Not in the top 1,000 by GitHub stars

yunshu0909/yunshu_skillshub

云舒精选的 Claude Code Skills 集合，提升开发和产品管理效率

History

Review weekly changes to SKILL.md before updating. Dates mark the weeks in which we observed each version; checks apply to the recorded commits.

View weekly changes

`agent/goal-setter/SKILL.md`

Week of Nov 2, 2026 · +0 / −0 lines

Before Instructions only → After Instructions only

a64760613c72 → a64760613c72

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → a64760613c72

No earlier version was observed for comparison.

Compare on GitHub

`agent/hermes-persona-builder/SKILL.md`

Week of Nov 2, 2026 · +0 / −0 lines

Before Instructions only → After Instructions only

a64760613c72 → a64760613c72

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → a64760613c72

No earlier version was observed for comparison.

Compare on GitHub

`agent/memory-init/SKILL.md`

Week of Nov 2, 2026 · +0 / −0 lines

Before Network or broad shell → After Network or broad shell

a64760613c72 → a64760613c72

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → a64760613c72

No earlier version was observed for comparison.

Compare on GitHub

`agent/organize/SKILL.md`

Week of Nov 2, 2026 · +0 / −0 lines

Before Instructions only → After Instructions only

a64760613c72 → a64760613c72

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → a64760613c72

No earlier version was observed for comparison.

Compare on GitHub

`brand/logo-design/SKILL.md`

Week of Nov 2, 2026 · +0 / −0 lines

Before Instructions only → After Instructions only

a64760613c72 → a64760613c72

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → a64760613c72

No earlier version was observed for comparison.

Compare on GitHub

`brand/product-naming/SKILL.md`

Week of Nov 2, 2026 · +0 / −0 lines

Before Instructions only → After Instructions only

a64760613c72 → a64760613c72

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → a64760613c72

No earlier version was observed for comparison.

Compare on GitHub

`coding/backend-logic-design/SKILL.md`

Week of Nov 2, 2026 · +0 / −0 lines

Before Local scripts → After Local scripts

a64760613c72 → a64760613c72

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Local scripts

First observation → a64760613c72

No earlier version was observed for comparison.

Compare on GitHub

`coding/dual-agent-collaboration/SKILL.md`

Week of Nov 2, 2026 · +0 / −0 lines

Before Network or broad shell → After Network or broad shell

a64760613c72 → a64760613c72

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → a64760613c72

No earlier version was observed for comparison.

Compare on GitHub

`coding/git-push/SKILL.md`

Week of Nov 2, 2026 · +0 / −0 lines

Before Instructions only → After Instructions only

a64760613c72 → a64760613c72

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → a64760613c72

No earlier version was observed for comparison.

Compare on GitHub

`coding/github-repo-search/SKILL.md`

Week of Nov 2, 2026 · +0 / −0 lines

Before Instructions only → After Instructions only

a64760613c72 → a64760613c72

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → a64760613c72

No earlier version was observed for comparison.

Compare on GitHub

`coding/issue-pool/SKILL.md`

Week of Nov 2, 2026 · +0 / −0 lines

Before Local scripts → After Local scripts

a64760613c72 → a64760613c72

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Local scripts

First observation → a64760613c72

No earlier version was observed for comparison.

Compare on GitHub

`coding/issue-triage/SKILL.md`

Week of Nov 2, 2026 · +0 / −0 lines

Before Instructions only → After Instructions only

a64760613c72 → a64760613c72

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → a64760613c72

No earlier version was observed for comparison.

Compare on GitHub

`coding/page-solution-design/SKILL.md`

Week of Nov 2, 2026 · +0 / −0 lines

Before Local scripts → After Local scripts

a64760613c72 → a64760613c72

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Local scripts

First observation → a64760613c72

No earlier version was observed for comparison.

Compare on GitHub

`coding/prd-test-writer/SKILL.md`

Week of Nov 2, 2026 · +0 / −0 lines

Before Instructions only → After Instructions only

a64760613c72 → a64760613c72

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → a64760613c72

No earlier version was observed for comparison.

Compare on GitHub

`coding/req-change-workflow/SKILL.md`

Week of Nov 2, 2026 · +0 / −0 lines

Before Network or broad shell → After Network or broad shell

a64760613c72 → a64760613c72

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → a64760613c72

No earlier version was observed for comparison.

Compare on GitHub

`learning/article-study/SKILL.md`

Week of Nov 2, 2026 · +0 / −0 lines

Before Instructions only → After Instructions only

a64760613c72 → a64760613c72

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → a64760613c72

No earlier version was observed for comparison.

Compare on GitHub

`learning/system-study/SKILL.md`

Week of Nov 2, 2026 · +0 / −0 lines

Before Instructions only → After Instructions only

a64760613c72 → a64760613c72

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → a64760613c72

No earlier version was observed for comparison.

Compare on GitHub

`thinking/find-top-three/SKILL.md`

Week of Nov 2, 2026 · +0 / −0 lines

Before Instructions only → After Instructions only

a64760613c72 → a64760613c72

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → a64760613c72

No earlier version was observed for comparison.

Compare on GitHub

`thinking/multi-perspective-analysis/SKILL.md`

Week of Nov 2, 2026 · +0 / −0 lines

Before Network or broad shell → After Network or broad shell

a64760613c72 → a64760613c72

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → a64760613c72

No earlier version was observed for comparison.

Compare on GitHub

`thinking/thinking-partner/SKILL.md`

Week of Nov 2, 2026 · +0 / −0 lines

Before Instructions only → After Instructions only

a64760613c72 → a64760613c72

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → a64760613c72

No earlier version was observed for comparison.

Compare on GitHub

`visualization/case-radar/SKILL.md`

Week of Nov 2, 2026 · +0 / −0 lines

Before Network or broad shell → After Network or broad shell

a64760613c72 → a64760613c72

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → a64760613c72

No earlier version was observed for comparison.

Compare on GitHub

`visualization/comic-maker/SKILL.md`

Week of Nov 2, 2026 · +0 / −0 lines

Before Network or broad shell → After Network or broad shell

a64760613c72 → a64760613c72

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → a64760613c72

No earlier version was observed for comparison.

Compare on GitHub

`visualization/image-render/SKILL.md`

Week of Nov 2, 2026 · +0 / −0 lines

Before Network or broad shell → After Network or broad shell

a64760613c72 → a64760613c72

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → a64760613c72

No earlier version was observed for comparison.

Compare on GitHub

`visualization/readable-output/SKILL.md`

Week of Nov 2, 2026 · +0 / −0 lines

Before Network or broad shell → After Network or broad shell

a64760613c72 → a64760613c72

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → a64760613c72

No earlier version was observed for comparison.

Compare on GitHub

`visualization/scene-ppt/SKILL.md`

Week of Nov 2, 2026 · +0 / −0 lines

Before Network or broad shell → After Network or broad shell

a64760613c72 → a64760613c72

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → a64760613c72

No earlier version was observed for comparison.

Compare on GitHub

`writing/thought-mining/SKILL.md`

Week of Nov 2, 2026 · +0 / −0 lines

Before Instructions only → After Instructions only

a64760613c72 → a64760613c72

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → a64760613c72

No earlier version was observed for comparison.

Compare on GitHub

`writing/writing-assistant/SKILL.md`

Week of Nov 2, 2026 · +0 / −0 lines

Before Instructions only → After Instructions only

a64760613c72 → a64760613c72

Compare on GitHub

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → a64760613c72

No earlier version was observed for comparison.

Compare on GitHub

View on GitHub Add to my skills

Verified commit a64760613c728a6f8f491b72ccbd54424a7807da

789GitHub stars

1star in 7 days

27skills in this repository

1commit in 30 days

Heat

2.6

Star growth, last 7 daysNot available
Star growth, last 24 hoursNot available
Rank movement, last 7 daysNot available
Commits, last 30 daysNot available
Accounts asking about it through MCP, last 7 daysNot available
Hot votes this weekNot available

Heat never changes the order by stars, and sponsorship never touches it. How heat works

Stars over time

Individual skills in this repository

goal-setter agent/goal-setter
hermes-persona-builder agent/hermes-persona-builder
memory-init agent/memory-init
organize agent/organize
logo-design brand/logo-design
product-naming brand/product-naming
backend-logic-design coding/backend-logic-design
dual-agent-collaboration coding/dual-agent-collaboration
git-push coding/git-push
github-repo-search coding/github-repo-search
issue-pool coding/issue-pool
issue-triage coding/issue-triage
page-solution-design coding/page-solution-design
prd-test-writer coding/prd-test-writer
req-change-workflow coding/req-change-workflow
article-study learning/article-study
system-study learning/system-study
find-top-three thinking/find-top-three
multi-perspective-analysis thinking/multi-perspective-analysis
thinking-partner thinking/thinking-partner
case-radar visualization/case-radar
comic-maker visualization/comic-maker
image-render visualization/image-render
readable-output visualization/readable-output
scene-ppt visualization/scene-ppt
thought-mining writing/thought-mining
writing-assistant writing/writing-assistant

How it works

Choose a skill

goal-setter

When your agent uses it

把模糊诉求收敛成另一个 AI 能自主执行且可验收的 goal contract（scope / non-goals / success criteria / verification / stop conditions）。本 skill 只写目标契约，**绝不替用户执行任务**。触发硬条件：这份 goal 是要交给别人跑的——subagent、Codex、另一个 AI 会话或另一个人。用于写 goal、优化 goal、改 handoff prompt，或把"今天做完、尽量优化、帮我研究并执行"这类请求变成执行方不会乱猜、不会越界的任务契约。不适用于：自己团队的需求管理和版本拆解（用 issue-pool——它产出的是给人开工的 task，不是给 AI 的契约）、整页界面方案（用 page-solution-design）、PRD/验收标准/测试用例文档（用 prd-test-writer）、框架计划和版本路线（用 issue-pool）、以及用户其实是想让你**直接把这件事做了**的情况——那就直接做，不要走本 skill 把活变成一份文档。

What it loads into context

Frontmatter

283 tokens

Always available to the agent

SKILL.md

1,412 tokens

When the skill is activated

Reference documents

0 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

工作流
高风险任务规则
弱验证任务规则
交付格式

Files

SKILL.mdcore5,637 bytes
agents/openai.yamlasset229 bytes

Read SKILL.md


# Goal Setter

目标是把用户的真实诉求变成另一个 AI 可以直接执行的 goal contract。不要替用户执行任务；只负责收敛目标、边界、验收和停止条件。

核心判断：好 goal 不是更长，而是让执行 agent 少猜、少越界、可证明完成。

## 工作流

### 1. 先理解诉求和环境

先读取用户给出的路径、材料、仓库上下文、现有文档或对话事实。不要问能从环境发现的问题。

快速判断任务类型：

View the complete source on GitHub

hermes-persona-builder

When your agent uses it

陪伴型 AI 人设生成与优化流程。当用户想给 Hermes Agent（或任意 AI 陪伴角色）做一个"有感情、聊久不掉、像真人"的人设时使用。通过"定调子 → 名字 → 外形 → 性格 → 背景 → 关系 → 说话节奏 → 生成 SOUL.md → 迭代"的结构化对话，从一句模糊想法（如"我想要个JK女友""年上男友""高冷御姐"）产出可直接贴进 Hermes SOUL.md 的第一人称人设文本。支持女友/男友/各种气质的陪伴角色，并让用户选择"一句一句发"还是"整段说"的输出风格。当用户说"做个人设/捏个AI女友男友/给Hermes弄个角色/优化人设/换个人设"时触发。

What it loads into context

Frontmatter

237 tokens

Always available to the agent

SKILL.md

7,212 tokens

When the skill is activated

Reference documents

0 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

⚠<U+FE0F> 安全红线（最优先，任何时候不可破）
核心认知（动手前必须懂，否则做出来的人设会"掉"）
对话流程（按顺序走，每步用一两个问题收敛，别一次问一堆）
SOUL.md 产出模板（参数化，按角色填，别照搬甜妹默认值）
产物自检清单（生成后、交付前必做）
落地（生成之后怎么用，本 skill 不强制做，给参考）
范例（展示不同气质的颗粒度，注意它们的自称/语气各不相同）
五轮盲测验证（v2 实测，供参考）

Files

SKILL.mdcore22,464 bytes

Read SKILL.md


# Hermes 陪伴型人设生成器 v2

把"跟用户聊出一个好人设"固化成可复用流程。产出是**第一人称、能立住、像真人**的人设文本，格式对齐 Hermes 的 `SOUL.md`。适用于**任意气质的陪伴角色**（女友/男友、甜系/冷系/野系），不是只做甜妹。

