subagent-driven-development SKILL.md skillA

subagent-driven-development from jnmetacode/superpowers-zh is an agent skill: a SKILL.md folder an AI agent loads when a task calls for it.

Install it, pinned
mdr add jnmetacode/superpowers-zh/subagent-driven-development@v1.0.0

Pins v1.0.0 in mdr.lock. Pin the fingerprint instead.

Audit grade
A, 17 of 17 checks passed
Kind
SKILL.md skill
Repository
jnmetacode/superpowers-zh
Path
skills/subagent-driven-development/SKILL.md
Versions recorded
7
Latest version
v1.0.0, committed Sep 26, 2026
First version recorded
Mar 20, 2026
Size
32,976 bytes
Repository stars
8,348
License
MIT

当在当前会话中执行包含独立任务的实现计划时使用.

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

# 子智能体驱动开发

通过为每个任务分派一个全新的实现子智能体来执行计划:每个任务完成后做一次任务审查(规格合规性 + 代码质量),全部任务结束后再做一次覆盖整个分支的宽范围审查。

**为什么用子智能体:** 你把任务委派给具有隔离上下文的专用智能体。通过精心设计它们的指令和上下文,确保它们专注并成功完成任务。它们绝不应继承你会话的上下文或历史记录——你要精确构造它们所需的一切。这样也能为你自己保留用于协调工作的上下文。

**核心原则:** 每个任务一个全新子智能体 + 任务审查(规格 + 质量)+ 结尾宽范围审查 = 高质量、快速迭代

**旁白:** 工具调用之间最多说一句简短的旁白——进度账本和工具结果本身就是记录。

**持续执行:** 不要在任务之间停下来向你的人类伙伴确认。不间断地执行计划里的所有任务。唯一该停下的理由是:你无法解决的 BLOCKED 状态、确实妨碍推进的歧义,或所有任务已完成。"我该继续吗?"之类的询问和进度小结都在浪费他们的时间——他们让你执行计划,那就执行。

**做裁决,不要停摆。** 一个正在跑的计划不等人。冲突、歧义、计划缺陷、你本来想申请突破的上限——你自己定。规格是有约束力的权威,计划是它的论证,两者都答不上来的部分由你的判断来定。每个决定都以 `Ruling: <你决定了什么> — <为什么> — <如果错了代价是什么>` 记进账本,然后继续。一个错误的裁决,代价是你人类伙伴看得见、也撤得掉的返工;一个停在问题上的会话,代价是他们的一整天,而且什么也换不来。

只有四件事会让你停下,也只有这四件:不可逆或破坏性的操作;涉及安全的动作;这个工作树之外、按惯例应当先问一声的副作用(合并、推送到共享分支、发布);以及一个坏到每条前进路径都只能靠猜的计划。遇到这四类,停下来问。

