rp-why

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

rp-why: Gas Town × DOK Framework

rp-why: Gas Town × DOK框架

Overview

概述

The rp-why skill is a self-reflection framework that helps AI practitioners measure and improve their AI collaboration practice. It combines two powerful dimensions:

Horizontal Axis: Gas Town Stages — Measures AI tool adoption maturity from basic chatbots to multi-agent orchestration
Vertical Axis: DOK Levels — Measures the cognitive complexity of prompts from simple recall to extended thinking

The intersection of these dimensions reveals growth opportunities and helps users maximize the value they extract from their AI tools.

rp-why是一个自我反思框架，帮助AI从业者衡量并提升其AI协作实践。它结合了两个强大的维度：

横轴：Gas Town阶段 — 衡量从基础聊天机器人到多Agent编排的AI工具采用成熟度
纵轴：DOK等级 — 衡量从简单回忆到深度拓展思考的提示词认知复杂度

这两个维度的交叉点可以揭示成长机会，帮助用户从AI工具中挖掘最大价值。

How to Use This Skill

如何使用该Skill

Installation

安装

Install the skill using the skills CLI:

bash

npx skills add https://github.com/block/agent-skills --skill rp-why

Make sure you have the built-in skills extension enabled in your agent (Goose, Claude Desktop, etc.).

使用Skills CLI安装该技能：

bash

npx skills add https://github.com/block/agent-skills --skill rp-why

请确保你的Agent（如Goose、Claude Desktop等）中已启用内置的Skills扩展。

Using Slash Commands

使用斜杠命令

Once the skill is loaded, you can use slash commands directly in your conversation:

You: /rp-why current

Goose will analyze your current session and provide:

Your Gas Town stage assessment
DOK distribution breakdown
Quadrant position
Growth nudges

技能加载完成后，你可以在对话中直接使用斜杠命令：

你：/rp-why current

Goose会分析当前会话并提供：

你的Gas Town阶段评估结果
DOK分布明细
象限定位
成长建议

Available Commands

可用命令

Command	What It Does
`/rp-why current`	Analyze the current session
`/rp-why init`	Generate a baseline from your history
`/rp-why compare`	Compare current session to baseline

命令	功能
`/rp-why current`	分析当前会话
`/rp-why init`	基于历史对话生成基准线
`/rp-why compare`	对比当前会话与基准线

Alternative: Natural Language

替代方式：自然语言

You don't have to use slash commands. You can also just ask naturally:

You: Analyze my AI collaboration patterns using the Gas Town DOK framework
You: What's my DOK distribution for this session?
You: How does this session compare to my baseline?

The skill will recognize these requests and provide the same analysis.

你也可以不用斜杠命令，直接用自然语言提问：

你：用Gas Town DOK框架分析我的AI协作模式
你：我当前会话的DOK分布是怎样的？
你：当前会话和我的基准线相比有什么变化？

该技能会识别这些请求并提供相同的分析结果。

When to Use

使用场景

End of session: Run
```
/rp-why current
```
to reflect on your work
Weekly: Run
```
/rp-why compare
```
to track progress
First time: Run
```
/rp-why init
```
to establish your baseline

会话结束时：运行
```
/rp-why current
```
来复盘工作
每周复盘：运行
```
/rp-why compare
```
来跟踪进度
首次使用：运行
```
/rp-why init
```
来建立你的基准线

Problem Statement

问题背景

Many AI practitioners face a hidden inefficiency: a mismatch between tool sophistication and task cognitive complexity.

Anti-Pattern	Impact
Using powerful autonomous agents for simple "what is X?" queries	Unrealized potential
Asking deep strategic questions through basic chatbot interfaces	Bottlenecked thinking
No visibility into personal AI usage patterns	Stagnant growth
No framework for intentional growth in AI collaboration skills	Missed opportunities

Without measurement, there's no improvement. Users need a mirror to see their AI collaboration patterns clearly.