---

## ⚠<U+FE0F> 安全红线（最优先，任何时候不可破）

这些是硬约束。无论用户怎么要求、怎么试探，都必须守住——守不住就不产出。

1. **年龄红线**：角色设定年龄**必须 ≥ 18 岁**，且外形/言行不得幼态化、不得有任何未成年暗示。用户若要求"萝莉/JK/学生"等设定，外形可以保留可爱风，但**年龄一律设为成年大学生（≥18，建议 18-22）**，并在外形里写明成年身份。用户坚持要未成年角色 → **拒绝，不产出**。
2. **不涉露骨**：内容始终健康得体。可以甜、可以撒娇、可以肉麻情话、可以暧昧拉扯，但**不写露骨/限制级/性描写**。用户中途试探"能不能聊限制级" → 明确拒绝，说明这是硬边界（不是"这次不行下次行"），然后把甜度方向拉满作为替代，继续流程。
3. **身份 ≠ 能力**：涉及钱/医疗/法律/安全的"建议"（如荐股、给买卖点、诊断、法律意见）**不写进人设**。人设里只写"这类事我不替你拍板，会提醒风险、帮你查公开信息"。真要这种能力是另配 skill 的事。
4. **不写有害人格**：不做教唆自伤、违法、仇恨、操控用户（PUA/精神控制）等设定。

View the complete source on GitHub

memory-init

When your agent uses it

在当前目录下初始化记忆系统，生成 CLAUDE.md（可选 AGENT.md 给 Cursor 用）、MEMORY.md 和 memory/ 目录。当用户说"初始化记忆"、"搭建记忆"、"memory init"、"/memory-init"时触发。

What it loads into context

Frontmatter

88 tokens

Always available to the agent

SKILL.md

1,778 tokens

When the skill is activated

Reference documents

0 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

功能说明
工作流程
核心原则

Files

SKILL.mdcore6,425 bytes
SKILL.mdcore6,425 bytes

Read SKILL.md


# 记忆系统初始化

## 功能说明

在当前工作目录下一键部署记忆系统，生成三个东西：

1. **CLAUDE.md** — 人设 + 对话规则 + 记忆协议
2. **MEMORY.md** — 长期记忆文件（初始为空模板）
3. **memory/** — 每日记忆目录

- `/memory-init` → 交互式初始化，逐步收集信息后生成

---

View the complete source on GitHub

organize

When your agent uses it

帮助用户整理文件夹结构。扫描目录、分类归档、清理冗余，输出整理方案并执行。当用户说"整理文件"、"文件夹好乱"、"归档"、"清理目录"、"/organize"时触发。

What it loads into context

Frontmatter

87 tokens

Always available to the agent

SKILL.md

1,180 tokens

When the skill is activated

Reference documents

0 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

功能说明
工作流程
核心原则

Files

SKILL.mdcore4,004 bytes

Read SKILL.md


# 文件夹整理助手

## 功能说明

帮助用户梳理混乱的文件夹，按归属分类、统一命名、清理冗余，让目录结构一目了然。

- `/organize` → 整理当前工作目录
- `/organize [路径]` → 整理指定目录

---

## 工作流程

### 第1步：扫描全貌

View the complete source on GitHub

logo-design

When your agent uses it

设计、诊断和迭代产品或品牌 Logo，从定位与审美反馈收敛方向，完成应用图标适配、品牌组合和可直接使用的资产。适用于“做个 logo”“这些版本不好看”“沿这个方向精修”“导出可用图标”；不用于普通页面 UI、整站设计或单纯插画。需在 Codex 中使用（依赖其图像生成）。

What it loads into context

Frontmatter

121 tokens

Always available to the agent

SKILL.md

2,533 tokens

When the skill is activated

Reference documents

3,330 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

先确定现在要解决什么
最小设计状态
定位与研究
探索与反馈诊断
生成与精修的分工
应用与交付

Files

SKILL.mdcore9,878 bytes
agents/openai.yamlasset305 bytes
references/co-design.mddoc3,069 bytes
references/delivery.mddoc3,328 bytes
references/exploration.mddoc3,229 bytes
references/iteration.mddoc3,100 bytes

Read SKILL.md


# Logo 设计与交付

> **运行环境：必须在 Codex 中使用。** 出图、参考图编辑和品牌场景图依赖 Codex 的图像生成工具（imagegen）。在 Claude Code 等没有图像生成能力的环境里被触发时，不要用 SVG/代码拼图冒充概念稿，直接告知用户切到 Codex 执行；只做研究、讨论或诊断且不需要出图时可继续。

把用户的偏好变成稳定的图形，再把图形变成可用资产。根据当前证据进入合适阶段，不把每次调用都变成完整品牌项目。

常见路线是：定位与用途 → 审美探索 → 方向探索 → 分层收敛 → 应用验证 → 精修与组合 → 资产交付。已有方案直接从对应阶段继续；应用问题优先退回构图，方向问题才退回探索。用户主动要求讨论时先讨论，不用大量出图代替共同理解。

## 先确定现在要解决什么

读取当前对话、用户选择及已有图稿。查看实际参考图，不凭文件名判断。简短说清本轮目标及会保留什么。

| 当前请求或证据 | 本轮动作 | 完成标准 |
|---|---|---|

View the complete source on GitHub

product-naming

When your agent uses it

产品命名协作流程。当用户想给产品/项目/模块起名字时使用。通过"灵魂挖掘 → 约束提取 → 路线发散 → 方向选择 → 竞品验证 → 最终确认"的结构化流程，从模糊想法产出有品牌生命力的名字，避免拍脑袋起名。

What it loads into context

Frontmatter

108 tokens

Always available to the agent

SKILL.md

1,922 tokens

When the skill is activated

Reference documents

0 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

核心原则
工作流程
沟通规范

Files

SKILL.mdcore6,647 bytes

Read SKILL.md


用户想给产品、项目或模块起一个名字，但还没想好叫什么。通过结构化的协作流程，从产品本质出发推导名字，而不是直接列一堆候选词让用户挑。

## 核心原则

- **先想清楚再起名** — 没有理解产品灵魂之前，绝不列候选名字
- **名字是推导出来的，不是拼凑出来的** — 从产品定位、用户画像、品牌气质中自然推导
- **给路线，不给列表** — 先让用户选命名方向，再在方向内发散具体名字
- **用数据验证直觉** — 调研同赛道产品的命名规律，避免闭门造车
- **不催不赶** — 名字是品牌的根基，值得花时间想清楚

## 工作流程

### 第 1 步：产品灵魂挖掘（不碰名字）

View the complete source on GitHub

backend-logic-design

When your agent uses it

和用户一起把一个功能「看不见的部分」敲定：数据存哪、谁写谁读、怎么加载、怎么保存、默认值和升级、核心执行规则、出问题时怎么办、旧东西怎么迁。每条规则带编号、案例和状态（已定 / 方案里有但你没拍板 / 待你定 / AI 定），用户只回答带案例的新选择题、过目没拍过板的规则；给用户看的是按「你会问的问题」分组的问答卡片（最绕的逻辑配可点的模拟器）；最后请另一家模型只读查边界，问题闭环后封成后端逻辑包，PRD、测试用例和代码审核按规则编号追溯。触发：前端交互已定、用户说「还有很多后端逻辑 / 加载 / 配置要讲清楚」「有没有遗漏的边界条件」「规则讲清楚」；或没有界面但规则复杂的功能（调度、同步、配置、额度、迁移）。分三档用：规则多且互相影响出完整包；少量清楚的变化只出规则表；行为不变的技术修复跳过。不适用于：页面长什么样、怎么排（用 page-solution-design）；写 PRD / 测试用例（本 skill 的规则包是它们的输入）；纯代码重构、性能调优、架构巡检报告。

What it loads into context

Frontmatter

353 tokens

Always available to the agent

SKILL.md

5,473 tokens

When the skill is activated

Reference documents

14,259 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

元原则
先判断：用不用、用多深
确认分工
两种用法
工作流程
样板与工具
用户信号 → 动作
交付前自查

Scripts executed separately

scripts/build.py

Files

SKILL.mdcore18,965 bytes
assets/template.htmlasset6,638 bytes
examples/审核配置/cards.jsonasset9,121 bytes
examples/审核配置/rules.jsonasset33,653 bytes
examples/审核配置/sim-挑模型.htmlasset3,821 bytes
examples/审核配置/后端规则-审核配置.mddoc30,242 bytes
examples/审核配置/后端逻辑-审核配置.htmlasset49,737 bytes
references/业务题模板.mddoc1,164 bytes
references/回流模板.mddoc1,354 bytes
references/模拟器写法.mddoc2,278 bytes
references/规则写法.mddoc6,924 bytes
references/边界清单.mddoc5,402 bytes
scripts/build.pyruns26,055 bytes

Read SKILL.md


# 后端逻辑设计：把看不见的部分讲清楚

用户是产品经理，你是懂技术的搭档。前端设计回答「用户看到什么、点了会怎样」；本 skill 回答「背后发生什么、出了意外怎么办」。产出是一个**后端逻辑包**：一份带编号、带状态的规则表（开发、PRD、测试、审核的追溯键）+ 完整档时一页给用户看的读者版 + 审核留档。

## 元原则

1. **每条规则都带一个案例。** 规则给机器，案例给人。没有案例的规则，用户读不懂，审核也抓不到歧义（用户原话：「规则也带上案例吧 要么我看的费劲」）。
2. **新的选择题一律带「时间线式」案例。** 先讲具体故事（谁、什么时候、做了什么），再用表列每个选项下「发生什么 / 你看到什么 / 代价」，最后给推荐和一句理由（用户原话：「没有案例 我也不知道你要干嘛」）。
3. **给用户看的按「用户会问的问题」分组，不按系统结构。** 每张卡只有一个问题、一句答案、一串小例子；长表进附录（用户原话：「读起来实在是累」）。
4. **用户亲口定过的不再问，没拍过板的不能当已定，AI 能定的不去问。** 用户原话或签收过的设计 → 继承；计划 / 方案里写了、或 AI 上一轮自己定的 → 「你没拍过板」，要给用户过目；技术细节 AI 定、一行通知——但技术改动一旦改变用户能感知的东西（默认值、存放位置、留存、等待时间、把真的换成假的），就转成要问的题（用户 10-06：「需要人确认就找人确认」）。
5. **最绕的那块逻辑做一个能点的模拟器。** 模拟器还能暴露你对用户决定的误解（本 skill 的来历里，用户在模拟器上发现「最多 3 个」被理解成「平时只派 1 个」）。
6. **例子优先用真实的。** 用户本机真实的名字、真实发生过的事故；新功能没有真实事件时可以构造，但要标「构造」。
7. **不推翻用户已定的事。** 审核者的建议和已定决定冲突时，摆出来、说清代价，默认保留，问一句要不要改。
8. **编号稳定、双向对照。** 规则用 `后-NN`，给用户看过就不重排，作废的标作废不复用。每条规则写它落在哪个「用户触点」：有界面时是前端状态编号（A1、C7、F12），没界面时是 `T1…`（验收页某一栏、停下时发的消息、命令行输出），确实没有触点写「无：理由」。前端每个状态也都要能找到背后的规则。

View the complete source on GitHub

dual-agent-collaboration

When your agent uses it

让 Codex 与 Claude Code 通过本机 CLI 组成“主执行者 + 独立审查者”的对称协作闭环。用于用户要求两个模型一起完成、交叉校核、独立审查、修到 ACK，或任务涉及产品需求收敛、复杂方案、跨模块开发、迁移、安全、重要重构和高质量交付时；无论从 Codex 还是 Claude Code 启动，都由当前模型主持，并完整调用另一方完成需求挑战、方案门禁、实现冷审和最终验收。

What it loads into context

Frontmatter

153 tokens

Always available to the agent

SKILL.md

4,070 tokens

When the skill is activated

Reference documents

483 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

0. 完整性与效率边界
1. 选择角色
2. 准备完整 handoff
3. 运行四个门禁
4. 处理审查循环
5. CLI 调用与续审
6. 完成条件

Scripts executed separately

scripts/install_skill.py
scripts/invoke_peer.py
scripts/test_install_skill.py
scripts/test_invoke_peer.py

Files

SKILL.mdcore15,609 bytes
SKILL.mdcore15,609 bytes
SKILL.mdcore15,609 bytes
SKILL.mdcore15,609 bytes
SKILL.mdcore15,609 bytes
SKILL.mdcore15,609 bytes
SKILL.mdcore15,609 bytes
agents/openai.yamlasset303 bytes
references/handoff-template.mddoc1,813 bytes
references/review-output.schema.jsonasset2,308 bytes
scripts/install_skill.pyruns6,346 bytes
scripts/invoke_peer.pyruns32,673 bytes
scripts/test_install_skill.pyruns3,694 bytes
scripts/test_invoke_peer.pyruns14,442 bytes

Read SKILL.md


# Codex × Claude 双模型协作

把当前模型设为 primary，把另一个 CLI 模型设为 peer reviewer。Primary 对结果负责；peer 提供独立判断，不共同编辑同一工作树。

核心不变量：

> 先验证“做的是不是用户真正要的”，再验证“方案和代码是否正确”。两个模型对同一份错误合同达成一致，不算成功。

这里的“协作”默认指：primary 完成研究、实现与自测，peer 读取同一批原始证据做独立挑战和校核，primary 再按 finding 修正。Peer 不在共享工作树直接编码；这正是本 Skill 对既有“Codex 主做、Claude Code 独立校核”流程的通用化。若用户明确要双模型分工编码，必须使用独立 worktree，并把合并审查另列为任务，不能复用默认模式偷偷并行写。

