-future`。 |
| `nonGoals[]` | 由用户撰写，或用户声明没有范围外项。 |
| `cost` | 一个带有结构性依据支撑的分级，加上仍然存在的未知项。 |

`cost` 是一个粗略的分级，不是工作计划据以排期的工作量估算；需求无法支撑以人天为单位的估算。其未知项在决策权重上比其规模更重要。

依据所保留的用户原话进行分类，而不是依据分析器对原话的转述。要求评估的措辞、描述设想性想法的措辞，或建议某种机制的措辞，在当前收敛上下文中仅作为供判断参考的候选项保留。只有在用户明确确认之后，候选项才能进入 `requirements[]` 及长期保留的文档。

每个字段都带有各自的就绪标签：`ready`、`weak`，或 `weak-but-explicit`（弱，且用户同意暂不解决）。只有用户可以设置 `weak-but-explicit`。当所有适用字段都为 `ready` 或 `weak-but-explicit` 时，需求即视为已收敛。

各字段的判断规则见：[references/criteria.md](references/criteria.md)。

## 询问协议

使用现有的范围与成本依据来引出信息并做出判断，然后只就仓库无法回答的产品选择进行提问。`requirements[]` 和排除项的询问都依据用户提出的能力构建；分析中发现的相邻能力两者都不纳入，因此用户无需拒绝一项自己并未提出的能力。仅当某个答案会改变分析目标或所需的范围依据时，才重新分析范围与成本。

在开始前登记以下步骤，并在每个步骤完成时记录其依据：

| 步骤 | 动作 | 完成依据 |
|------|--------|---------------------|
| 1 | 陈述现有的范围事实，然后单独说明这些事实对需求意味着什么 | 列出事实及其分析依据 |
| 2 | 就低于 `ready` 的字段提问，每条消息最多两个问题 | 每个低于 `ready` 的字段一个问题 |
…

Read the whole file at its exact version.

How to install

mdr add shinpr/ai-coding-project-boilerplate/requirement-convergence@sha256:f654b6ed2db7d4cc

[![mdr](https://markdownregistry.com/badge/art_gmjpmvftyd7nldjh.svg)](https://markdownregistry.com/a/art_gmjpmvftyd7nldjh)

Versions

The registry has recorded 5 versions of this file, between Aug 31, 2026 and Sep 23, 2026, about one every 6 days.

Audit of the latest version

Source

API

GET https://markdownregistry.com/api/v1/artifacts/art_gmjpmvftyd7nldjh
GET https://markdownregistry.com/api/v1/resolve?ref=shinpr/ai-coding-project-boilerplate/requirement-convergence
GET https://markdownregistry.com/api/v1/blob/f654b6ed2db7d4ccc342fce8ac51f23bc1bb2f44b7bb575303f7f8f9a4999063

Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.

Guides: the SKILL.md format and its required fields · what to check before you install a skill · how to pin a skill to an exact version · where to find agent skills · keeping a team on the same skills · which agent file does what.

More from shinpr/ai-coding-project-boilerplate

Every file in shinpr/ai-coding-project-boilerplate

Other files named requirement-convergence

Browse by kind, by grade A, or by owner.