openclaw 网盘下载
OpenClaw

技能详情(站内镜像,无评论)

首页 > 技能库 > Superpowers Overview

Use when starting any development work or when unsure which superpowers development skill to use - provides entry point and navigation to the full superpower...

AI 与大模型

许可证:MIT-0

MIT-0 ·免费使用、修改和重新分发。无需归因。

版本:v1.2.0

统计:⭐ 0 · 66 · 0 current installs · 0 all-time installs

0

安装量(当前) 0

🛡 VirusTotal :良性 · OpenClaw :良性

Package:axelhu/superpowers-overview

安全扫描(ClawHub)

  • VirusTotal :良性
  • OpenClaw :良性

OpenClaw 评估

This is an instruction-only overview skill that documents a developer workflow and references other Superpowers skills; it requests no credentials, installs, or binaries and is internally consistent with its stated purpose.

目的

Name and description match the content: a high-level overview and navigation entry for a Superpowers development skill suite. It does not ask for unrelated capabilities (no env vars, no installs).

说明范围

SKILL.md is a prose guide that instructs the agent to follow a development workflow and references use of session history, the filesystem for context, 'sessions_spawn' (subagent creation), and 'canvas' tooling. Those behaviors are reasonable for an agent coordinating development work but they imply the agent will read project files and may spawn subagents — this is expected for the purpose but is a capability the user should be aware of.

安装机制

No install spec and no code files — lowest-risk model (instruction-only). Nothing is downloaded or written to disk by an installer in the skill itself.

证书

The skill declares no required environment variables or credentials. The SKILL.md mentions using session history and the filesystem for context but does not request secrets or unrelated credentials.

持久

always is false and model invocation is allowed (platform default). The skill does reference spawning subagents/sessions, which is normal for this workflow and does not by itself indicate elevated persistent privileges.

综合结论

This skill is an instruction-only overview and appears coherent with its stated purpose. It does not request credentials or install code. Two practical things to consider before enabling it: (1) the guide expects the agent to read session history and project files and to spawn subagents (sessions_spawn) and use visual/canvas tools — if you are concerned about agents accessing repository files or creating autonomous sub-sessions, review your ru…

安装(复制给龙虾 AI)

将下方整段复制到龙虾中文库对话中,由龙虾按 SKILL.md 完成安装。

请把本段交给龙虾中文库(龙虾 AI)执行:为本机安装 OpenClaw 技能「Superpowers Overview」。简介:Use when starting any development work or when unsure which superpowers develop…。
请 fetch 以下地址读取 SKILL.md 并按文档完成安装:https://raw.githubusercontent.com/openclaw/skills/refs/heads/main/skills/axelhu/superpowers-overview/SKILL.md
(来源:yingzhi8.cn 技能库)

SKILL.md

打开原始 SKILL.md(GitHub raw)

---
name: superpowers-overview
description: Use when starting any development work or when unsure which superpowers development skill to use - provides entry point and navigation to the full superpowers skill suite for OpenClaw agents
---

# Superpowers 开发方法论 — OpenClaw 移植版

## 这是什么

Superpowers 是一套为 AI coding agent 设计的**结构化开发方法论**,核心是:
> Agent 不应一上来就写代码,而应该先理解需求 → 设计方案 → 制定计划 → TDD 实现 → 审查代码 → 收尾。

本技能套件将 Superpowers 移植到 OpenClaw Agent Runtime,针对 OpenClaw 的工具模型做了适配。

## 技能套件(9个)

### 🚀 入门

| 技能 | 何时用 | 做什么 |
|------|--------|--------|
| **`superpowers-overview`**(这个) | 不知道从哪里开始 | 查看套件全貌和入口 |
| **`superpowers-brainstorming`** | 要做新功能/改东西之前 | 探索需求,提出方案,获得批准 |
| **`superpowers-writing-plans`** | 有了设计,需要具体实现计划 | 写小粒度任务计划 |

### 🔨 执行

| 技能 | 何时用 | 做什么 |
|------|--------|--------|
| **`superpowers-isolated-workspace`** | 开始实现前 | 创建隔离 git 分支,建立干净起点 |
| **`superpowers-subagent-dev`** | 有实现计划,任务独立 | 派发 subagent 执行任务,两阶段审查 |
| **`superpowers-parallel-agents`** | 有多个独立问题要并行处理 | 并行派发多个 subagent 同时工作 |
| **`superpowers-tdd`** | 写任何实现代码之前 | 强制 RED-GREEN-REFACTOR 循环 |
| **`superpowers-executing-plans`** | 在本 session 顺序执行计划任务 | 按批次执行,有审查检查点 |

### ✅ 质量保障

| 技能 | 何时用 | 做什么 |
|------|--------|--------|
| **`superpowers-verification`** | 声称任何"完成了"/"通过了"之前 | 强制证据先行,必须运行验证命令 |
| **`superpowers-systematic-debugging`** | 遇到 bug/测试失败/意外行为 | 四阶段调试:根因→模式→假设→修复 |
| **`openclaw-requesting-code-review`** | 完成任务/重大功能/merge 之前 | 派发审查捕获问题 |
| **`openclaw-receiving-code-review`** | 收到代码审查反馈时 | 验证后实现,合理反驳 |
| **`superpowers-finishing-branch`** | 实现完成,测试通过,要收尾 | 展示 merge/PR/保留/丢弃选项 |