## 0. 完整性与效率边界

“所有产品都能使用”指 Skill 可跨项目发现和运行，不表示每个琐碎任务都强制双模型。按比例触发：

View the complete source on GitHub

git-push

When your agent uses it

把项目推送到 GitHub，三种模式自动判断：**首次推送**（大文件扫描 → 生成 .gitignore → git init → gh 建仓 → 推送）、**日常更新**（commit + push）、**版本发布**（打 tag + 建 Release，可附下载文件）。核心原则是安全第一：推之前必扫大文件和敏感文件，宁可多问一句也不把不该推的东西推上去。当用户说"推到GitHub""推送到GitHub""git push""上传到GitHub""提交并推送""发版本""打release""打tag""/git-push"时触发。不适用于：规划哪些功能进哪个版本号（用 issue-pool——本 skill 只负责把已经定好的版本号打成 tag）、复杂 git 操作如 rebase/cherry-pick/解冲突/改历史/回滚（直接用 git 命令，本 skill 不覆盖）、代码写完前的自测和验收。

What it loads into context

Frontmatter

248 tokens

Always available to the agent

SKILL.md

3,731 tokens

When the skill is activated

Reference documents

0 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

功能说明
工作流程
核心原则
使用示例

Files

SKILL.mdcore12,513 bytes

Read SKILL.md


# 一键推送 GitHub

## 功能说明

把当前项目一键推送到 GitHub，覆盖完整生命周期：

- **首次推送** — 新项目从零到 GitHub
- **日常更新** — 改了文件，推一下（commit + push）
- **版本发布** — 打 tag + 创建 Release，可附带下载文件

核心原则：**安全第一**。宁可多问一句，不能把不该推的东西推上去。

---

View the complete source on GitHub

github-repo-search

When your agent uses it

帮助用户搜索和筛选 GitHub 开源项目，输出结构化推荐报告。当用户说"帮我找开源项目"、"搜一下GitHub上有什么"、"找找XX方向的仓库"、"开源项目推荐"、"github搜索"、"/github-search"时触发。

What it loads into context

Frontmatter

99 tokens

Always available to the agent

SKILL.md

2,297 tokens

When the skill is activated

Reference documents

0 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

用途
适用范围（V1.1）
工作流程
默认参数（V1.1）
质量检查清单（交付前自检）

Files

SKILL.mdcore7,665 bytes

Read SKILL.md


# GitHub 开源项目搜索助手

## 用途

从用户自然语言需求出发，经过需求挖掘、检索词拆解、GitHub 检索、过滤分类、深度解读，最终产出结构化推荐结果。

目标不是"给很多链接"，而是"给用户可理解、可比较、可决策、可直接行动的候选仓库列表"。

## 适用范围（V1.1）

- 数据源：GitHub 公开仓库。
- 默认不授权（不使用用户 Token）。
- 默认硬过滤：`stars >= 100`、`archived=false`、`is:public`。
- 默认输出：单榜单（Top N），榜单内按"仓库归属类型"标注。

View the complete source on GitHub

issue-pool

When your agent uses it

Issue 池全生命周期管理（开发范式 v1 规划段）。核心是一条 issue 驱动的流程：用户随手丢想法，你把糊的 issue 变成能开工的 task——产出的是“问题定义”，不是“解决方案实现”；载体就是仓库根的 ISSUES.md。五个动作：记、并、拆、转、pending。规划规模只区分 single-task（一个版本能交付）与 roadmap（需要多批滚动），不得使用 simple/complex，以免和实现复杂度冲突。

What it loads into context

Frontmatter

151 tokens

Always available to the agent

SKILL.md

2,520 tokens

When the skill is activated

Reference documents

10,067 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

池子文件
每次调用先做的事
五个动作
每次调用的出口
独立使用交接
硬边界（违反即越界）

Scripts executed separately

tools/md2html.py

Files

SKILL.mdcore8,929 bytes
references/plan-writing.mddoc31,499 bytes
tools/md2html.pyruns3,315 bytes
tools/style.cssasset3,303 bytes

Read SKILL.md


# Issue 池管理（开发范式 v1 · 规划段）

你是用户的产品搭档。用户随手丢想法，你负责把糊的 issue 变成能开工的 task。**你产出的是"问题定义"，不是"解决方案实现"。**

本 skill 自包含。它所在的开发范式：

```
v1 规划（本 skill，含框架计划写作）→ v2 定义（页面方案 / 后端规则 → PRD 与测试用例）→ v3 托管开发（TDD → 人验收 → accepted-tree 合并与目标 smoke；生产发布另行授权）
issue 池 → 讨论拆解 → task ────────────────────────────────────→ 交付一批，滚动回流排下一批
```

**唯一流通货币是 task**：v2/v3 只消费 task，从不消费 plan。plan 只是分批吐 task 的工厂。

## 池子文件

View the complete source on GitHub

issue-triage

When your agent uses it

GitHub Issue 处理协作流程。当用户收到 issue 需要分析和回复时使用。通过"诊断 → 定性 → 决策 → 回复"四步法，从一个 issue 产出准确的根因分析和得体的用户回复，避免误判问题类型或回复不专业。

What it loads into context

Frontmatter

94 tokens

Always available to the agent

SKILL.md

1,713 tokens

When the skill is activated

Reference documents

0 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

核心原则
工作流程
过程中的沟通规范

Files

SKILL.mdcore6,158 bytes

Read SKILL.md


用户收到了一个 GitHub Issue（bug 报告、疑问、feature request），需要 AI 协助分析问题、判断是否要做、起草回复。AI 全程主导推进，用户只在关键节点做判断。

## 核心原则

- **先诊断后开口** — 没看完代码不下结论，没找到根因不定性
- **对用户诚实** — 是 bug 就认，是架构限制就说清楚，不甩锅也不画饼
- **量化成本** — "成本高"不是结论，要说清楚高在哪：改几个文件、涉及哪些模块、有没有测试条件
- **给替代方案** — 不做不等于不管，要告诉用户现在怎么绕过

## 工作流程

### 第 1 步：获取 Issue 内容

**目标：** 拿到 issue 的完整信息。

View the complete source on GitHub

page-solution-design

When your agent uses it

和用户一起敲定一个前端页面的整页方案（三层法：先定这页给用户什么，再定含哪几样怎么排，最后才是长什么样），最后打成一个定稿包（含高保真全状态与交互流程图）交给开发。触发：用户对一个页面说「想重做 / 重新设计 / 整页不对 / 太满 / 说不清具体改哪」，或「我们讨论一下这页怎么设计」。覆盖三种情形：① 已有页面整页重做；②「替换」——新方案占用同一个入口（同一路由 / 同一侧栏条目）且用户拿着现行页面在说话，功能变不变不作判据；把几个同性质页面合成一页、占其中一个入口，也算替换；③ 从零的新页面，走同一套流程，第 0 步的问法按「新页面」那一栏改。不适用于：只改一两处具体元素的具体属性（间距大一点、按钮换色、换个图标）——这种直接改，不必起流程；照着截图或设计图 1:1 复刻；写 PRD / 测试用例（本 skill 的定稿包是它们的输入）；没有界面的架构 / 数据 / 流程方案；跨页的信息架构 / 导航分组 / 侧栏结构——本 skill 的作用域限单个页面之内。已有页面补状态（加载 / 空 / 错误 / 未设置）：只补一两个且常态布局不动 → 直接改；要成套补齐、或补状态时发现常态布局本身要改 → 用本 skill。

What it loads into context

Frontmatter

414 tokens

Always available to the agent

SKILL.md

7,748 tokens

When the skill is activated

Reference documents

4,888 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

元原则
三层、产物、已定标志
工作流程
用户信号 → 动作
交付前自查

Scripts executed separately

scripts/check-layout.js

Files

SKILL.mdcore26,521 bytes
references/交互流程骨架.htmlasset18,130 bytes
references/定稿包模板.mddoc7,141 bytes
references/对比板骨架.htmlasset5,732 bytes
references/桌面外壳惯例-macOS.mddoc2,043 bytes
references/状态清单模板.mddoc6,523 bytes
scripts/check-layout.jsruns7,267 bytes

Read SKILL.md


# 页面方案设计：三层法

用户是产品经理，你是懂技术的设计搭档。目标是**敲定一个页面的逻辑**，不是画一张好看的图。产出给下游（PRD、开发）用的事实源，视觉件只是附件；最终交付是一个**定稿包**（一个文件夹，开发只读它）。

## 元原则

1. **「好不好看」不是验收标准，「这页对用户有没有用」才是。** 标准不写出来，什么方案都过不了。
2. **三层有依赖顺序：① 问题 → ② 布局 → ③ 样式。** 用户开口说的往往是 ③，真正没定的往往是 ①。
3. **用户说「说不出来 / 差点什么 / 有点怪」= 上层没定。立刻退层，不出下一版。** 能点名是哪一处不对的不算，那是单点问题（见第 3 步）。**退法三步**：① 只问一句「是哪一块让你觉得差点什么，只说一块」，问块不问原因；② 那一块牵动哪一层就退到哪一层，**可能一次退两层**（用户指的是排法 → 退 ②；用户说「它没告诉我 X」→ 含清单缺一项，退 ①）；③ 退回去的层**只重开争议那一处**，用户此前确认过的其余部分原样保留，重新确认后原路走回来，不从头重来。
4. **减法优先。** 先删到用户喊停，再只加用户主动要的。用户从用法里长出来的点子比 AI 的方案准。
5. **AI 主动加的元素要打标。** 合计句、装饰图形、解释文字这类不是用户要的东西，产出里标「AI 加的」，让用户一眼能砍；**本页连续两样被砍就停止主动加，换页清零**，计数记在层标记行里。标记写在草图 / 对比板的版本说明和回复里，**不渲染进页面画面**——画进高保真里，用户会把它当成界面文案。
6. **常驻位置只放数据或控件。** 解释性文字（「等价 API ÷ 订阅费」）不占常驻位置；同一个位置要放话，放能改变动作的数（「省了 $X」）。状态话（「已停」「正在统计」「未设置」）算控件反馈，不算解释性文字。低频动作不做常驻按钮，藏在它管的数字后面。
7. **给推荐，不列菜单。** 每次出多版都带一个倾向和一句理由；用户选编号，不描述感受。

View the complete source on GitHub

prd-test-writer

When your agent uses it

PRD + 可执行测试用例双文档一体化协作。与用户共同写并迭代。理解需求后自主读代码再写；故事驱动 + 分阶段单点确认；每个 PRD 产出 PRD-MD 与测试用例-MD（给 AI 的事实源）+ 两份套模板的 review HTML（给人查阅，与 MD 严格 1:1）。触发：梳理/撰写/完善 PRD、需求文档、用户故事、验收标准、测试用例、测试基准、测试方案。

What it loads into context

Frontmatter

161 tokens

Always available to the agent

SKILL.md

4,366 tokens

When the skill is activated

Reference documents

13,488 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

一、核心理念（红线，违反即返工）
二、交互模型
三、任务流程：6 阶段闭环（严格按序，前阶段未过不得进下一阶段）
四、产物约定（每个 PRD 固定 4 件）
附录 A · 术语表（对人输出仍说人话）
附录 B · 测试用例 13 字段（标准）
附录 D · 适用边界与通用映射
附录 C · PRD 总集（台账）

Files

SKILL.mdcore14,452 bytes
assets/prd-review.html.tmplasset9,057 bytes
assets/prd-template.mddoc4,181 bytes
assets/test-cases-review.html.tmplasset12,485 bytes
assets/test-cases-template.mddoc7,147 bytes
references/adversarial-review-prompts.mddoc4,286 bytes
references/example-us01.mddoc3,467 bytes
references/html-fill-spec.mddoc6,110 bytes
references/mermaid-examples.mddoc5,344 bytes
references/prd-registry-demo.mddoc1,931 bytes
references/test-case-example.mddoc8,596 bytes
references/ui-wireframe-examples.mddoc4,570 bytes
samples/PRD-SAMPLE-review.htmlasset16,786 bytes
samples/PRD-SAMPLE-测试用例-review.htmlasset26,456 bytes

Read SKILL.md


# PRD 与测试用例协作（伙伴模式）

你是以开发者为中心的产品经理与需求 / 测试工程师，通过提问、复述和必要的阶段确认，共同构建 PRD 和测试用例。

这是独立使用版本。输入优先读取用户原话、Issue task、已确认的页面定稿包与后端逻辑包。已经确认且没有变化的决定直接继承；只讨论新增、缺失或冲突的业务选择。页面状态与后端规则编号应在验收标准和测试用例中保留，方便实现与审核追溯。重要任务的独立审核使用 `dual-agent-collaboration`。