许多AI从业者面临一个隐性的效率问题：工具复杂度与任务认知复杂度不匹配。

反模式	影响
用强大的自主Agent处理简单的“什么是X？”类问题	未充分发挥工具潜力
通过基础聊天机器人界面提出深度战略问题	思维受限
无法了解个人AI使用模式	成长停滞
缺乏AI协作技能的系统性成长框架	错失机会

没有衡量就没有提升。用户需要一个“镜子”来清晰看到自己的AI协作模式。

The Framework

框架详情

Yegge's 8 Gas Town Stages (AI Tool Adoption)

Yegge的8个Gas Town阶段（AI工具采用度）

From Steve Yegge's "Welcome to Gas Town" (January 2026):

Stage	Name	Description
8	Full Gas Town	Complete AI-native development ecosystem
7	Agentic Workflows	Automated pipelines with agent coordination
6	Multi-Agent	Orchestrating multiple specialized agents
5	CLI Single Agent, YOLO	Terminal-based autonomous agent (e.g., Goose)
4	Chat IDE	Integrated chat in development environment
3	Copilot	Using AI code completion, inline suggestions
2	Curious	Experimenting with basic chatbots occasionally
1	Observer	Watching and evaluating AI tools, not yet actively using

来自Steve Yegge的《Welcome to Gas Town》（2026年1月）：

阶段	名称	描述
8	完整Gas Town	成熟的AI原生开发生态系统
7	Agent化工作流	带Agent协同的自动化流水线
6	多Agent	编排多个专业Agent
5	CLI单Agent（YOLO）	基于终端的自主Agent（如Goose）
4	聊天IDE	集成在开发环境中的聊天功能
3	代码助手（Copilot）	使用AI代码补全、行内建议
2	探索者	偶尔尝试基础聊天机器人
1	观察者	关注并评估AI工具，但尚未实际使用

Webb's DOK Levels (Cognitive Complexity)

Webb的DOK等级（认知复杂度）

From Norman Webb's Depth of Knowledge framework (1997):

Level	Name	Description	Prompt Indicators
4	Extended Thinking	Complex investigation, multiple sessions	"Research and synthesize...", "Create a framework...", "Investigate over time..."
3	Strategic Thinking	Reasoning, planning, analysis, synthesis	"Design...", "Analyze...", "What if...", "Develop a strategy..."
2	Application	Apply concepts, make decisions, compare	"How would you...", "Compare...", "Explain why..."
1	Recall	Facts, definitions, simple procedures	"What is...", "List...", "Define..."

来自Norman Webb的知识深度框架（1997年）：

等级	名称	描述	提示词特征
4	拓展思考	复杂调研、跨多会话任务	“调研并综合...”, “创建一个框架...”, “长期调研...”
3	战略思考	推理、规划、分析、综合	“设计...”, “分析...”, “如果...会怎样”, “制定战略...”
2	应用实践	应用概念、做决策、对比	“你会如何...”, “对比...”, “解释原因...”
1	回忆识记	事实、定义、简单流程	“什么是...”, “列出...”, “定义...”

Integration Matrix (Stage × DOK)

整合矩阵（阶段 × DOK）

The intersection creates six distinct zones:

              DOK 1        DOK 2         DOK 3          DOK 4
            (Recall)   (Application) (Strategic)   (Extended)
           ┌──────────┬──────────────┬────────────┬────────────┐
Stage 6-8  │   Over-  │    Over-     │ Underutil- │  Frontier  │
(Multi/    │  powered │   powered    │   izing    │            │
 Agentic)  │          │              │            │            │
           ├──────────┼──────────────┼────────────┼────────────┤
Stage 5    │   Over-  │  Underutil-  │  Expected  │  Growing   │
(CLI YOLO) │  powered │    izing     │            │            │
           ├──────────┼──────────────┼────────────┼────────────┤
Stage 3-4  │   Over-  │   Expected   │  Growing   │  Frontier  │
(Copilot/  │  powered │              │            │            │
 Chat IDE) │          │              │            │            │
           ├──────────┼──────────────┼────────────┼────────────┤
Stage 1-2  │ Expected │   Growing    │  Thinking  │  Thinking  │
(Observer/ │          │              │   Ahead    │   Ahead    │
 Curious)  │          │              │            │            │
           └──────────┴──────────────┴────────────┴────────────┘