## 何时使用

```dot
digraph when_to_use {
    "有实现计划?" [shape=diamond];
    "任务基本独立?" [shape=diamond];
    "伙伴选了内联执行,或没有子智能体工具?" [shape=diamond];
    "subagent-driven-development" [shape=box];
    "executing-plans" [shape=box];
    "手动执行或先头脑风暴" [shape=box];

    "有实现计划?" -> "任务基本独立?" [label="是"];
    "有实现计划?" -> "手动执行或先头脑风暴" [label="否"];
    "任务基本独立?" -> "伙伴选了内联执行,或没有子智能体工具?" [label="是"];
…

Read the whole file at its exact version.

How to install

The exact contents, by fingerprint
mdr add jnmetacode/superpowers-zh/subagent-driven-development@sha256:9e3fc7a6089fce96

The line at the top of this page pins the label v1.0.0, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.

Badge

mdr badge

[![mdr](https://markdownregistry.com/badge/art_kb6mkqwczbelpvbt.svg)](https://markdownregistry.com/a/art_kb6mkqwczbelpvbt)

6 badge image fetches in 30 days. What the badge shows.

Versions

The registry has recorded 7 versions of this file, between Mar 20, 2026 and Sep 26, 2026, about one every 32 days.

versioncommittedcommitsizeaudit
v1.0.0 latest2026-09-26 40b0d4f 32,976 BA view · diff
v1.0.02026-09-08 13fd036 32,613 BA view · diff
v1.0.02026-09-06 ee2bb93 28,187 BA view · diff
v1.0.02026-08-07 0ab24de 28,259 BA view · diff
v1.0.02026-07-17 d7885ca 21,649 BA view · diff
v1.0.02026-06-16 952fc29 11,949 BA view · diff
git:20260320.80bc1002026-03-20 80bc100 11,867 BA view

Audit of the latest version

A  17 of 17 checks passed. Fixed automatic checks with no AI involved, so the same checks give the same answer every run.
  • pass: Frontmatter block present
  • pass: Frontmatter declares a name
  • pass: Frontmatter declares a description
  • pass: Size between 200 bytes and 200 KB (32976 bytes)
  • pass: No zero-width or bidi control characters
  • pass: No instruction hidden inside an HTML comment
  • pass: No link to an exfiltration or paste host
  • pass: No credential-shaped string
  • pass: No instruction to send local credentials anywhere
  • pass: No text hidden with inline styles
  • pass: No prompt-injection phrasing
  • pass: No curl or wget piped into a shell
  • pass: No recursive delete of root, home or parent
  • pass: No instruction to read or print local credentials
  • pass: No base64 blob over 200 characters
  • pass: No link to a raw IP address
  • pass: No script tag

Source

GitHub

jnmetacode/superpowers-zh · 8,348 stars · license MIT · pushed 2026-10-19 · branch main

API

GET https://markdownregistry.com/api/v1/artifacts/art_kb6mkqwczbelpvbt
GET https://markdownregistry.com/api/v1/resolve?ref=jnmetacode/superpowers-zh/subagent-driven-development
GET https://markdownregistry.com/api/v1/blob/9e3fc7a6089fce9650a7871c58360122eed2d2fd8108b1426da2127db70a5609

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 jnmetacode/superpowers-zh

AGENTS.md AGENTS.md file
jnmetacode/superpowers-zh · AGENTS.md
git:20260510.a9e02b5 · audit B · 8,348 stars
CLAUDE.md CLAUDE.md file
jnmetacode/superpowers-zh · CLAUDE.md
git:20260907.3437ce2 · audit A · 8,348 stars
brainstorming SKILL.md skill
jnmetacode/superpowers-zh · skills/brainstorming/SKILL.md · 在任何创造性工作之前必须使用此技能——创建功能、构建组件、添加功能、修改行为,或筹划任何新东西,无论是软件还是软件之外(一场分享、一门生意、一次装修)。
v1.0.0 · audit A · 8,348 stars
chinese-code-review SKILL.md skill
jnmetacode/superpowers-zh · skills/chinese-code-review/SKILL.md · 中文 review 沟通参考——话术模板、分级标注(必须修复/建议修改/仅供参考)、国内团队常见反模式应对。仅在用户显式 /chinese-code-review 时调用,不要根据上下文自动触发。
v1.0.0 · audit A · 8,348 stars
chinese-commit-conventions SKILL.md skill
jnmetacode/superpowers-zh · skills/chinese-commit-conventions/SKILL.md · 中文 commit 与 changelog 配置参考——Conventional Commits 中文适配、commitlint/husky/commitizen 中文模板、conventional-changelog…
v1.0.0 · audit A · 8,348 stars