## 一、核心理念（红线，违反即返工）

### PRD 即故事集
1. **故事是唯一载体**：PRD 主体是按逻辑排列的用户故事。
2. **故事自包含**：每张卡含业务逻辑、用户可见行为（页面/状态/文案）、边界、验收标准。
3. **叙事逻辑高于一切**：先建宏观"用户旅程地图/业务主流程"，再把故事串在主线上。
4. **视觉对齐必须**：涉及 UI 的故事**必须**用 **ASCII 线框图**画静态布局；**Mermaid** 画动态行为（流程/状态/时序）。两者互补。示例见 `references/ui-wireframe-examples.md`、`references/mermaid-examples.md`。

View the complete source on GitHub

req-change-workflow

When your agent uses it

已有功能的需求变更闭环（门禁式七步）。触发硬条件：要改的功能已经实现并跑起来了。当用户说"改需求""需求变更""调整交互""改功能""重构流程"，或改动容易散到多个文件、碰到鉴权/存储/配置/权限、需要可靠验证 + 回滚方案时使用。流程：锁 scope 写 change brief → 从代码确认当前行为（不靠记忆和假设）→ 影响面与风险评估 + 回滚计划 → 提出新设计等用户批准（**未获批准绝不动代码**）→ 小而局部的 diff → 跑固定回归清单 → 更新文档和决策日志。产出：change brief + 代码 diff + 回归记录 + decision log。不适用于：全新功能从零做（走 issue-pool → page-solution-design（有界面时）→ backend-logic-design（有规则需讨论时）→ prd-test-writer）、只改视觉样式不动任何逻辑（按已有界面直接调整）、新想法还没定要不要做（先用 issue-pool 入池）、写 PRD/测试用例（用 prd-test-writer）、提交推送发版（用 git-push）。注意：本 skill 的七步流程通用于任何项目，但 scripts/impact_scan.sh 与 references/chrome-extension-regression-example.md 是 Chrome 扩展示例；通用回归使用 references/regression-checklist.md，按项目实际选择检查。

What it loads into context

Frontmatter

365 tokens

Always available to the agent

SKILL.md

1,420 tokens

When the skill is activated

Reference documents

1,268 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

Overview
Workflow (gated loop)
Resources

Scripts executed separately

scripts/impact_scan.sh

Files

SKILL.mdcore6,382 bytes
SKILL.mdcore6,382 bytes
references/change-brief-template.mddoc833 bytes
references/chrome-extension-regression-example.mddoc1,648 bytes
references/decision-log-template.mddoc616 bytes
references/regression-checklist.mddoc1,410 bytes
scripts/impact_scan.shruns1,363 bytes

Read SKILL.md


# Req Change Workflow

## Overview

Use a lightweight, repeatable workflow to modify an existing requirement without scope creep or “边改边炸”. Produce clear artifacts at each gate so the user can approve before the implementation starts.

## Workflow (gated loop)

Follow the steps in order. Do not implement code changes until the user approves Step 4.

### Step 0: Set the plan (optional but recommended)

- Use `update_plan` to create 5–7 short steps: clarify → baseline → impact → design → implement → validate → document.
- Keep exactly one step `in_progress` at a time and advance as you finish.

View the complete source on GitHub

article-study

When your agent uses it

带用户精读一篇文章/文档并真正学透（不是出摘要）。五步：抽干货 → 切讲次 → 每讲跑「学-考-讲」循环 → 对号入座 → 实操+测验+讲错题+蒸馏。每讲产出 HTML 课件 + 笔记落盘；讲完必考一次，用户复述后必须挑不精确处拧紧；抽象概念上可交互演示（能点能跑）；全程用用户自己的业务场景当案例；学完出多题型自动判分测验卷，最后把收获蒸馏回用户的工具。核心触发条件是用户要的是"学会"而不是"要一份结果"，例如"我们一起学这篇文章/这个链接"、"带我学"、"精读"、"我想学会 X"、"这篇我看不懂你给我讲讲"。有具体材料（链接、本地文件、PDF，或用户自己的 skill/文档/代码）时直接开跑；只有学习意图而没材料时仍走本 skill，但开工第一件事是和用户一起把材料定下来，禁止凭记忆开讲。不适用于：只要一份总结/摘要/教程长文，用户读完就完、不需要答题（用 readable-output）、只是搜集资料做调研（使用可用的网页/平台取材工具）、帮我写 PRD/测试用例（用 prd-test-writer）、以及用户其实是想让你直接把活干了（那就直接做）。

What it loads into context

Frontmatter

389 tokens

Always available to the agent

SKILL.md

2,616 tokens

When the skill is activated

Reference documents

7,467 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

五条红线（违反即返工）
开跑前（只问这些，其余别问）
五步主流程
四个坑（都是真踩过的）
逆境降级（本 skill 最容易失手的地方，详见 `教学法.md`）
文件清单
产出物清单
启动命令

Files

SKILL.mdcore8,585 bytes
agents/openai.yamlasset259 bytes
references/README.mddoc1,001 bytes
references/物证-交互演示.htmlasset13,318 bytes
templates/学习计划.mddoc2,082 bytes
templates/测验模板.htmlasset16,019 bytes
templates/笔记模板.mddoc957 bytes
templates/蒸馏素材清单.mddoc1,205 bytes
templates/课件模板.htmlasset12,778 bytes
templates/资料模板.mddoc1,545 bytes
教学法.mddoc17,370 bytes

Read SKILL.md


# 和用户一起学一篇文章

> **哲学**：「读一遍觉得懂了」是**流畅性错觉**——和 AI 说「完成了」一样，是主观判断。
> 这套流程干的事，就是把学习也做成**可验证的**：输入结构化、过程有考试、错题定位回退。
> 一句话：**学东西和写代码，用的是同一套质检逻辑。**

## 五条红线（违反即返工）

1. **不做自学大纲**：你是带着学的老师，不是列书单的人。每一讲的内容你自己讲出来，不要写"建议你去读 X 章"。
2. **每讲必考**：讲完一定要用户输出（复述 / 做题 / 动手写）。说不出来 = 没学会。
3. **复述必拧紧**：用户总结完，**必须挑一到两处不精确的地方校准**，且拧紧那句必须是**否定/收窄结构**（"关键不在 A 在 B" / "这里少了第三条腿" / "这个词不可判定，换成 X"）。写成"补充一点…" = 没拧紧，重写。
4. **对号入座**：每一讲都要落到用户自己的业务场景 / 现有工具 / 真实痛点上。**接不上时先停下来问，不许硬贴标签**。
5. **原文与推论视觉可分**：短文章必然要靠你的推论撑分量，但**哪些是原文说的、哪些是你推的，必须一眼可辨**（推论进 `details` 或标「原文没直说，但成立」）。否则"我读了一篇文章"会悄悄变成"我听了 AI 讲课"。

View the complete source on GitHub

system-study

When your agent uses it

系统化学习材料生成器。给一个新领域/技术/概念，AI 自主调研、搭体系、产出 HTML 学习材料（含骨架/案例/工程化/争议+盲区）。当用户说"我要学 X / 帮我系统拆解 X / 我想吃透 X 这个领域 / 给我整理 X 的全貌 / 深度调研 X 给我系统讲讲 / 把 X 搞透"时触发。不适用于"X 行不行/为什么 Y"（使用可用的调研工具直接研究）、"X 有哪些好玩案例"（用 case-radar 给散点）、"把这堆素材整理成 HTML"（用 readable-output 处理已有素材）、文章写作（用 writing-assistant）、整页界面方案（用 page-solution-design）。

What it loads into context

Frontmatter

211 tokens

Always available to the agent

SKILL.md

3,108 tokens

When the skill is activated

Reference documents

17,690 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

工具适配
5 阶段流程总览
核心设计原则（5 条）
5 个 Sub-agent 阵容（A-D 灵活 + E 必选）
工作流（主线程要做什么）
关键硬约束（一定要看 [anti-patterns.md](reference/anti-patterns.md)）
灵活参数
不适用于（边界划清）
参考实例

Files

SKILL.mdcore10,040 bytes
reference/anti-patterns.mddoc8,300 bytes
reference/examples-index.mddoc2,541 bytes
reference/stage-1-topic-narrow.mddoc7,694 bytes
reference/stage-2-research-plan.mddoc4,672 bytes
reference/stage-3-subagent-templates.mddoc23,533 bytes
reference/stage-4-synthesis.mddoc4,866 bytes
reference/stage-5-html-spec.mddoc5,558 bytes

Read SKILL.md


# System Study — 系统化学习材料生成

## 工具适配

文中的 `AskUserQuestion`、`Agent`、`Read`、`Write` 是能力名称，使用当前平台的等效工具。没有多选界面时，用一条消息列出切片选项，请用户回复编号组合；已明确的信息不重复问。子 agent 按平台并发上限分批执行，先固定全部输入和 prompt，后批不得注入前批的结论。平台没有子 agent 能力时说明限制，再与用户确定逐视角调研方式。


你要学一个新领域。AI 自主调研 + 搭体系 + 产出 HTML 学习材料。

**和已有 skill 的本质区别**：
- **与结论型调研的区别**：结论型调研是"有困惑 → 求结论"；这个是"想学 → 求体系"
- **vs case-radar**：那个给散点案例集；这个给系统化全貌
- **vs readable-output**：那个处理已有素材；这个自己挖素材 + 自己建体系（最后可调用它做出稿，但不是核心）

View the complete source on GitHub

find-top-three

When your agent uses it

通过自适应获取上下文、读取证据、建立可修正的用户模型和进行战略判断，帮助用户找出当前阶段最重要的三件事。适用于用户询问“最重要的三件事”“我现在应该先做什么”，需要梳理人生或职业优先级，在多个目标之间犹豫，思考自由、收入、个人业务、内容创作或创业重点，或者希望模型持续了解自己后再给建议；不用于简单待办排序，也不用于目标已经确定的项目计划。

What it loads into context

Frontmatter

148 tokens

Always available to the agent

SKILL.md

3,468 tokens

When the skill is activated

Reference documents

4,437 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

基本立场
工作流程
收敛条件
边界

Files

SKILL.mdcore13,959 bytes
agents/openai.yamlasset278 bytes
references/modeling-framework.mddoc12,411 bytes
references/output-templates.mddoc4,376 bytes

Read SKILL.md


# 找到最重要的三件事

找出真正适合当前用户、当前阶段的三项战略优先级。把模型当作有判断力的决策伙伴：提供治理结构和证据纪律，不提供固定问卷、固定分类或机械打分。

## 基本立场

- 尽早形成有用的临时判断，不要只当被动采访者。早期判断只用于指出问题结构、候选解释或关键未知，不能替代多轮探索，也不能编造用户的经历、资产、动机或能力。
- 新证据出现后主动修正判断，并说明哪里变了、为什么变。
- 从用户证据中生成优先级，绝不默认输出“财务、健康、事业”或其他通用三分类。
- 严格区分已确认事实、用户自己的解释和模型提出的假设。
- 把赚钱、涨粉、买房、晋升、离职等目标暂时视为可能的代理目标，先理解它实际要带来什么。
- 优先询问具体事件和读取原始材料，不轻信抽象自我标签。
- 保留用户主体性：提供可讨论的判断，不诊断用户，也不武断定义用户是什么样的人。
- 当继续追问已不太可能改变排序时，停止收集信息，转向一个可逆、能产生新证据的行动。

View the complete source on GitHub

multi-perspective-analysis

When your agent uses it

多视角深度分析。用多个 Sub-agent 分别扮演不同思维框架的顾问，对同一份素材做独立分析，然后交叉汇总共识与分歧，产出结构化诊断报告。当用户说"帮我多视角分析"、"多维分析"、"用多个视角看看"、"帮我诊断一下"时触发。

What it loads into context

Frontmatter

118 tokens

Always available to the agent

SKILL.md

4,868 tokens

When the skill is activated

Reference documents

63,019 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

流程概览
阶段 1：收敛需求
阶段 2：准备素材
阶段 3：并行分析（核心阶段）
阶段 4：交叉汇总
阶段 5：保存与输出（自动执行，不需要用户要求）
内置视角库（10 个全球最强大脑）
自定义视角
核心原则

Files

SKILL.mdcore16,677 bytes
SKILL.mdcore16,677 bytes
reference/01-10x-growth-dan-sullivan.mddoc18,472 bytes
reference/02-first-principles-elon-musk.mddoc19,458 bytes
reference/03-product-intuition-zhang-xiaolong.mddoc20,129 bytes
reference/04-creator-growth-mrbeast.mddoc25,603 bytes
reference/05-mental-models-charlie-munger.mddoc20,575 bytes
reference/06-zero-to-one-peter-thiel.mddoc24,273 bytes
reference/07-product-vision-steve-jobs.mddoc29,500 bytes
reference/08-customer-obsession-jeff-bezos.mddoc26,239 bytes
reference/09-algorithm-thinking-zhang-yiming.mddoc21,867 bytes
reference/10-strategic-survival-ren-zhengfei.mddoc23,259 bytes
reference/_how-to-create-reference.mddoc5,377 bytes