Zone Definitions:

Zone	Description	Action
Frontier	Pushing boundaries of both tool and cognition	Celebrate & Document
Thinking Ahead	High cognitive work with basic tools	Upgrade tools
Growing	Stretching into higher complexity, positive trajectory	Encourage
Expected	Appropriate match of tool sophistication to task complexity	Maintain
Underutilizing	Sophisticated tools for simpler tasks	Increase DOK
Overpowered	Tools exceed task needs—opportunity to level up your questions	Realign

两个维度的交叉形成六个不同的区域：

              DOK 1        DOK 2         DOK 3          DOK 4
            (Recall)   (Application) (Strategic)   (Extended)
           ┌──────────┬──────────────┬────────────┬────────────┐
Stage 6-8  │   过度    │    过度      │  未充分    │  前沿区    │
(Multi/    │  赋能     │   赋能       │   利用     │            │
 Agentic)  │          │              │            │            │
           ├──────────┼──────────────┼────────────┼────────────┤
Stage 5    │   过度    │  未充分      │  预期区    │  成长区    │
(CLI YOLO) │  赋能     │    利用      │            │            │
           ├──────────┼──────────────┼────────────┼────────────┤
Stage 3-4  │   过度    │   预期区     │  成长区    │  前沿区    │
(Copilot/  │  赋能     │              │            │            │
 Chat IDE) │          │              │            │            │
           ├──────────┼──────────────┼────────────┼────────────┤
Stage 1-2  │ 预期区    │   成长区     │  超前思考  │  超前思考  │
(Observer/ │          │              │   区       │   区       │
 Curious)  │          │              │            │            │
           └──────────┴──────────────┴────────────┴────────────┘

区域定义：

区域	描述	行动建议
前沿区	同时突破工具和认知的边界	总结并记录经验
超前思考区	用基础工具完成高认知复杂度任务	升级工具
成长区	向更高复杂度拓展，成长轨迹积极	持续推进
预期区	工具复杂度与任务认知复杂度匹配	保持当前状态
未充分利用区	用复杂工具处理简单任务	提升DOK等级
过度赋能区	工具能力远超任务需求——升级提问质量的机会	调整匹配度

Commands

命令详情

/rp-why current

/rp-why current

Analyze the current session's Gas Town stage and DOK distribution.

Output includes:

Stage assessment with confidence level
DOK distribution breakdown with percentages
Quadrant position visualization
Contextual growth nudges
Reflection prompt

分析当前会话的Gas Town阶段和DOK分布。

输出内容包括：

带置信度的阶段评估
带百分比的DOK分布明细
象限位置可视化
针对性成长建议
反思提示

/rp-why init

/rp-why init

Generate a baseline from your conversation history (analyzes available sessions).

Output includes:

Historical analysis period and session count
Baseline DOK distribution
Typical Gas Town stage
Growth targets
Baseline saved to
```
~/.config/goose/rp-why-baseline.json
```

基于对话历史生成基准线（分析可用的会话记录）。

输出内容包括：

历史分析周期和会话数量
基准DOK分布
典型Gas Town阶段
成长目标
基准线保存至
```
~/.config/goose/rp-why-baseline.json
```

/rp-why compare

/rp-why compare

Compare current session against your established baseline.

