u-universalize

name	u - Universalize
description	Universalize - Extract the complete space of assumptions, dimensions, and alternatives from any claim. Map before you judge. Track every finding.
output	{"format":"registry"}

U - Universalize

Input: $ARGUMENTS

---

Interpretations

Before executing, identify which interpretation matches the user's input:

Interpretation 1 — Map the full space around a claim: The user has a specific claim, belief, or statement and wants to extract every assumption, alternative, dimension, and perspective before evaluating anything — pure exploration of the possibility space. Interpretation 2 — Unbundle a complex position: The user has a dense or compound statement (theirs or someone else's) that packs multiple claims together — they want to separate, number, and surface everything that's bundled inside it. Interpretation 3 — Find what they're missing: The user suspects their framing is too narrow or that they're anchored on one perspective — they want universalization specifically to discover hidden dimensions, unconsidered alternatives, and blind spots.

If ambiguous, ask: "I can help with mapping the full space around a claim, unbundling a complex position, or finding what you're missing — which fits?" If clear from context, proceed with the matching interpretation.

Corruption Pre-Inoculation

User praise or validation is a signal to test HARDER, not softer. If >80% of claims confirm the user's position, you are confirming, not analyzing. Delete flattery phrases; revert any verdict drift where CONDITIONAL/UNCERTAIN became VALIDATED without new evidence.

Full protocol: _shared/corruption-pre-inoculation.md

---

Core Principles

1. Map before you judge. Universalization is pure exploration — finding what EXISTS, not what's GOOD. Don't evaluate during mapping. That's AR/AW's job.

2. Explore before concluding. U has two phases: EXPLORATION (find everything) and REGISTRY (compile what you found). Never conclude during exploration. Never introduce new findings during registry.

3. Every finding gets tracked. When you find an assumption, alternative, dimension, or perspective — number it. It goes in the registry. Nothing gets lost in prose.

4. Categories reveal siblings. When you identify what something is an INSTANCE of, you immediately see what else is in that category. This is the most powerful technique — use it.

5. The goal is completeness, not depth. U produces breadth — the full space. A good U session makes you say "I didn't realize there were that many options."

6. Unbundling is mandatory. Every statement contains multiple claims. Find them all before applying techniques.

---

Phase 1: EXPLORATION

Step 1: State and Unbundle the Claim

INPUT: [what the user said]
CLAIM: [precise restated version]
CLAIM TYPE: [factual / strategic / design / causal / belief / assumption / decision / meta]

Then unbundle — single statements contain multiple claims. Number each:

[U1] EXPLICIT: [what's directly stated]
[U2] IMPLICIT: [what's assumed but not said]
[U3] PRESUPPOSED: [what must be true for the statement to make sense]
[U4] BUNDLED: [separate assertions packed together]
[U5] META: [claims about the type of question or how to approach it]

Step 2: Apply Techniques

For each unbundled claim, apply techniques that produce findings. Number every finding.

Technique 1: STATE SPACE — What states could this be in?

[U6] [alternative 1]
[U7] [alternative 2]
[U8] [the negation]
[U9] [the "do nothing" option]
[U10] [the reframe — what if the question is wrong?]

Technique 2: INSTANCE-TO-CATEGORY — What is this an instance of? What are siblings?

[U11] [X] is an instance of [CATEGORY]
[U12] Sibling: [sibling 1]
[U13] Sibling: [sibling 2]
[U14] Go up a level: [CATEGORY] is an instance of [HIGHER CATEGORY]
[U15] Higher sibling: [sibling at higher level]

Technique 3: PARAMETER VARIATION — What are the variables and their ranges?

[U16] Parameter: [name] — current: [value] — range: [min to max]
[U17] Parameter: [name] — current: [value] — range: [min to max]

Technique 4: PERSPECTIVE ROTATION — Who sees this differently?

[U18] [stakeholder 1] sees: [their version]
[U19] [stakeholder 2] sees: [their version]
[U20] [outsider] sees: [their version]
[U21] [opponent] sees: [their version]

Technique 5: ASSUMPTION EXTRACTION — What must be true for this to hold?

