ugav2-universal-goal-analysis-v2-actionable-questions

name	ugav2 - Universal Goal Analysis v2 (Actionable Questions)
description	Goal analysis focused on generating fact-yielding questions. Each question must produce a testable fact, not a vague reflection. Consolidated into /uga but usable standalone for question-focused goal analysis.
tier	tier4
categories	["Goal Processing"]
tags	["goal","questions","actionable","analysis"]
output	{"format":"prose"}

Universal Goal Analysis v2: Actionable Fact-Yielding Questions

Input: $ARGUMENTS

---

Core Principles

1. Every question must yield a fact. A question like "what do you want?" yields a feeling. A question like "what measurable outcome would make you say this succeeded?" yields a testable fact. v2's entire contribution is this distinction. If a question can be answered with "I feel like..." it is not actionable.

2. Questions precede analysis. Do not analyze the goal until you have asked fact-yielding questions about it. Analysis without facts is speculation dressed as insight. The question phase is not warmup — it IS the work.

3. Binary questions outperform open questions. "Is there a deadline?" yields a fact. "Tell me about the timeline" yields rambling. Start binary, then drill into the yes/no answers.

4. Unanswerable questions are findings. When a question cannot be answered, that gap IS the most important output. A goal with three unanswerable questions is a goal with three unresolved risks.

5. Questions must be ordered by dependency. Ask existence questions before detail questions. "Is there a budget?" before "How large is the budget?" Skipping existence questions is the primary way analyses hallucinate structure that isn't there.

---

Phase 1: Goal Capture

State the goal as received, then immediately test its specificity.

[A] STATED_GOAL: [verbatim from user]
[B] SPECIFICITY_TEST:
    - Can you measure completion? [Y/N — if N, goal is vague]
    - Is there a deadline? [Y/N — if N, no urgency signal]
    - Is there one actor or many? [1/many — affects delegation questions]
    - Has this been attempted before? [Y/N — if Y, ask what happened]
[C] GOAL_CLASS: [vague | directional | specific | already-in-progress]

---

Phase 2: Fact-Yielding Question Battery

For each category, ask binary existence questions first, then detail questions only where existence = Y.

2a. Outcome Facts

[D] OUTCOME_QUESTIONS:
    1. Is there a measurable definition of success? → [Y/N] → [if Y: what?]
    2. Is there a measurable definition of failure? → [Y/N] → [if Y: what?]
    3. Is there a minimum viable outcome (less than full success but still worth it)? → [Y/N]
    4. Will someone other than you judge success? → [Y/N] → [if Y: who? what criteria?]
    5. Is there a point of no return? → [Y/N] → [if Y: when?]

2b. Resource Facts

[E] RESOURCE_QUESTIONS:
    1. Is there a budget? → [Y/N] → [if Y: how much?]
    2. Is there a time limit? → [Y/N] → [if Y: what?]
    3. Are there people available to help? → [Y/N] → [if Y: how many? what skills?]
    4. Is there existing work to build on? → [Y/N] → [if Y: what state is it in?]
    5. Are there tools/systems already in place? → [Y/N] → [if Y: what?]

2c. Constraint Facts

[F] CONSTRAINT_QUESTIONS:
    1. Is there something that cannot change? → [Y/N] → [if Y: what?]
    2. Is there a dependency on someone else's action? → [Y/N] → [if Y: who? what?]
    3. Is there a regulatory or policy constraint? → [Y/N] → [if Y: what?]
    4. Is there a technical constraint? → [Y/N] → [if Y: what?]
    5. Is there a political/organizational constraint? → [Y/N] → [if Y: what?]

2d. Context Facts

[G] CONTEXT_QUESTIONS:
    1. Has this goal been attempted before? → [Y/N] → [if Y: what happened?]
    2. Is someone else pursuing this same goal? → [Y/N] → [if Y: who? are they ahead?]
    3. Is this goal instrumental to a deeper goal? → [Y/N] → [if Y: what's the real goal?]
    4. Was there a triggering event? → [Y/N] → [if Y: what? when?]
    5. Is there an alternative approach being considered? → [Y/N] → [if Y: what?]