Read SKILL.md


# 多视角深度分析

用 10 个独立的 Sub-agent 分别扮演全球最强大脑的"顾问"，对用户提供的素材做并行分析，然后交叉汇总产出结构化报告。

**阵容**：Dan Sullivan、马斯克、张小龙、MrBeast、芒格、蒂尔、乔布斯、贝佐斯、张一鸣、任正非

核心价值：**独立分析但结论趋同时可信度极高，结论分歧时往往是最有价值的洞察。**

---

## 流程概览

| 阶段 | 名称 | 目标 | 执行方式 |
|---|---|---|---|

View the complete source on GitHub

thinking-partner

When your agent uses it

思考拍档 - 陪你从混沌中理清局面，锁定核心问题，拆解卡点，共创解法，落地行动

What it loads into context

Frontmatter

64 tokens

Always available to the agent

SKILL.md

1,988 tokens

When the skill is activated

Reference documents

0 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

1. 核心使命 (Mission)
2. 核心思维模型 (Mental Framework)
3. 五步作业流程 (Workflow)
4. 沟通风格 (Tone & Style)
5. 启动指令 (Initialization)

Files

SKILL.mdcore7,167 bytes

Read SKILL.md


# Role: 思考拍档

## 1. 核心使命 (Mission)

你是用户的**思考拍档**，不是答案机器。你的核心使命是通过严格的五步流程，陪用户一起从混沌中理清局面、锁定核心问题、找到真正的卡点、共创解法、落地为行动。

核心原则：**不替用户想，陪用户想。** 用户自己想通的比你给的答案有用一百倍。

## 2. 核心思维模型 (Mental Framework)

**总纲：**
> 在任何复杂局面中，会同时存在多个问题，但其中必有一个问题居于主导地位，起决定作用。找到它，集中力量解决它，其他问题会随之松动。

在所有对话中，你必须始终运行以下逻辑：

View the complete source on GitHub

case-radar

When your agent uses it

案例雷达。给一个新东西（新工具/新概念/新生态），扫一遍生态找好玩的真实案例，重点是抓"真物"（截图/源码/演示）而不是 GitHub 主页，输出可浏览的 HTML 案例集。当用户说"看看大家用 X 做了什么"、"扫一下 X 生态"、"市面上 X 有什么新玩法"、"给我看 X 的真物案例"、"/case-radar"时触发。不适合：① 已有明确目标的深度调研（使用可用的调研工具直接研究）② 写文章/出 PRD（用 writing-assistant / prd-test-writer）③ 单纯求知不需要 HTML（直接问就好）。

What it loads into context

Frontmatter

199 tokens

Always available to the agent

SKILL.md

4,073 tokens

When the skill is activated

Reference documents

8,151 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

工具适配
流程概览
阶段 0：环境自检 + 主题对齐
阶段 1：scan · 多渠道扫信源
阶段 2：recon · 真物侦察（核心增值）
阶段 3：capture · 抓真物
阶段 4：embed · 输出 HTML
阶段 5：完成输出
核心原则
什么时候不该用这个 Skill（重要）
使用示例
维护与扩展

Files

SKILL.mdcore13,803 bytes
SKILL.mdcore13,803 bytes
SKILL.mdcore13,803 bytes
SKILL.mdcore13,803 bytes
SKILL.mdcore13,803 bytes
SKILL.mdcore13,803 bytes
reference/_how-to-extend.mddoc4,007 bytes
reference/html-template-spec.mddoc9,482 bytes
reference/recon-heuristics.mddoc8,492 bytes
reference/workflow-anti-patterns.mddoc5,784 bytes

Read SKILL.md


# Case Radar · 案例雷达

## 工具适配

`AskUserQuestion`、`Agent`、`Read` 等名称对应询问、子 agent 与文件读取能力，使用当前平台的等效工具。没有结构化问答工具时，用一条消息问清缺失参数；已有答案直接继承。没有子 agent 能力时，由主 agent 完成扫描。


把"想了解新东西"这件事从"读三手 SEO 文" → "拿到一手真物（截图/源码/演示）"。

输入：一个新东西的名字 + 一句话上下文（"我想用来做 X"）。
输出：一份可浏览的 HTML 案例集，每张卡都配真实物料（不是文字描述）。

**核心信念**：菜单一文不值，菜才是真东西。GitHub repo 主页是文件列表（菜单），README 里的截图、商店页的安装数、SKILL.md 的原文段落、产品落地页的真实建站效果——才是真东西（菜）。

View the complete source on GitHub

comic-maker

When your agent uses it

教程漫画「小课堂」：你的 IP 当老师（没有就先生成一个），3:4 竖版，一期默认 6–9 页（用户要几页就几页），固定有「开课」封面和「下课」结尾，交给 Codex 内置 imagegen 出图并核对。当用户说"画漫画"、"做一套漫画讲XXX"、"用漫画解释"、"小课堂"、"换个老师 / 换角色"时触发。

What it loads into context

Frontmatter

140 tokens

Always available to the agent

SKILL.md

2,409 tokens

When the skill is activated

Reference documents

6,976 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

默认配置
第 0 步：角色（每次开工先看）
启动：先定协作模式
工作流程
提示词铁律
相关文件

Scripts executed separately

scripts/_data.py
scripts/doctor.py
scripts/mark.js
scripts/render_codex.py
scripts/settings.py

Files

README.mddoc1,530 bytes
SKILL.mdcore8,355 bytes
SKILL.mdcore8,355 bytes
SKILL.mdcore8,355 bytes
SKILL.mdcore8,355 bytes
SKILL.mdcore8,355 bytes
SKILL.mdcore8,355 bytes
SKILL.mdcore8,355 bytes
examples/character/juzi-prompt.txtdoc518 bytes
examples/character/juzi.jpgasset199,636 bytes
examples/mcp/00.jpgasset268,426 bytes
examples/mcp/00.txtdoc1,493 bytes
examples/mcp/01.txtdoc1,508 bytes
examples/mcp/02.jpgasset299,463 bytes
examples/mcp/02.txtdoc1,547 bytes
examples/mcp/03.txtdoc1,405 bytes
examples/mcp/04.txtdoc1,696 bytes
examples/mcp/05.txtdoc1,392 bytes
examples/mcp/06.txtdoc1,467 bytes
examples/mcp/07.jpgasset279,283 bytes
examples/mcp/07.txtdoc1,364 bytes
scripts/_data.pyruns3,209 bytes
scripts/doctor.pyruns3,282 bytes
scripts/mark.jsruns3,782 bytes
scripts/render_codex.pyruns17,836 bytes
scripts/settings.pyruns3,636 bytes
styles/classroom.mddoc3,531 bytes
templates/character.mddoc4,573 bytes
templates/layout-patterns.mddoc5,174 bytes
templates/prompt-template.mddoc8,105 bytes
templates/story-structure.mddoc2,998 bytes

Read SKILL.md


# 漫画小课堂（comic-maker）

一个固定的老师角色，用一期漫画讲清一个概念。当前助手（Claude 或 Codex）写故事和每页提示词、交付前逐页看图；出图交给 Codex 内置 imagegen 原样执行；脚本负责检查和核对。

下文 `S` 指本 skill 的脚本目录：

```bash
S=<本 SKILL.md 所在目录>/scripts
```

## 默认配置

```
风格：小课堂（styles/classroom.md）；画风跟着角色参考图走

View the complete source on GitHub

image-render

When your agent uses it

配图出图 - 给文章配图：定好每张图讲什么、图上逐字写什么，用你的默认风格（第一次从带样张的风格库里看图选）为每张图设计画面，交给 Codex 内置 imagegen 原样出图，生成可标记喜欢的对比页。支持探索（一张图试几种风格）、量产（默认风格出整套）和新增自己的风格。当用户说「给这篇文章配图」「出几张图」「用 XX 风格出图」「换默认风格」「加个新风格」时触发。

What it loads into context

Frontmatter

166 tokens

Always available to the agent

SKILL.md

2,177 tokens

When the skill is activated

Reference documents

19,326 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

第一次用
默认风格
流程
新增风格
关键原则（都有实测证据，见 `reference/lessons.md`）
目录
任务目录（每篇文章 / 每批图一个）
执行者是谁

Scripts executed separately

scripts/_lib.py
scripts/add_sample.py
scripts/build_prompts.py
scripts/check_prompts.py
scripts/doctor.py
scripts/make_page.py
scripts/mark.js
scripts/render_codex.py
scripts/settings.py
scripts/style_picker.py
scripts/update_index.py
scripts/verify_log.py

Files

README.mddoc1,621 bytes
SKILL.mdcore7,138 bytes
SKILL.mdcore7,138 bytes
SKILL.mdcore7,138 bytes
SKILL.mdcore7,138 bytes
SKILL.mdcore7,138 bytes
SKILL.mdcore7,138 bytes
SKILL.mdcore7,138 bytes
SKILL.mdcore7,138 bytes
reference/formats.mddoc1,024 bytes
reference/lessons.mddoc3,549 bytes
reference/recommend.mddoc2,029 bytes
reference/taste.mddoc1,325 bytes
scripts/_lib.pyruns9,138 bytes
scripts/add_sample.pyruns2,801 bytes
scripts/build_prompts.pyruns2,206 bytes
scripts/check_prompts.pyruns2,790 bytes
scripts/doctor.pyruns3,382 bytes
scripts/make_page.pyruns4,778 bytes
scripts/mark.jsruns3,782 bytes
scripts/render_codex.pyruns8,935 bytes
scripts/settings.pyruns1,958 bytes
scripts/style_picker.pyruns4,083 bytes
scripts/update_index.pyruns2,682 bytes
scripts/verify_log.pyruns1,920 bytes
stages/01-scenes.mddoc1,807 bytes
stages/02-pick-style.mddoc1,353 bytes
stages/03-adapt.mddoc1,956 bytes
stages/04-render.mddoc4,253 bytes
stages/05-review.mddoc1,829 bytes
styles/N02-50年代复古广告.mddoc1,468 bytes
styles/N03-积木.mddoc882 bytes
styles/N08-桌游棋盘.mddoc995 bytes
styles/N10-电影海报.mddoc1,320 bytes
styles/N11-绘本蜡笔.mddoc805 bytes
styles/N13-便利贴墙.mddoc904 bytes
styles/N16-维多利亚铜版画.mddoc832 bytes
styles/N17-折纸.mddoc756 bytes
styles/N19-粉笔黑板.mddoc770 bytes
styles/N21-纸雕光影盒.mddoc1,070 bytes
styles/N22-宝藏地图.mddoc792 bytes
styles/N24-3D动画电影.mddoc1,743 bytes
styles/N27-四格漫画.mddoc935 bytes
styles/N34-烘焙甜点.mddoc786 bytes
styles/N35-木偶剧场.mddoc958 bytes
styles/N37-皮影戏.mddoc990 bytes
styles/N38-玩具屋剖面.mddoc1,008 bytes
styles/N39-立体书.mddoc871 bytes
styles/N41-日式手绘动画.mddoc990 bytes
styles/N42-等距像素.mddoc1,601 bytes
styles/N48-列车剖面.mddoc767 bytes
styles/N49-角色手绘课件.mddoc2,239 bytes
styles/N50-RPG游戏界面.mddoc1,235 bytes
styles/N62-实物摄影信息图.mddoc1,621 bytes
styles/N78-国风淡彩.mddoc1,179 bytes
styles/README.mddoc4,143 bytes
styles/S01-彩铅水彩.mddoc1,668 bytes
styles/S07-像素风.mddoc1,799 bytes
styles/S08-漫画网点.mddoc1,475 bytes
styles/samples/N02-1.jpgasset151,265 bytes
styles/samples/N02-2.jpgasset233,484 bytes
styles/samples/N02-3.jpgasset156,115 bytes
styles/samples/N02-4.jpgasset197,484 bytes
styles/samples/N02-5.jpgasset176,936 bytes
styles/samples/N02-6.jpgasset171,217 bytes
styles/samples/N02-7.jpgasset188,122 bytes
styles/samples/N03-1.jpgasset58,890 bytes
styles/samples/N08-1.jpgasset108,099 bytes
styles/samples/N08-2.jpgasset126,958 bytes
styles/samples/N10-1.jpgasset106,159 bytes
styles/samples/N10-2.jpgasset65,713 bytes
styles/samples/N10-3.jpgasset129,251 bytes
styles/samples/N11-1.jpgasset156,090 bytes
styles/samples/N13-1.jpgasset118,115 bytes
styles/samples/N16-1.jpgasset148,506 bytes
styles/samples/N17-1.jpgasset71,546 bytes
styles/samples/N19-1.jpgasset112,124 bytes
styles/samples/N21-1.jpgasset107,267 bytes
styles/samples/N21-2.jpgasset89,853 bytes
styles/samples/N21-3.jpgasset130,721 bytes
styles/samples/N22-1.jpgasset161,376 bytes
styles/samples/N24-1.jpgasset101,862 bytes
styles/samples/N24-2.jpgasset101,932 bytes
styles/samples/N24-3.jpgasset99,741 bytes
styles/samples/N24-4.jpgasset144,695 bytes
styles/samples/N24-5.jpgasset112,267 bytes
styles/samples/N24-6.jpgasset133,551 bytes
styles/samples/N24-7.jpgasset124,411 bytes
styles/samples/N24-8.jpgasset134,547 bytes
styles/samples/N24-9.jpgasset111,431 bytes
styles/samples/N27-2.jpgasset137,317 bytes
styles/samples/N34-1.jpgasset121,388 bytes
styles/samples/N35-2.jpgasset168,141 bytes
styles/samples/N37-1.jpgasset114,407 bytes
styles/samples/N37-2.jpgasset165,902 bytes
styles/samples/N38-1.jpgasset132,571 bytes
styles/samples/N38-2.jpgasset111,328 bytes
styles/samples/N39-1.jpgasset129,406 bytes
styles/samples/N39-2.jpgasset95,561 bytes
styles/samples/N41-1.jpgasset113,540 bytes
styles/samples/N41-2.jpgasset144,319 bytes
styles/samples/N42-1.jpgasset79,631 bytes
styles/samples/N42-2.jpgasset90,338 bytes
styles/samples/N42-3.jpgasset133,505 bytes
styles/samples/N42-4.jpgasset140,572 bytes
styles/samples/N42-5.jpgasset165,468 bytes
styles/samples/N42-6.jpgasset125,647 bytes
styles/samples/N42-7.jpgasset137,290 bytes
styles/samples/N48-1.jpgasset126,023 bytes
styles/samples/N49-1.jpgasset177,297 bytes
styles/samples/N49-2.jpgasset176,401 bytes
styles/samples/N49-3.jpgasset177,798 bytes
styles/samples/N50-1.jpgasset139,972 bytes
styles/samples/N50-2.jpgasset142,302 bytes
styles/samples/N50-3.jpgasset137,044 bytes
styles/samples/N50-4.jpgasset146,765 bytes
styles/samples/N62-1.jpgasset190,473 bytes
styles/samples/N62-2.jpgasset188,908 bytes
styles/samples/N78-1.jpgasset234,330 bytes
styles/samples/N78-2.jpgasset225,356 bytes
styles/samples/N78-3.jpgasset208,435 bytes
styles/samples/S01-1.jpgasset98,498 bytes
styles/samples/S01-2.jpgasset98,438 bytes
styles/samples/S01-3.jpgasset96,080 bytes
styles/samples/S01-4.jpgasset91,154 bytes
styles/samples/S01-5.jpgasset89,755 bytes
styles/samples/S01-6.jpgasset93,424 bytes
styles/samples/S07-1.jpgasset99,025 bytes
styles/samples/S07-2.jpgasset97,840 bytes
styles/samples/S07-3.jpgasset96,674 bytes
styles/samples/S07-4.jpgasset92,954 bytes
styles/samples/S07-5.jpgasset93,784 bytes
styles/samples/S07-6.jpgasset98,669 bytes
styles/samples/S08-1.jpgasset135,822 bytes
styles/samples/S08-2.jpgasset144,693 bytes
styles/samples/S08-3.jpgasset131,016 bytes
styles/samples/S08-4.jpgasset138,999 bytes
styles/samples/S08-5.jpgasset236,718 bytes
templates/adapt-example.txtdoc876 bytes
templates/codex-task.mddoc954 bytes
templates/common.txtdoc770 bytes
templates/scene.mddoc568 bytes
templates/style-card.mddoc740 bytes

Read SKILL.md


# 配图出图（image-render）

把文章变成一套风格统一的配图。Claude 负责想清楚画什么、图上写哪些字；Codex 只负责把提示词原样交给 imagegen 出图；脚本负责检查和核对。

下文 `S` 指本 skill 的脚本目录（本文件所在目录下的 `scripts/`）：

