rp-why（回答为什么）

rp-why: Gas Town × DOK Framework

Overview

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.

How to Use This Skill

Installation

Install the skill using the skills CLI:

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.).

Using Slash Commands

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

Available Commands

Alternative: Natural Language

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.

When to Use

• 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

Problem Statement

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

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

The Framework

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

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

Webb's DOK Levels (Cognitive Complexity)

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

Integration Matrix (Stage × 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:

Commands

/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

/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

/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

Sample Output

╔══════════════════════════════════════════════════════════════════╗
║                    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?

Target User Profiles

Growth Nudge Reference

Frontier (High Stage, High DOK)

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

Thinking Ahead (Low Stage, High DOK)

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

Underutilizing (High Stage, Lower DOK)

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

Learning Zone (Low Stage, Low 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"

Overpowered (High Stage, Low 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"

Upgrading Your Prompts

Attribution

• 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.

Version History