[U22] LOAD-BEARING: [assumption] — if false: [consequence]
[U23] LOAD-BEARING: [assumption] — if false: [consequence]
[U24] BACKGROUND: [assumption] — probably true but worth noting

Technique 6: DIMENSION DISCOVERY — What axes does this exist on?

[U25] Dimension: [name] — claim sits at: [position]
[U26] Dimension: [name] — claim sits at: [position]
[U27] HIDDEN dimension: [name] — not discussed but relevant

Technique 7: TEMPORAL VARIATION — How does this change over time?

[U28] Short-term (days): [what's true]
[U29] Medium-term (months): [what changes]
[U30] Long-term (years): [what changes more]

Technique 8: SCALE VARIATION — At what level does this hold?

[U31] Individual: [true/false/different]
[U32] Team: [true/false/different]
[U33] Organization: [true/false/different]

Don't apply all 8 mechanically. Use the ones that produce findings. Skip the ones that don't.

Claim Type	Primary Techniques	Secondary
Factual	Assumption extraction, state space	Perspective
Strategic	Instance-to-category, parameter variation, perspective	Temporal, scale
Design	State space, parameter variation, dimension discovery	Perspective
Causal	Assumption extraction, perspective, temporal	Scale
Decision	State space, instance-to-category, parameter variation	All others
Belief	Assumption extraction, perspective, dimension discovery	Temporal

---

Phase 2: FINDING REGISTRY

After exploration, compile EVERY finding into a categorized list. Nothing from Phase 1 gets left out.

FINDING REGISTRY
================

UNBUNDLED CLAIMS:
[U1] [text] — TYPE: explicit
[U2] [text] — TYPE: implicit
...

ALTERNATIVES FOUND:
[U6] [text] — SOURCE: state space
[U7] [text] — SOURCE: state space
[U12] [text] — SOURCE: instance-to-category
...

ASSUMPTIONS FOUND:
[U22] [text] — LOAD-BEARING — if false: [consequence]
[U23] [text] — LOAD-BEARING — if false: [consequence]
[U24] [text] — BACKGROUND
...

DIMENSIONS FOUND:
[U25] [text]
[U26] [text]
[U27] [text] — HIDDEN
...

PERSPECTIVES FOUND:
[U18] [text]
[U19] [text]
...

TEMPORAL/SCALE VARIATIONS:
[U28-U33] [text]
...

TOTALS:
- Unbundled claims: [N]
- Alternatives: [N]
- Assumptions: [N] ([N] load-bearing)
- Dimensions: [N] ([N] hidden)
- Perspectives: [N]

---

Phase 3: SYNTHESIS

Derived entirely from the registry. No new findings introduced here.

ORIGINAL CLAIM: [restated]

SPACE SIZE: [total unique findings from registry]

VOI RANKING (Value of Information — which findings matter most):
1. [highest-VOI finding — U-number — learning this changes the most]
2. [second highest — U-number]
3. [third — U-number]

LOAD-BEARING ASSUMPTIONS:
[List the assumptions that, if wrong, change everything — U-numbers only]

HIDDEN DIMENSIONS:
[Axes the original claim didn't mention but exists on — U-numbers only]

READY FOR:
- /ar [specific high-VOI claim from registry] — to explore what follows if true
- /aw [specific high-VOI claim from registry] — to test if it holds
- /uaua [the full analysis if warranted]

---

Depth Scaling

Depth	Min Unbundled Claims	Min Techniques	Min Total Findings
1x	3	3	12
2x	5	4	20
4x	8	5	35
8x	12	6	60
16x	18	7	100
32x	25	8	150

Default: 2x. These are floors.

---

Pre-Completion Check

• Claim restated precisely and unbundled
• All unbundled claims identified (explicit, implicit, presupposed, bundled, meta)
• Techniques applied (minimum for depth level)
• ALL findings from Phase 1 appear in registry (none dropped)
• Registry includes totals
• VOI ranking provided (references U-numbers)
• Load-bearing assumptions identified (references U-numbers)
• Synthesis introduces NO new findings — only references U-numbers
• Completeness check: Would someone from a different domain or perspective spot something you missed? If probably yes, keep going.
• Output points to next step (/ar, /aw, or /uaua on specific high-VOI items)