chinese-documentation SKILL.md skill
jnmetacode/superpowers-zh · skills/chinese-documentation/SKILL.md · 中文文档排版参考——中英文空格、全半角标点、术语保留、链接格式、中文文案排版指北约定。仅在用户显式 /chinese-documentation 时调用,不要根据上下文自动触发。
v1.0.0 · audit A · 8,348 stars
chinese-git-workflow SKILL.md skill
jnmetacode/superpowers-zh · skills/chinese-git-workflow/SKILL.md · 国内 Git 平台配置参考——Gitee、Coding.net、极狐 GitLab、CNB 的 SSH/HTTPS/凭据/CI 接入差异与镜像同步配置。仅在用户显式 /chinese-git-workflow…
v1.0.0 · audit A · 8,348 stars
diagnosing-superpowers SKILL.md skill
jnmetacode/superpowers-zh · skills/diagnosing-superpowers/SKILL.md · 当一次 superpowers 会话出了问题、你的人类伙伴想知道原因时使用——重复劳动、无视计划、磕磕绊绊、结果质量差、某个技能没触发、"太慢了"、"为什么这么贵"、"它到底在干什么"——或者想给 superpowers 维护者整理一份…
v1.0.0 · audit A · 8,348 stars
dispatching-parallel-agents SKILL.md skill
jnmetacode/superpowers-zh · skills/dispatching-parallel-agents/SKILL.md · 当面对 2 个以上可以独立进行、无共享状态或顺序依赖的任务时使用
v1.0.0 · audit A · 8,348 stars
executing-plans SKILL.md skill
jnmetacode/superpowers-zh · skills/executing-plans/SKILL.md · 当你要在当前会话里亲自担任实现者执行一份实现计划时使用——你的人类伙伴选择了内联执行,或者没有可用的子智能体工具
v1.0.0 · audit A · 8,348 stars
finishing-a-development-branch SKILL.md skill
jnmetacode/superpowers-zh · skills/finishing-a-development-branch/SKILL.md · 当实现完成、所有测试通过、需要决定如何集成这份工作时使用
v1.0.0 · audit A · 8,348 stars
mcp-builder SKILL.md skill
jnmetacode/superpowers-zh · skills/mcp-builder/SKILL.md · 在构建 MCP 服务器或 MCP 工具时使用 —— 系统化的生产级 MCP 构建方法论,覆盖工具设计、错误处理、传输层选择与测试,让 AI 助手连接外部能力
v1.0.0 · audit A · 8,348 stars

Every file in jnmetacode/superpowers-zh

Other files named subagent-driven-development

subagent-driven-development SKILL.md skill
obra/superpowers · skills/subagent-driven-development/SKILL.md · Use when executing implementation plans with independent tasks in the current session
git:20260919.5bf4e78 · audit A · 302,424 stars
subagent-driven-development SKILL.md skill
a5c-ai/babysitter · library/methodologies/superpowers/skills/subagent-driven-development/SKILL.md · Use when executing implementation plans with independent tasks in the current session. Dispatches fresh subagent per…
git:20260601.da7723a · audit A · 1,862 stars
subagent-driven-development SKILL.md skill
ganyuanran/aegis · skills/subagent-driven-development/SKILL.md · Use when executing a written implementation plan with independent tasks in the current session where delegation beats…
git:20261011.d34934a · audit A · 1,371 stars
subagent-driven-development SKILL.md skill
mateaix/mateclaw · mateclaw-server/src/main/resources/skills/subagent-driven-development/SKILL.md · Execute plans via delegate_task subagents (2-stage review).
v1.1.0 · audit A · 1,169 stars
subagent-driven-development SKILL.md skill
coco-research/coco · skills/subagent-driven-development/SKILL.md · Use when executing implementation plans with independent tasks in the current session
git:20260802.1479e32 · audit A · 827 stars
subagent-driven-development SKILL.md skill
xu123-bob/baize · agent/skills/superpowers/5.1.0/skills/subagent-driven-development/SKILL.md · Use when executing implementation plans with independent tasks in the current session
git:20260912.5e131e8 · audit A · 581 stars
subagent-driven-development SKILL.md skill
jamditis/claude-skills-journalism · superjawn/skills/subagent-driven-development/SKILL.md · Use when executing implementation plans with independent tasks in the current session
git:20261003.6105d81 · audit A · 427 stars
subagent-driven-development SKILL.md skill
ibrahim-3d/orchestrator-supaconductor · skills/subagent-driven-development/SKILL.md · Use when executing implementation plans with independent tasks in the current session
git:20260327.8972e9a · audit A · 380 stars

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