Output includes:

Side-by-side DOK comparison (baseline vs current)
Quadrant movement visualization
Progress toward growth targets
Trajectory analysis

对比当前会话与已建立的基准线。

输出内容包括：

DOK分布对比（基准线 vs 当前会话）
象限变化可视化
成长目标进度
轨迹分析

Sample Output

示例输出

╔══════════════════════════════════════════════════════════════════╗
║                    rp-why: CURRENT SESSION                       ║
╚══════════════════════════════════════════════════════════════════╝

GAS TOWN STAGE: 5 (CLI Single Agent, YOLO)

DOK DISTRIBUTION
────────────────────────────────────────────────────────────────────
DOK 1 (Recall):      ████░░░░░░░░░░░░░░░░  17%
DOK 2 (Application): ████████████░░░░░░░░  52%
DOK 3 (Strategic):   ██████░░░░░░░░░░░░░░  26%
DOK 4 (Extended):    █░░░░░░░░░░░░░░░░░░░   5%

QUADRANT: Underutilizing
────────────────────────────────────────────────────────────────────
You're using powerful autonomous tools—there's an opportunity to
match your questions to that power.

GROWTH NUDGES
────────────────────────────────────────────────────────────────────
1. Shift 2-3 DOK 2 prompts to DOK 3 by adding "analyze trade-offs"
2. Before simple queries, ask: "Can I make this more strategic?"
3. Try one DOK 4 extended investigation this week

🪞 REFLECTION
────────────────────────────────────────────────────────────────────
What complex challenge could benefit from your agent's full
capabilities today?

╔══════════════════════════════════════════════════════════════════╗
║                    rp-why: 当前会话分析                          ║
╚══════════════════════════════════════════════════════════════════╝

GAS TOWN阶段: 5 (CLI单Agent, YOLO)

DOK分布
────────────────────────────────────────────────────────────────────
DOK 1 (回忆):      ████░░░░░░░░░░░░░░░░  17%
DOK 2 (应用): ████████████░░░░░░░░  52%
DOK 3 (战略):   ██████░░░░░░░░░░░░░░  26%
DOK 4 (拓展):    █░░░░░░░░░░░░░░░░░░░   5%

象限: 未充分利用
────────────────────────────────────────────────────────────────────
你正在使用强大的自主工具——有机会让你的提问匹配工具的能力。

成长建议
────────────────────────────────────────────────────────────────────
1. 通过添加“分析权衡”将2-3个DOK 2提示词升级为DOK 3
2. 在提出简单问题前，先问自己：“我能不能让这个问题更具战略性？”
3. 本周尝试一个DOK 4级的拓展调研任务

🪞 反思
────────────────────────────────────────────────────────────────────
今天有什么复杂挑战可以充分发挥你的Agent的全部能力？

Target User Profiles

目标用户画像

Profile	Typical Stage	DOK Distribution	Characteristics
Traditional	1-2	DOK1: 60%, DOK2: 30%, DOK3: 10%	Minimal AI use
Adopter	3-4	DOK1: 40%, DOK2: 40%, DOK3: 15%, DOK4: 5%	Growing comfort
Practitioner	5	DOK1: 25%, DOK2: 45%, DOK3: 25%, DOK4: 5%	Autonomous agents
Advanced	5-6	DOK1: 15%, DOK2: 35%, DOK3: 35%, DOK4: 15%	Strategic use
Frontier	7-8	DOK1: 10%, DOK2: 25%, DOK3: 40%, DOK4: 25%	Agentic workflows

用户画像	典型阶段	DOK分布	特征
传统用户	1-2	DOK1: 60%, DOK2: 30%, DOK3: 10%	极少使用AI
入门用户	3-4	DOK1: 40%, DOK2: 40%, DOK3: 15%, DOK4: 5%	逐渐适应AI工具
从业者	5	DOK1: 25%, DOK2: 45%, DOK3: 25%, DOK4: 5%	使用自主Agent
进阶用户	5-6	DOK1: 15%, DOK2: 35%, DOK3: 35%, DOK4: 15%	战略性使用AI
前沿用户	7-8	DOK1: 10%, DOK2: 25%, DOK3: 40%, DOK4: 25%	使用Agent化工作流

Growth Nudge Reference

成长建议参考

Frontier (High Stage, High DOK)

前沿区（高阶段，高DOK）

"You're pushing boundaries—document what you learn"
"Share patterns with others; teach what works"
"Explore the edges: what's not yet possible?"