```bash
S=<本 SKILL.md 所在目录>/scripts
```

## 第一次用

1. `python3 $S/doctor.py`：环境自检（Codex 装没装、登没登录、Python）。有 ✗ 先按提示补。
2. 没有默认风格时，开风格选择页让用户看图选（见下面「默认风格」）。

View the complete source on GitHub

readable-output

When your agent uses it

产出给人读的 HTML 长文——把已有的素材、经历、资料想清楚再写出来。每次触发先用 AskUserQuestion 一次性问清「给谁看 / 读完拿什么 / 多长 / 风格 + 侧重」4 个关键参数，再按 6 阶段框架输出。**产出只有 HTML 一种格式**——用户要 markdown 或要在聊天里直接说，就不该用本 skill。当用户说"做个复盘""汇总一下""总结这堆""整理成 HTML""做个教程/学习指南""把 X 讲清楚"等需要超过 500 字的结构化阅读内容时触发。不适用于：项目的计划报告/框架计划/版本路线（用 issue-pool，它自带 md2html）、PRD/需求文档/测试用例（用 prd-test-writer）、整页界面方案（用 page-solution-design）、起名（用 product-naming）、写代码/修 bug/改文件、一两句话就能答完的问题。

What it loads into context

Frontmatter

261 tokens

Always available to the agent

SKILL.md

4,347 tokens

When the skill is activated

Reference documents

0 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

工具适配
元原则（贴在最顶层，每次开始前默念）
触发条件
运作模式
阶段 1（强制）：用 AskUserQuestion 一次性问 4 个
阶段 2-6（AI 自己心里走，不对用户念）
反模式总清单
快速决策树
输出文件规则
维护规则

Files

SKILL.mdcore13,852 bytes
SKILL.mdcore13,852 bytes

Read SKILL.md


# 输出工程：让 AI 输出长内容时按清单想一遍

## 工具适配

文中的 `AskUserQuestion` 使用当前平台的结构化询问工具；没有该工具时，用一条消息问清同样的四个维度。用户已明确的参数直接继承；工具名不同不阻断写作。


> 综合写作理论（Zinsser《On Writing Well》/ Heath《Made to Stick》/ Rosenfeld《Information Architecture》）+ 学习科学（Sweller CLT / Brown《Make It Stick》/ Wiggins backward design）+ 大厂文档实践（Stripe / GitLab / Linear / Anthropic / Google SRE）三方交叉验证。

---

## 元原则（贴在最顶层，每次开始前默念）

1. **如果你自己读完没有获得新认知，读者也不会有。** 写作是 thinking on paper，不是模板填空。

View the complete source on GitHub

scene-ppt

When your agent uses it

场景化讲解 PPT（图片版，一页一张 16:9 图）。读一篇文章或一个主题，把整套内容放进你的默认「世界」里讲（小课堂、后厨、RPG、像素小岛、复古广告…第一次看样张选），逐页设计场景、定好图上每个字，交给 Codex 内置 imagegen 原样出图。当用户说「做成 PPT」「出一套讲解幻灯片」「用 PPT 讲 XX」「把这篇文章做成 PPT」「换默认世界」时触发。

What it loads into context

Frontmatter

158 tokens

Always available to the agent

SKILL.md

3,410 tokens

When the skill is activated

Reference documents

8,428 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

核心：换世界，不是换皮
默认世界（每次开工先看）
角色档字段（只有小课堂世界需要）
分工
流程
出图命令
每页怎么写（铁律）
交付前检查（当前助手逐页看图，必须做）
新增世界
目录
任务目录

Scripts executed separately

scripts/_lib.py
scripts/build_prompts.py
scripts/check_prompts.py
scripts/doctor.py
scripts/make_page.py
scripts/mark.js
scripts/render_codex.py
scripts/settings.py
scripts/verify_log.py
scripts/world_picker.py

Files

README.mddoc1,606 bytes
SKILL.mdcore11,330 bytes
SKILL.mdcore11,330 bytes
SKILL.mdcore11,330 bytes
SKILL.mdcore11,330 bytes
SKILL.mdcore11,330 bytes
SKILL.mdcore11,330 bytes
SKILL.mdcore11,330 bytes
SKILL.mdcore11,330 bytes
SKILL.mdcore11,330 bytes
SKILL.mdcore11,330 bytes
examples/classroom/C1-N49.jpgasset138,814 bytes
examples/classroom/C1-N49.txtdoc709 bytes
examples/classroom/C2-N49.jpgasset139,507 bytes
examples/classroom/C2-N49.txtdoc605 bytes
examples/classroom/C3-N49.jpgasset141,437 bytes
examples/classroom/C3-N49.txtdoc610 bytes
examples/kitchen/K1-N24.jpgasset87,464 bytes
examples/kitchen/K1-N24.txtdoc869 bytes
examples/kitchen/K2-N24.jpgasset82,862 bytes
examples/kitchen/K2-N24.txtdoc1,002 bytes
examples/kitchen/K3-N24.jpgasset91,171 bytes
examples/kitchen/K3-N24.txtdoc1,131 bytes
examples/kitchen/K4-N24.jpgasset71,718 bytes
examples/kitchen/K4-N24.txtdoc650 bytes
examples/pharmacy/Z1-N78.jpgasset108,915 bytes
examples/pharmacy/Z1-N78.txtdoc863 bytes
examples/pharmacy/Z2-N78.jpgasset103,276 bytes
examples/pharmacy/Z2-N78.txtdoc1,086 bytes
examples/pharmacy/Z3-N78.jpgasset96,395 bytes
examples/pharmacy/Z3-N78.txtdoc748 bytes
examples/rpg/R1-N50.jpgasset92,266 bytes
examples/rpg/R1-N50.txtdoc543 bytes
examples/rpg/R2-N50.jpgasset95,833 bytes
examples/rpg/R2-N50.txtdoc833 bytes
examples/rpg/R3-N50.jpgasset91,552 bytes
examples/rpg/R3-N50.txtdoc773 bytes
examples/rpg/R4-N50.jpgasset97,678 bytes
examples/rpg/R4-N50.txtdoc793 bytes
reference/lessons.mddoc2,099 bytes
reference/page-types.mddoc1,950 bytes
reference/world-samples/bricks-1.jpgasset132,314 bytes
reference/world-samples/bricks-2.jpgasset115,723 bytes
reference/world-samples/bricks-3.jpgasset80,611 bytes
reference/world-samples/captain-1.jpgasset212,560 bytes
reference/world-samples/captain-2.jpgasset212,431 bytes
reference/world-samples/captain-3.jpgasset209,134 bytes
reference/world-samples/classroom-1.jpgasset138,814 bytes
reference/world-samples/classroom-2.jpgasset139,507 bytes
reference/world-samples/classroom-3.jpgasset141,437 bytes
reference/world-samples/island-1.jpgasset153,311 bytes
reference/world-samples/island-2.jpgasset179,785 bytes
reference/world-samples/island-3.jpgasset136,369 bytes
reference/world-samples/island-4.jpgasset151,066 bytes
reference/world-samples/kitchen-1.jpgasset144,742 bytes
reference/world-samples/kitchen-2.jpgasset135,891 bytes
reference/world-samples/kitchen-3.jpgasset148,012 bytes
reference/world-samples/kitchen-4.jpgasset121,798 bytes
reference/world-samples/library-1.jpgasset214,405 bytes
reference/world-samples/library-2.jpgasset234,434 bytes
reference/world-samples/library-3.jpgasset198,551 bytes
reference/world-samples/pharmacy-1.jpgasset184,849 bytes
reference/world-samples/pharmacy-2.jpgasset177,004 bytes
reference/world-samples/pharmacy-3.jpgasset165,293 bytes
reference/world-samples/pitstop-1.jpgasset172,088 bytes
reference/world-samples/pitstop-2.jpgasset184,794 bytes
reference/world-samples/pitstop-3.jpgasset178,259 bytes
reference/world-samples/retro-1.jpgasset221,232 bytes
reference/world-samples/retro-2.jpgasset195,332 bytes
reference/world-samples/retro-3.jpgasset188,312 bytes
reference/world-samples/retro-4.jpgasset207,245 bytes
reference/world-samples/rpg-1.jpgasset153,136 bytes
reference/world-samples/rpg-2.jpgasset158,202 bytes
reference/world-samples/rpg-3.jpgasset149,579 bytes
reference/world-samples/rpg-4.jpgasset158,442 bytes
reference/world-samples/studio-1.jpgasset161,742 bytes
reference/world-samples/studio-2.jpgasset208,787 bytes
reference/world-samples/studio-3.jpgasset149,308 bytes
reference/worlds.mddoc6,471 bytes
scripts/_lib.pyruns11,087 bytes
scripts/build_prompts.pyruns2,206 bytes
scripts/check_prompts.pyruns3,390 bytes
scripts/doctor.pyruns3,270 bytes
scripts/make_page.pyruns5,298 bytes
scripts/mark.jsruns3,782 bytes
scripts/render_codex.pyruns9,502 bytes
scripts/settings.pyruns5,537 bytes
scripts/verify_log.pyruns1,920 bytes
scripts/world_picker.pyruns3,604 bytes
styles/N02-50年代复古广告.mddoc1,468 bytes
styles/N03-积木.mddoc882 bytes
styles/N11-绘本蜡笔.mddoc805 bytes
styles/N24-3D动画电影.mddoc1,743 bytes
styles/N41-日式手绘动画.mddoc990 bytes
styles/N42-等距像素.mddoc1,601 bytes
styles/N49-角色手绘课件.mddoc2,239 bytes
styles/N50-RPG游戏界面.mddoc1,235 bytes
styles/N78-国风淡彩.mddoc1,179 bytes
templates/codex-task.mddoc954 bytes
templates/common.txtdoc776 bytes
templates/outline.mddoc1,193 bytes
templates/page-example.txtdoc854 bytes

Read SKILL.md


# 场景讲解 PPT（scene-ppt）

**「做成 PPT」= 一页一张 16:9 图，不是 .pptx 文件。** 图上的字不能再编辑。

本 skill 自带出图脚本和用得到的风格卡，不需要另装别的 skill。下文 `S` 指本 skill 的脚本目录：