---

Phase 3: Gap Analysis

[H] UNANSWERED_QUESTIONS: [list every question from Phase 2 that could not be answered]
[I] RISK_RANKING: [rank unanswered questions by: if this assumption is wrong, how bad?]
[J] FACT_SUMMARY:
    KNOWN_FACTS: [N]
    UNKNOWN_FACTS: [M]
    RATIO: [N/(N+M)] — below 0.5 = goal is under-specified

---

Phase 4: Actionable Recommendation

Based on the fact/gap ratio, recommend next step:

[K] RECOMMENDATION:
    IF ratio >= 0.7: Goal is well-specified → proceed to strategy (/uga Step 4)
    IF ratio 0.4-0.7: Goal has gaps → answer top 3 unknowns before strategizing
    IF ratio < 0.4: Goal is under-specified → return to goal clarification (/gu)

[L] TOP_3_UNKNOWNS_TO_RESOLVE:
    1. [question] — how to answer: [specific action]
    2. [question] — how to answer: [specific action]
    3. [question] — how to answer: [specific action]

---

Phase 5: Report

UGA v2 FACT-YIELDING ANALYSIS:
Goal: [stated goal]
Specificity: [vague/directional/specific/in-progress]

Known facts: [N] | Unknown: [M] | Ratio: [X]

Key facts established:
- [fact 1]
- [fact 2]
- [fact 3]

Critical unknowns:
1. [unknown] — risk if wrong: [consequence]
2. [unknown] — risk if wrong: [consequence]
3. [unknown] — risk if wrong: [consequence]

Recommendation: [proceed / resolve gaps / clarify goal]
Next action: [specific action to take]

→ INVOKE: /uga $ARGUMENTS (for full analysis after fact-gathering)

---

Failure Modes

Failure	Signal	Fix
Reflective questions	Question can be answered with "I feel like..."	Rewrite to demand a measurable fact
Skipping existence checks	Asking "how large is the budget?" without first asking "is there a budget?"	Always binary first, detail second
Accepting vague answers	"The timeline is flexible" treated as an answer	Push: "Is there a date after which this loses value? Y/N"
Question overload	40+ questions with no prioritization	Cap at 20, rank by dependency and risk
Analysis before facts	Strategizing when ratio < 0.4	Block strategy until ratio >= 0.4
Ignoring gaps	Unanswered questions not flagged as risks	Every unanswered question is a finding

---

Depth Scaling

Depth	Questions Asked	Gap Analysis	Recommendation
1x	10 binary questions across 2 categories	List unknowns	Single next step
2x	20 questions across all 4 categories	Ranked unknowns with risk	Top 3 actions
4x	20 questions + 10 follow-up detail questions	Risk-scored gap matrix	Sequenced action plan
8x	Full battery + custom domain questions	Gap dependency tree	Full strategy with contingencies

Default: 2x. These are floors.

---

Pre-Completion Checklist

• Every question is fact-yielding (not reflective)
• Existence questions asked before detail questions
• Unanswered questions explicitly listed as findings
• Fact/unknown ratio calculated
• Top 3 unknowns identified with resolution methods
• Recommendation matches the ratio threshold
• No strategy offered when ratio < 0.4

---

Integration

• Consolidated into: /uga (which includes v2's question approach in Steps 1-2)
• Use standalone when: You need pure fact-gathering before analysis
• Routes to: /uga (full analysis), /gu (goal clarification if under-specified)
• Invoked by: Users wanting question-first goal analysis
• Differs from /uga: /uga runs the full 17-step framework; ugav2 focuses exclusively on fact-yielding questions
• Differs from /gu: /gu clarifies what the goal IS; ugav2 establishes what facts support it