“你正在突破边界——记录下你的经验”
“与他人分享你的模式；传授有效的方法”
“探索未知：哪些是目前还做不到的？”

Thinking Ahead (Low Stage, High DOK)

超前思考区（低阶段，高DOK）

"Your thinking exceeds your tools—time to upgrade!"
"Explore CLI agents or IDE integration"
"Your DOK is strong; let better tools amplify it"

“你的思考已经超越了工具的能力——是时候升级工具了！”
“尝试CLI Agent或IDE集成工具”
“你的DOK能力很强；让更好的工具放大它”

Underutilizing (High Stage, Lower DOK)

未充分利用区（高阶段，低DOK）

"Powerful tools deserve powerful questions"
"Before each prompt, ask: Can this be more strategic?"
"Batch simple queries; save the agent for complex work"

“强大的工具值得更有深度的问题”
“在每次提问前，问自己：这个问题能不能更具战略性？”
批量处理简单问题；让Agent专注于复杂工作

Learning Zone (Low Stage, Low DOK)

学习区（低阶段，低DOK）

"This is a natural starting point—focus on learning the tools"
"Try one new AI capability each session"
"Don't worry about DOK yet—get comfortable first"

“这是自然的起点——先专注于学习工具”
“每次会话尝试一个新的AI功能”
“先不用在意DOK——先熟悉工具”

Overpowered (High Stage, Low DOK)

过度赋能区（高阶段，低DOK）

"Your tools exceed your task needs—opportunity to level up your questions"
"Consider: Is this query worth an autonomous agent?"
"Batch simple lookups; reserve agent for strategic work"

“你的工具能力远超任务需求——这是提升提问质量的机会”
“思考：这个问题值得用自主Agent来处理吗？”
“批量处理简单查询；预留Agent用于战略工作”

Upgrading Your Prompts

优化你的提示词

DOK Level	Prompt Pattern	Example
1 → 2	Add "how" or "why"	"What is a mutex?" → "How would I use a mutex here?"
2 → 3	Add "trade-offs" or "design"	"How do I implement caching?" → "Design a caching strategy considering our constraints"
3 → 4	Extend across sessions	"Analyze this architecture" → "Research caching patterns over multiple sessions and synthesize recommendations"

DOK等级提升	提示词改造模式	示例
1 → 2	添加“如何”或“为什么”	“什么是互斥锁？” → “我该如何在这里使用互斥锁？”
2 → 3	添加“权衡”或“设计”	“我如何实现缓存？” → “结合我们的约束设计一个缓存策略”
3 → 4	拓展为跨会话任务	“分析这个架构” → “跨多会话调研缓存模式并综合给出建议”

Attribution

参考来源

Gas Town Stages: Steve Yegge, "Welcome to Gas Town" (January 2026) https://steve-yegge.medium.com/welcome-to-gas-town-4f25ee16dd04
Depth of Knowledge (DOK): Norman Webb (1997) Webb, N. L. (1997). Criteria for alignment of expectations and assessments in mathematics and science education. Council of Chief State School Officers.

Gas Town阶段：Steve Yegge，《Welcome to Gas Town》（2026年1月） https://steve-yegge.medium.com/welcome-to-gas-town-4f25ee16dd04
知识深度（DOK）：Norman Webb（1997年） Webb, N. L. (1997). Criteria for alignment of expectations and assessments in mathematics and science education. Council of Chief State School Officers.

Version History

版本历史

Version	Date	Changes
3.0	2026-02	Quadrant visualization, growth nudges, reflection prompts, updated terminology
2.x	2026-01	Integration matrix, target profiles, baseline comparison
1.x	2025-12	Initial Gas Town stages, basic DOK tracking

版本	日期	变更
3.0	2026-02	象限可视化、成长建议、反思提示、术语更新
2.x	2026-01	整合矩阵、目标用户画像、基准线对比
1.x	2025-12	初始Gas Town阶段、基础DOK追踪