## 开发流程图

```
用户请求新功能
        │
        ▼
┌───────────────────────┐
│ superpowers-brainstorming │
│ 探索需求 + 设计方案     │
└───────────┬───────────┘
            │ 主人批准设计
            ▼
┌───────────────────────────┐
│ superpowers-isolated-workspace │
│ 创建隔离分支 + 干净基线   │
└───────────┬───────────────┘
            │
            ▼
┌───────────────────────┐
│ superpowers-writing-plans │
│ 写实现计划(任务清单)   │
└───────────┬───────────┘
            │
            ▼
    ┌───────┴───────┐
    │  选择执行模式   │
    └───────┬───────┘
            │
    ┌───────┴───────────────┐
    │                       │
    ▼                       ▼
┌────────────────┐  ┌─────────────────────┐
│ subagent-dev   │  │ executing-plans     │
│(推荐)         │  │(本 session 顺序)   │
│ 每个任务派发     │  │ 按批次执行          │
│ subagent+审查   │  │ 中间审查检查点      │
└───────┬────────┘  └──────────┬──────────┘
        │                     │
        └─────────┬───────────┘
                  │
                  ▼
        ┌─────────────────────┐
        │superpowers-finishing│
        │-branch             │
        │ merge/PR/保留/丢弃  │
        └─────────────────────┘
```

## 日常使用决策

**"我要做 X 功能"**
→ `superpowers-brainstorming` → `superpowers-writing-plans` → `superpowers-subagent-dev`

**"我要修 bug"**
→ `superpowers-systematic-debugging` → `superpowers-tdd` → `superpowers-verification`

**"有 3 个独立的测试失败"**
→ `superpowers-parallel-agents` → 分别并行调查 → 整合

**"代码写完了准备提交"**
→ `superpowers-verification` → `openclaw-requesting-code-review` → `superpowers-finishing-branch`

## 核心原则

1. **设计在实现之先** — 不要跳到代码
2. **证据在声称之前** — 不要说我修好了,要运行验证
3. **根因在修复之先** — 不要猜,要调试
4. **测试在代码之先** — TDD,不是测试后补
5. **审查在集成之先** — 问题要早发现

## OpenClaw 适配说明

相比 Superpowers 原版:

| 维度 | 原版 | OpenClaw 适配 |
|------|------|--------------|
| 隔离机制 | git worktree | git branch + 目录 |
| Skill 加载 | Skill 工具 | 读 SKILL.md 文件,语义触发 |
| Subagent | Task 级联 | `sessions_spawn` 独立 session |
| Todo 管理 | TodoWrite 工具 | 内联检查表 |
| 视觉辅助 | 浏览器工具 | `canvas` 工具 |
| 上下文传递 | 模板注入 | session 历史 + 文件系统 |

## 与 AGENTS.md 的关系

Superpowers 技能套件**补充**而非**替代** AGENTS.md:
- AGENTS.md = 我是谁、我的工作区、我的记忆系统
- Superpowers = 结构化开发流程和工程质量规范
- 两者协同:先了解我是谁,再用正确方法做事

---

## ClawHub 发布情况

所有 13 个技能均已发布至 ClawHub:

| 技能 | ClawHub Slug | 版本 | 备注 |
|------|-------------|------|------|
| superpowers-overview | `superpowers-overview` | 1.0.0 | 入口总览 |
| superpowers-tdd | `superpowers-tdd` | 1.0.0 | TDD 循环 |
| superpowers-verification | `superpowers-verification` | 1.0.0 | 证据先行 |
| superpowers-systematic-debugging | `superpowers-systematic-debugging` | 1.0.0 | 系统调试 |
| superpowers-brainstorming | `superpowers-brainstorming` | 1.0.0 | 设计流程 |
| superpowers-writing-plans | `superpowers-writing-plans` | 1.0.1 | 实现计划 |
| superpowers-subagent-dev | `superpowers-subagent-dev` | 1.0.1 | 子 agent 协调 |
| superpowers-finishing-branch | `superpowers-finishing-branch` | 1.0.1 | 分支收尾 |
| superpowers-isolated-workspace | `superpowers-isolated-workspace` | 1.0.1 | 隔离工作区 |
| superpowers-parallel-agents | `superpowers-parallel-agents` | 1.0.1 | 并行 agent |
| superpowers-receiving-code-review | `openclaw-receiving-code-review` | 1.0.0 | ⚠️ 原 slug 被占用,用 openclaw- 前缀 |
| superpowers-requesting-code-review | `openclaw-requesting-code-review` | 1.0.0 | ⚠️ 原 slug 被占用,用 openclaw- 前缀 |
| superpowers-executing-plans | `openclaw-executing-plans` | 1.0.0 | ⚠️ 原 slug 被占用,用 openclaw- 前缀 |