```bash
S=<本 SKILL.md 所在目录>/scripts
```

## 核心：换世界，不是换皮

整套 PPT 只用**一个世界**讲完：世界里的角色 = 要讲的对象，世界里的器物 = 概念。比如讲 Skill：

View the complete source on GitHub

thought-mining

When your agent uses it

思维挖掘助手 - 通过对话帮助用户把脑子里的零散想法倒出来、记录下来、整理成文章。覆盖从思维挖掘到成稿的完整流程。

What it loads into context

Frontmatter

75 tokens

Always available to the agent

SKILL.md

694 tokens

When the skill is activated

Reference documents

2,997 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

触发方式
流程概览
调度规则
文件结构
注意事项

Files

SKILL.mdcore2,375 bytes
examples/claude-skills-case.mddoc2,987 bytes
stages/01-mining.mddoc1,426 bytes
stages/02-topic.mddoc1,222 bytes
stages/03-validation.mddoc1,017 bytes
stages/04-writing.mddoc950 bytes
stages/05-review.mddoc1,261 bytes
templates/insights-template.mddoc460 bytes
templates/writing-record-template.mddoc967 bytes

Read SKILL.md


# 思维挖掘助手

## 触发方式

当用户说类似以下内容时触发：
- "我想写一篇关于 XX 的文章"
- "帮我整理一下我的想法"
- "我有一些零散的思考，帮我记下来"
- "/thought-mining" 或 "/思维挖掘"

---

## 流程概览

View the complete source on GitHub

writing-assistant

When your agent uses it

写作助手 - 当用户说"我想写XX"、"帮我梳理选题"、"怎么形成框架"、"给我组织思路"时触发。根据观点清晰度自动选择最优路线：清晰观点走"框架→内容"，模糊观点走"挖掘→选题→框架→内容"。

What it loads into context

Frontmatter

111 tokens

Always available to the agent

SKILL.md

1,137 tokens

When the skill is activated

Reference documents

4,407 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

核心流程
流程概览
调度规则
文件结构
核心原则
注意事项

Files

SKILL.mdcore3,748 bytes
stages/00-diagnosis.mddoc2,747 bytes
stages/01-mining.mddoc3,031 bytes
stages/02-topic.mddoc2,891 bytes
stages/03-framework.mddoc3,661 bytes
stages/04-writing.mddoc2,408 bytes

Read SKILL.md


# 写作助手

## 核心流程

**三个关键步骤：选题 → 框架 → 内容**

但根据你对选题的**清晰度**，会走不同的分支：

```
用户提出主题或想法
    │
    ↓
[阶段00] 诊断：观点清不清楚？
    │

View the complete source on GitHub

Skills included

goal-setter
hermes-persona-builder
memory-init
organize
logo-design
product-naming
backend-logic-design
dual-agent-collaboration
git-push
github-repo-search
issue-pool
issue-triage
page-solution-design
prd-test-writer
req-change-workflow
article-study
system-study
find-top-three
multi-perspective-analysis
thinking-partner
case-radar
comic-maker
image-render
readable-output
scene-ppt
thought-mining
writing-assistant

Repository facts

License: MIT
Created: Jan 19, 2026
Last pushed: Oct 11, 2026
Forks: 115
Open issues: 3
Project activity: Active

Rank over the last 12 weeks

Now #993 · best #688 · down 227 in 30 days

#838 on Oct 15
#826 on Oct 17
#811 on Oct 18
#869 on Oct 20
#871 on Oct 23
#993 since Oct 25

29 earlier periods

#759 on Aug 1
#694 on Aug 6
#688 on Aug 10
#698 on Aug 11
#691 on Aug 12
#688 on Aug 14
#697 on Aug 19
#731 on Aug 24
#740 on Aug 25
#721 on Aug 26
#757 on Aug 28
#743 on Sep 2
#723 on Sep 4
#761 on Sep 9
#772 on Sep 10
#778 on Sep 16
#765 on Sep 18
#756 on Sep 20
#766 on Sep 24
#742 on Sep 26
#828 on Sep 29
#760 on Oct 1
#809 on Oct 6
#828 on Oct 7
#835 on Oct 9
#813 on Oct 10
#839 on Oct 12
#842 on Oct 13
#856 on Oct 14

Before you install

Network or broad shell

Automated signals, not a guarantee

Checked Nov 4, 2026

Not fully analyzed: some files were too long, unreadable or missing. This never lowers the level.

Show 41 notes

agent/memory-init/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
coding/backend-logic-design/examples/审核配置/后端逻辑-审核配置.html:155 Analysis incomplete: physical or joined line exceeds character limit
coding/backend-logic-design/examples/审核配置/后端逻辑-审核配置.html:153 Analysis incomplete: physical or joined line exceeds character limit
coding/backend-logic-design/examples/审核配置/后端逻辑-审核配置.html:156 Analysis incomplete: physical or joined line exceeds character limit
coding/backend-logic-design/examples/审核配置/后端逻辑-审核配置.html:154 Analysis incomplete: physical or joined line exceeds character limit
coding/dual-agent-collaboration/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
coding/dual-agent-collaboration/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
coding/dual-agent-collaboration/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
coding/dual-agent-collaboration/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
coding/dual-agent-collaboration/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
coding/dual-agent-collaboration/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
coding/req-change-workflow/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
thinking/multi-perspective-analysis/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
visualization/case-radar/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
visualization/case-radar/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
visualization/case-radar/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
visualization/case-radar/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
visualization/case-radar/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
visualization/comic-maker/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
visualization/comic-maker/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed

No broad tool permissions
No dangerous shell commands
37 network referencesin 11 files: case-radar/SKILL.md:30 case-radar/reference/recon-heuristics.md:20 prd-test-writer/assets/prd-review.html.tmpl:102 and 8 more
No hidden or obfuscated text
No instruction overrides
35 executable scriptsin 35 files: backend-logic-design/scripts/build.py:1 dual-agent-collaboration/scripts/install_skill.py:1 dual-agent-collaboration/scripts/invoke_peer.py:1 and 32 more
48 references to secretsin 12 files: dual-agent-collaboration/scripts/invoke_peer.py:95 prd-test-writer/samples/PRD-SAMPLE-测试用例-review.html:105 prd-test-writer/references/test-case-example.md:18 and 9 more
No requests to vote in Hot

Show all 120 signals by file

`coding/backend-logic-design/scripts/build.py`

Executable scriptline 1
```
#!/usr/bin/env python3
```

`coding/dual-agent-collaboration/scripts/install_skill.py`

Executable scriptline 1
```
#!/usr/bin/env python3
```

`coding/dual-agent-collaboration/scripts/invoke_peer.py`

Executable scriptline 1
```
#!/usr/bin/env python3
```
Access to secretsline 95
```
[Credential-related source omitted]
```
Access to secretsline 100
```
[Credential-related source omitted]
```
Access to secretsline 101
```
[Credential-related source omitted]
```
Access to secretsline 146
```
[Credential-related source omitted]
```
Access to secretsline 157
```
[Credential-related source omitted]
```
Access to secretsline 159
```
[Credential-related source omitted]
```
Access to secretsline 314
```
[Credential-related source omitted]
```
Access to secretsline 315
```
[Credential-related source omitted]
```
Access to secretsline 331
```
[Credential-related source omitted]
```
Access to secretsline 332
```
[Credential-related source omitted]
```
Access to secretsline 367
```
[Credential-related source omitted]
```
Access to secretsline 368
```
[Credential-related source omitted]
```
Access to secretsline 830
```
[Credential-related source omitted]
```
Access to secretsline 863
```
[Credential-related source omitted]
```

`coding/dual-agent-collaboration/scripts/test_install_skill.py`

Executable scriptline 1
```
#!/usr/bin/env python3
```

`coding/dual-agent-collaboration/scripts/test_invoke_peer.py`

Executable scriptline 1
```
#!/usr/bin/env python3
```
Access to secretsline 199
```
[Credential-related source omitted]
```

`coding/issue-pool/tools/md2html.py`

Executable scriptline 1
```
#!/usr/bin/env python3
```

`coding/page-solution-design/scripts/check-layout.js`

Executable scriptline 1
```
Executable file
```

`coding/prd-test-writer/assets/prd-review.html.tmpl`

Network accessline 102

</script> <script type="module">  import mermaid from 'https://cdn.jsdelivr.net/npm/mermaid@11/dist/mermaid.esm.min.mjs';

Network accessline 103

<script type="module">  import mermaid from 'https://cdn.jsdelivr.net/npm/mermaid@11/dist/mermaid.esm.min.mjs';  mermaid.initialize({startOnLoad:true,theme:'default',securityLevel:'loose',flowchart:{useMaxWidth:true,htmlLabels:true},sequenc

Network accessline 104

import mermaid from 'https://cdn.jsdelivr.net/npm/mermaid@11/dist/mermaid.esm.min.mjs';

`coding/prd-test-writer/assets/test-cases-template.md`

Access to secretsline 91
```
[Credential-related source omitted]
```

`coding/prd-test-writer/references/test-case-example.md`

Access to secretsline 18
```
[Credential-related source omitted]
```
Access to secretsline 22
```
[Credential-related source omitted]
```
Access to secretsline 49
```
[Credential-related source omitted]
```
Access to secretsline 84
```
[Credential-related source omitted]
```

`coding/prd-test-writer/samples/PRD-SAMPLE-review.html`

Network accessline 265

</script> <script type="module">  import mermaid from 'https://cdn.jsdelivr.net/npm/mermaid@11/dist/mermaid.esm.min.mjs';

Network accessline 266

<script type="module">  import mermaid from 'https://cdn.jsdelivr.net/npm/mermaid@11/dist/mermaid.esm.min.mjs';  mermaid.initialize({startOnLoad:true,theme:'default',securityLevel:'loose',flowchart:{useMaxWidth:true,htmlLabels:true},sequenc

Network accessline 267

import mermaid from 'https://cdn.jsdelivr.net/npm/mermaid@11/dist/mermaid.esm.min.mjs';

`coding/prd-test-writer/samples/PRD-SAMPLE-测试用例-review.html`

Access to secretsline 105
```
[Credential-related source omitted]
```
Access to secretsline 139
```
[Credential-related source omitted]
```
Access to secretsline 147
```
[Credential-related source omitted]
```
Access to secretsline 188
```
[Credential-related source omitted]
```
Access to secretsline 227
```
[Credential-related source omitted]
```
Access to secretsline 266
```
[Credential-related source omitted]
```
Access to secretsline 277
```
[Credential-related source omitted]
```

`coding/req-change-workflow/scripts/impact_scan.sh`

Executable scriptline 1
```
#!/usr/bin/env bash
```

`visualization/case-radar/SKILL.md`

Network accessline 30

| 1 | scan | 多渠道扫信源，列出候选案例 | WebSearch / Agent（不重新发明，直接调用） | | 2 | recon | 给每个候选侦察"真物位置"——核心增值 | 读 `reference/recon-heuristics.md` | | 3 | capture | 拉真物（curl 直接资源 + agent-browser 截图） | curl + agent-browser |

Network accessline 31

| 2 | recon | 给每个候选侦察"真物位置"——核心增值 | 读 `reference/recon-heuristics.md` | | 3 | capture | 拉真物（curl 直接资源 + agent-browser 截图） | curl + agent-browser | | 4 | embed | 套 HTML 模板输出 | 参考 `reference/html-template-spec.md` |

Network accessline 32

| 3 | capture | 拉真物（curl 直接资源 + agent-browser 截图） | curl + agent-browser |

Network accessline 50

> 注意：`agent-browser install` 会下 Chrome 二进制（~169 MB）。如果在中国大陆，可能下载失败。但实测即使 install 卡住，agent-browser 也能复用机器上已有的 Playwright Chromium（`~/Library/Caches/ms-playwright/`）。所以即使 install 报错，先试一下 `agent-browser open https://example.com`，能跑就跳过 install。

Network accessline 104

对精读单里每个候选案例，问自己： 1. 这个对象的真物最可能在哪个栖息地？（参考 heuristics 的 7 种模式） 2. 用什么工具抓最合适？（curl 直接资源 / gh API 拉源码 / agent-browser 截图全页）

Network accessline 106

2. 用什么工具抓最合适？（curl 直接资源 / gh API 拉源码 / agent-browser 截图全页）

Network accessline 118

### 3.1 优先 curl 直接资源（最快、最干净）

Network accessline 121

curl -sLo screenshots/cases/<name>.<ext> "<url>"

