qo-question-ordering

name	qo - Question Ordering
description	Find the satisfying unresolved question that opens a document and order all questions in a dependency chain. Upstream of /w.
output	{"format":"prose"}

Question Ordering

Input: $ARGUMENTS

---

What This Skill Does

Before writing, you need to know: what question does this document resolve, and in what order should the reader encounter sub-questions? This skill finds both.

The writing skill assumes you already have the question and the steps. This skill finds them. Pipeline: question_ordering → high_quality_writing.

---

Core Principles

1. Problems are more satisfying than solutions as openers. A problem creates tension — the reader wants resolution. A solution closes tension — the reader has nothing to pursue. Open with the problem.

2. Question order is a dependency chain, not chosen by interest. Questions have prerequisites. "Why is X needed?" requires knowing what X is. "What is X?" requires having a problem X solves. Order follows dependencies.

3. The opening problem is found by tracing backward. Start at the target conclusion. Ask: "What must the reader know to arrive here?" Repeat until you hit universal experience. Read the chain forward for the question order.

4. Questions not in the dependency chain are digressions. If a question is not a prerequisite for the conclusion, it breaks momentum. Cut it.

5. Questions and answers have separate timing. A question can be posed (tension created) at a different point than it is answered (tension resolved). The opening problem is posed first and answered last.

---

The Process

Step 1: Define the Target

What should the reader believe, know, or conclude after reading?

TARGET CONCLUSION: [What the reader arrives at]
TARGET READER: [Who — what do they already know/believe?]

Step 2: Backward Chain

Starting from the target conclusion, trace prerequisites backward. At each step ask: "What must the reader know or feel to arrive at this?"

TARGET: [conclusion]
  ← REQUIRES: [what must be known to reach this]
    ← REQUIRES: [what must be known to reach THAT]
      ← REQUIRES: [...]
        ← REQUIRES: [universal experience — no prerequisites]

Stop when you reach something the reader has experienced without any setup — something with no prerequisites. This is the opening problem.

Step 3: Validate the Opening Problem

The opening problem must pass ALL four tests:

Test	Question	Must Be
Universal	Has the reader experienced this without any setup?	YES
Felt	Does stating it create tension — does the reader want resolution?	YES
Unsolved	Does the reader lack a known/obvious solution?	YES
Connected	Is this what the target conclusion ultimately resolves?	YES

If any test fails, the chain hasn't gone back far enough, or the wrong branch was taken. Try a different prerequisite path.

Failure modes for opening problems:

• Reader hasn't experienced it → no recognition → no tension
• Reader already has the answer → no tension → boring
• Obvious solution exists → reader dismisses ("just do X") → stops reading
• Not connected to conclusion → document is incoherent — opens one question, answers another

Step 4: Read the Chain Forward

The backward chain, read forward, is the question order:

QUESTION ORDER:
1. [Opening problem — from universal experience]
2. [First sub-question — enabled by feeling the problem]
3. [Next sub-question — enabled by answering the previous]
...
N. [Target conclusion — reader arrives here]

Each question should be enabled by the answer to the previous one. If a question could be asked without the prior answer, it's either in the wrong position or not in the dependency chain.

Step 5: Map Question Timing

For each question, note when it is POSED (tension created) and when it is ANSWERED (tension resolved):

QUESTION MAP:
| # | Question | Posed | Answered | Span | Dependencies |
|---|----------|-------|----------|------|-------------|
| 1 | [opening problem] | Opening | End | Full document | None |
| 2 | [sub-question] | After Q1 posed | Section 2 | Short | Q1 |
| 3 | [sub-question] | After Q2 answered | Section 3 | Short | Q1, Q2 |
| ... | ... | ... | ... | ... | ... |

Timing principles:

• Opening problem: posed first, answered last (maximum tension span)
• Intermediate questions: shorter spans — posed and answered within sections
• Each answer should create the next question (momentum)
• Final answer resolves all remaining tension

Step 6: Validate the Chain

Check each question against the six ordering principles:

Principle	Check
Dependency	Does each question's answer require the prior answer?
Recognition first	Is the opening from universal experience (no prerequisites)?
Tension before resolution	Is every question posed before it's answered?
Escalation	Do questions get deeper/harder as they go? (Should happen naturally from dependencies)
Opening lacks solution	Does the reader genuinely not know the answer to Q1?
Answers create questions	Does each answer create the tension that drives the next question?

If any principle is violated, reorder or remove questions.

---

Output Format

QUESTION ORDERING FOR: [document title/topic]

TARGET CONCLUSION: [what the reader should arrive at]
TARGET READER: [who]

BACKWARD CHAIN:
[target]
  ← [prerequisite]
    ← [prerequisite]
      ← [opening problem — universal experience]

OPENING PROBLEM:
[Statement of the problem]
- Universal: [why the reader has experienced this]
- Felt: [why it creates tension]
- Unsolved: [why the reader lacks a solution / why obvious solutions fail]
- Connected: [how the target conclusion resolves this]

QUESTION ORDER:
1. [Q1 — opening problem]
2. [Q2 — first sub-question, enabled by Q1]
3. [Q3 — next, enabled by Q2]
...

QUESTION MAP:
| # | Question | Posed | Answered | Span | Dependencies |
|---|----------|-------|----------|------|-------------|
| 1 | ... | ... | ... | ... | ... |

VALIDATION:
- [ ] Dependency: each question requires prior answer
- [ ] Recognition first: opening from universal experience
- [ ] Tension before resolution: every question posed before answered
- [ ] Escalation: questions deepen naturally
- [ ] Opening unsolved: reader genuinely doesn't know Q1's answer
- [ ] Answers create questions: each answer drives the next

READY FOR: /w [topic]

---

Pre-Completion Check

• Target conclusion and reader defined
• Backward chain reaches universal experience
• Opening problem passes all four tests (universal, felt, unsolved, connected)
• Question order follows dependency chain
• No questions outside the dependency chain (digressions removed)
• Question map includes timing (posed/answered) for each question
• All six ordering principles validated
• Output includes "READY FOR: /w"