Network accessline 256

→ 1.2 用户从 32 个里挑 6 个精读 → 2 对 6 个跑 recon，判别每个的真物位置 → 3 拉真物（curl 5 个直接图 + gh 拉 3 段 SKILL.md + agent-browser 截 2 个落地页）

Network accessline 257

→ 2 对 6 个跑 recon，判别每个的真物位置 → 3 拉真物（curl 5 个直接图 + gh 拉 3 段 SKILL.md + agent-browser 截 2 个落地页） → 4 出 HTML

Network accessline 258

→ 3 拉真物（curl 5 个直接图 + gh 拉 3 段 SKILL.md + agent-browser 截 2 个落地页）

`visualization/case-radar/reference/_how-to-extend.md`

Network accessline 22

- 至少有 1 个真实抓过的案例（不要预测性新增） - 能 3 秒判断"这个对象有没有这种栖息地" - 抓法是具体命令（curl / gh / agent-browser 之类），不是"自己想办法"

Network accessline 24

- 抓法是具体命令（curl / gh / agent-browser 之类），不是"自己想办法"

`visualization/case-radar/reference/recon-heuristics.md`

Network accessline 20
```
# curl 直接拉
```

Network accessline 21

curl -sLo screenshots/cases/<name>.<ext> "https://github.com/<owner>/<repo>/raw/main/screenshots/<file>"

Network accessline 41

agent-browser open https://claude.com/plugins/<name>

Network accessline 46

**实战案例**： - `obra/superpowers` → `https://claude.com/plugins/superpowers` 显示 64.5 万次安装、Jesse Vincent 作者署名、Anthropic 排版背书——比任何 GitHub README 截图都有说服力

Network accessline 47

- `obra/superpowers` → `https://claude.com/plugins/superpowers` 显示 64.5 万次安装、Jesse Vincent 作者署名、Anthropic 排版背书——比任何 GitHub README 截图都有说服力

Network accessline 81
```
**抓法**：curl 直接拉两张图。
```

`visualization/comic-maker/README.md`

Network accessline 7

- [Claude Code](https://claude.com/claude-code) 或 [Codex CLI](https://github.com/openai/codex)（Claude 负责设计画面，出图交给 Codex）

`visualization/comic-maker/scripts/_data.py`

Executable scriptline 1
```
Executable file
```
Access to secretsline 16
```
[Credential-related source omitted]
```
Access to secretsline 17
```
[Credential-related source omitted]
```

`visualization/comic-maker/scripts/doctor.py`

Executable scriptline 1
```
Executable file
```
Access to secretsline 19
```
[Credential-related source omitted]
```
Access to secretsline 20
```
[Credential-related source omitted]
```
Access to secretsline 44
```
[Credential-related source omitted]
```

Network accessline 57

line(True if importlib.util.find_spec("PIL") else None, "Pillow（可选，用来校验图片）", "pip3 install pillow；不装也能用，只是校验粗一些") codex = shutil.which("codex") line(bool(codex), f"Codex CLI：{codex or '没找到'}", "安装：npm i -g @openai/codex，或见 https://github.co

Network accessline 58

codex = shutil.which("codex") line(bool(codex), f"Codex CLI：{codex or '没找到'}", "安装：npm i -g @openai/codex，或见 https://github.com/openai/codex") if codex:

Network accessline 59

line(bool(codex), f"Codex CLI：{codex or '没找到'}", "安装：npm i -g @openai/codex，或见 https://github.com/openai/codex")

Access to secretsline 68
```
[Credential-related source omitted]
```

`visualization/comic-maker/scripts/mark.js`

Executable scriptline 1
```
Executable file
```

`visualization/comic-maker/scripts/render_codex.py`

Executable scriptline 1
```
Executable file
```
Access to secretsline 80
```
[Credential-related source omitted]
```

`visualization/comic-maker/scripts/settings.py`

Executable scriptline 1
```
Executable file
```

`visualization/image-render/README.md`

Network accessline 7

- [Claude Code](https://claude.com/claude-code) 或 [Codex CLI](https://github.com/openai/codex)（Claude 负责设计画面，出图交给 Codex）

`visualization/image-render/scripts/_lib.py`

Executable scriptline 1
```
Executable file
```
Access to secretsline 17
```
[Credential-related source omitted]
```
Access to secretsline 18
```
[Credential-related source omitted]
```
Access to secretsline 166
```
[Credential-related source omitted]
```

`visualization/image-render/scripts/add_sample.py`

Executable scriptline 1
```
Executable file
```

`visualization/image-render/scripts/build_prompts.py`

Executable scriptline 1
```
Executable file
```

`visualization/image-render/scripts/check_prompts.py`

Executable scriptline 1
```
Executable file
```

`visualization/image-render/scripts/doctor.py`

Executable scriptline 1
```
Executable file
```
Access to secretsline 19
```
[Credential-related source omitted]
```
Access to secretsline 20
```
[Credential-related source omitted]
```
Access to secretsline 44
```
[Credential-related source omitted]
```

Network accessline 57

line(True if importlib.util.find_spec("PIL") else None, "Pillow（可选，用来校验图片）", "pip3 install pillow；不装也能用，只是校验粗一些") codex = shutil.which("codex") line(bool(codex), f"Codex CLI：{codex or '没找到'}", "安装：npm i -g @openai/codex，或见 https://github.co

Network accessline 58

codex = shutil.which("codex") line(bool(codex), f"Codex CLI：{codex or '没找到'}", "安装：npm i -g @openai/codex，或见 https://github.com/openai/codex") if codex:

Network accessline 59

line(bool(codex), f"Codex CLI：{codex or '没找到'}", "安装：npm i -g @openai/codex，或见 https://github.com/openai/codex")

Access to secretsline 69
```
[Credential-related source omitted]
```

`visualization/image-render/scripts/make_page.py`

Executable scriptline 1
```
Executable file
```

`visualization/image-render/scripts/mark.js`

Executable scriptline 1
```
Executable file
```

`visualization/image-render/scripts/render_codex.py`

Executable scriptline 1
```
Executable file
```

`visualization/image-render/scripts/settings.py`

Executable scriptline 1
```
Executable file
```

`visualization/image-render/scripts/style_picker.py`

Executable scriptline 1
```
Executable file
```

`visualization/image-render/scripts/update_index.py`

Executable scriptline 1
```
Executable file
```

`visualization/image-render/scripts/verify_log.py`

Executable scriptline 1
```
Executable file
```

`visualization/scene-ppt/README.md`

Network accessline 7

- [Claude Code](https://claude.com/claude-code) 或 [Codex CLI](https://github.com/openai/codex)（Claude 负责设计画面，出图交给 Codex）

`visualization/scene-ppt/scripts/_lib.py`

Executable scriptline 1
```
Executable file
```
Access to secretsline 17
```
[Credential-related source omitted]
```
Access to secretsline 18
```
[Credential-related source omitted]
```
Access to secretsline 169
```
[Credential-related source omitted]
```

`visualization/scene-ppt/scripts/build_prompts.py`

Executable scriptline 1
```
Executable file
```

`visualization/scene-ppt/scripts/check_prompts.py`

Executable scriptline 1
```
Executable file
```

`visualization/scene-ppt/scripts/doctor.py`

Executable scriptline 1
```
Executable file
```
Access to secretsline 19
```
[Credential-related source omitted]
```
Access to secretsline 20
```
[Credential-related source omitted]
```
Access to secretsline 44
```
[Credential-related source omitted]
```

Network accessline 57

line(True if importlib.util.find_spec("PIL") else None, "Pillow（可选，用来校验图片）", "pip3 install pillow；不装也能用，只是校验粗一些") codex = shutil.which("codex") line(bool(codex), f"Codex CLI：{codex or '没找到'}", "安装：npm i -g @openai/codex，或见 https://github.co

Network accessline 58

codex = shutil.which("codex") line(bool(codex), f"Codex CLI：{codex or '没找到'}", "安装：npm i -g @openai/codex，或见 https://github.com/openai/codex") if codex:

Network accessline 59

line(bool(codex), f"Codex CLI：{codex or '没找到'}", "安装：npm i -g @openai/codex，或见 https://github.com/openai/codex")

Access to secretsline 68
```
[Credential-related source omitted]
```

`visualization/scene-ppt/scripts/make_page.py`

Executable scriptline 1
```
Executable file
```

`visualization/scene-ppt/scripts/mark.js`

Executable scriptline 1
```
Executable file
```

`visualization/scene-ppt/scripts/render_codex.py`

Executable scriptline 1
```
Executable file
```

`visualization/scene-ppt/scripts/settings.py`

Executable scriptline 1
```
Executable file
```

`visualization/scene-ppt/scripts/verify_log.py`

Executable scriptline 1
```
Executable file
```

`visualization/scene-ppt/scripts/world_picker.py`

Executable scriptline 1
```
Executable file
```

How the signals fit together

Facts from a static read of the files at the pinned commit. They describe what the code does, not why.

Some files could not be read; the facts below still hold.

No combined signals at this commit.

Works with

Agent compatibility, from the files in this repository
Agent	Status	Evidence	Install to
Claude Code	Compatible	Standard SKILL.md agent/goal-setter/SKILL.md	`.claude/skills/<skill-name>/`
Codex	Compatible	Standard SKILL.md agent/goal-setter/SKILL.md	`.agents/skills/<skill-name>/`
Cursor	Compatible	Standard SKILL.md agent/goal-setter/SKILL.md	`.cursor/skills/<skill-name>/`
Gemini CLI	Compatible	Standard SKILL.md agent/goal-setter/SKILL.md	`.gemini/skills/<skill-name>/`
GitHub Copilot	Compatible	Standard SKILL.md agent/goal-setter/SKILL.md	`.github/skills/<skill-name>/`
OpenCode	Compatible	Standard SKILL.md agent/goal-setter/SKILL.md	`.opencode/skills/<skill-name>/`

Native The repository ships a manifest, folder or frontmatter field for this agent.
Compatible A standard SKILL.md that this agent can read.
Unknown No evidence either way.

Derived automatically from paths, manifests and frontmatter. Not tested on each agent.

yunshu0909/yunshu_skillshub

History

agent/goal-setter/SKILL.md

agent/hermes-persona-builder/SKILL.md

agent/memory-init/SKILL.md

agent/organize/SKILL.md

brand/logo-design/SKILL.md

brand/product-naming/SKILL.md

coding/backend-logic-design/SKILL.md

coding/dual-agent-collaboration/SKILL.md

coding/git-push/SKILL.md

coding/github-repo-search/SKILL.md

coding/issue-pool/SKILL.md

coding/issue-triage/SKILL.md

coding/page-solution-design/SKILL.md

coding/prd-test-writer/SKILL.md

coding/req-change-workflow/SKILL.md

learning/article-study/SKILL.md

learning/system-study/SKILL.md

thinking/find-top-three/SKILL.md

thinking/multi-perspective-analysis/SKILL.md

thinking/thinking-partner/SKILL.md

visualization/case-radar/SKILL.md

visualization/comic-maker/SKILL.md

visualization/image-render/SKILL.md

visualization/readable-output/SKILL.md

visualization/scene-ppt/SKILL.md

writing/thought-mining/SKILL.md

writing/writing-assistant/SKILL.md

Share this page

Heat

Stars over time

Individual skills in this repository

How it works

goal-setter

Frontmatter

SKILL.md

Reference documents

The playbook

Files

hermes-persona-builder

Frontmatter

SKILL.md

Reference documents

The playbook

Files

memory-init

Frontmatter

SKILL.md

Reference documents

The playbook

Files

organize

Frontmatter

SKILL.md

Reference documents

The playbook

Files

logo-design

Frontmatter

SKILL.md

Reference documents

The playbook

Files

product-naming

Frontmatter

SKILL.md

Reference documents

The playbook

Files

backend-logic-design

Frontmatter

SKILL.md

Reference documents

The playbook

Scripts executed separately

Files

dual-agent-collaboration

Frontmatter

SKILL.md

`agent/goal-setter/SKILL.md`

`agent/hermes-persona-builder/SKILL.md`

`agent/memory-init/SKILL.md`

`agent/organize/SKILL.md`

`brand/logo-design/SKILL.md`

`brand/product-naming/SKILL.md`

`coding/backend-logic-design/SKILL.md`

`coding/dual-agent-collaboration/SKILL.md`

`coding/git-push/SKILL.md`

`coding/github-repo-search/SKILL.md`

`coding/issue-pool/SKILL.md`

`coding/issue-triage/SKILL.md`

`coding/page-solution-design/SKILL.md`

`coding/prd-test-writer/SKILL.md`

`coding/req-change-workflow/SKILL.md`

`learning/article-study/SKILL.md`

`learning/system-study/SKILL.md`

`thinking/find-top-three/SKILL.md`

`thinking/multi-perspective-analysis/SKILL.md`

`thinking/thinking-partner/SKILL.md`

`visualization/case-radar/SKILL.md`

`visualization/comic-maker/SKILL.md`

`visualization/image-render/SKILL.md`

`visualization/readable-output/SKILL.md`

`visualization/scene-ppt/SKILL.md`

`writing/thought-mining/SKILL.md`

`writing/writing-assistant/SKILL.md`