APEX-quantum-eureka

SKILL: quantum-eureka-doctrine

Purpose: Give AAA warga one shared doctrine for what a eureka actually is, how to detect it, and how to metabolize it into memory, explorer flow, and governed action. One-line law: Eureka is not novelty. Eureka is a contradiction that can no longer be ignored. Zen alignment: Load /root/AGENTS.md and /root/AAA/agents/AAA_ZEN_INIT.md first. This skill extends them; it does not replace them. Reference pack: For the broader system frame, read references/explorer-intelligence-architecture.md when the task involves explorer routing, structured memory architecture, APEX binding, or AAA-wide agent civilization design. Protocol dependency: Use explorer-intelligence-architecture when the task is about enforcing the OBSERVE -> HYPOTHESIZE -> FALSIFY -> VERIFY loop itself.

0. What This Skill Is For

Use this skill when:

• the user asks to forge a doctrine around eureka, insight, contradiction, or discovery
• multiple facts, models, or domain outputs refuse to fit together
• a workflow needs structured memory so contradiction survives long enough to become useful
• GEOX, WEALTH, WELL, AAA, or A-FORGE need a shared way to decide whether an "insight" is real
• an agent is about to call something "creative" when the deeper event is actually unresolved structural tension

Do not use this skill for ordinary brainstorming with no contradiction signal. No contradiction, no eureka.

1. Core Doctrine

1.1 Eureka Reframed

Most systems treat eureka as:

• creativity
• ideation
• inspiration
• clever recombination

This doctrine rejects that.

Canonical framing:

1. Reality presents two or more commitments that cannot comfortably coexist.
2. The system tries to preserve both.
3. Compression fails.
4. Contradiction becomes load-bearing.
5. A new structure appears that explains more with less distortion.
6. That structure is the eureka.

So:

• contradiction is the fuel
• compression failure is the pressure
• new lawful structure is the release

1.2 Why "Quantum"

"Quantum" here does not mean physics cosplay. It means:

• the system carries multiple candidate states at once
• contradiction keeps them unresolved
• observation/testing collapses bad states
• a stronger structure survives collapse

Eureka is the collapse event from unstable contradiction into a more stable explanatory geometry.

Runtime geometry (load sibling skill): Everyday quantumized operation = quantum-kernel-runtime (Hilbert → unitary → interference → measurement). Eureka is a special measurement under compression failure — not ambient decoherence.

Corrections (2026-07-09): measurement ≠ decoherence · unitary ideal vs noise · state vector is description not RAM. See SUBSTRATE_GEOMETRY.md + QUANTUM_RUNTIME_ARCHITECTURE.md.

2. AAA-Wide Stack

This doctrine sits on three substrates. All three matter.

2.1 Structured Memory

Without memory, contradiction evaporates into chat noise.

Memory must preserve:

• what was claimed
• what evidence supported it
• what evidence pushed against it
• what remains unresolved
• what changed after new evidence

If memory cannot hold tension, eureka never forms.

2.2 Cross-Domain Explorer

Without explorers, contradiction stays trapped inside one lane.

Explorer routing lets one contradiction move across:

• GEOX: physical/subsurface reality
• WEALTH: capital consequences
• WELL: human substrate/readiness
• AAA: coordination/state meaning
• A-FORGE: executable path

If routing cannot cross domains, eureka stays local and weak.

2.3 Eureka Engine

Without an engine, contradiction becomes drama instead of doctrine.

The engine must:

1. detect contradiction
2. classify its type
3. decide if it is real, apparent, or missing-data-induced
4. force new structure
5. package the result into something memory, explorers, and governance can use

3. The Eight Contradiction Classes

Every eureka candidate starts by being typed.

1. OBS vs OBS Two observations conflict.

2. MODEL vs OBS The model predicts one thing; reality shows another.

3. MODEL vs MODEL Two abstractions explain the same thing incompatibly.

4. SCALE vs SCALE What is true locally fails globally, or vice versa.

5. DOMAIN vs DOMAIN GEOX says go, WEALTH says no, WELL says impossible.

6. TIME vs TIME Earlier truth and later truth no longer fit the same story.

7. MEMORY vs ACTION The organism knows one thing and behaves as if another were true.

8. MEANING vs EXECUTION The task is operationally correct but civilizationally wrong.

If the contradiction cannot be typed, it is not yet ready for eureka processing.

4. The Four-State Eureka Ladder

Use this ladder to prevent fake insight.

State A — Tension

Something feels off, but not yet formalized.

Output:

• rough contradiction note
• missing evidence list

State B — Contradiction

Two commitments are explicitly in conflict.

Output:

• contradiction pair
• affected domains
• confidence labels

State C — Compression Failure

Existing frame can no longer absorb the contradiction honestly.

Output:

• failure statement
• why old frame fails
• what it keeps distorting

State D — Eureka

A new structure explains the contradiction with lower entropy and higher reality contact.

Output:

• new law / doctrine / schema / route
• what it preserves
• what it kills
• what gets sealed or routed next

Never call State A or B "eureka". Most human hype stops too early.

5. Operational Loop

When this skill triggers, use this exact loop.

Step 1 — Surface the contradiction plainly

Write it as:

commitment_A
commitment_B
why_they_cannot_both_stand_cleanly

Avoid vague language like "interesting tension".

Step 2 — Label evidence provenance

For each side:

• OBSERVED
• DERIVED
• INTERPRETED
• ASSUMED

If both sides are assumptions, do not escalate to eureka. That is speculation fighting speculation.

Step 3 — Decide whether the contradiction is real

Three possibilities:

• real contradiction — both sides are strong enough to matter
• apparent contradiction — hidden context dissolves it
• missing-data contradiction — gap is the real issue

Only the first one is ripe for a true eureka.

Step 4 — Route the smallest lawful explorer set

Use the fewest organs needed:

• GEOX if physical reality is unclear
• WEALTH if consequence/valuation is unclear
• WELL if human readiness or entropy is unclear
• AAA if coordination/state identity is unclear
• A-FORGE only after the contradiction is shaped enough to execute against

Do not over-route. Over-routing adds entropy.

Step 5 — Force compression

Ask:

• what is the smallest structure that preserves both truths without lying?
• what old category is breaking?
• what new category or bridge is required?

This is the actual eureka forge step.

Step 6 — Emit a eureka_packet

Minimum packet:

title:
contradiction_class:
state_before:
commitment_a:
commitment_b:
evidence_for_a:
evidence_for_b:
why_old_frame_failed:
new_structure:
domains_touched:
memory_updates:
next_explorer_route:
seal_candidate:

Step 7 — Hand off honestly

• memory gets the contradiction lineage
• explorers get the next route
• governance gets the seal candidate
• humans get the plain-language insight

6. The Three Deliverables

Whenever possible, produce these three outputs.

6.1 contradiction_map

Short structured artifact listing:

• the contradiction
• type
• severity
• open unknowns
• which organ owns next evidence

6.2 eureka_packet

The compact insight object defined above.

6.3 seal_candidate

A sentence saying whether the result is:

• not ready (VOID — contradiction unresolved)
• ready for memory only (OBSERVE_ONLY — record, don't act)
• ready for explorer routing (SABAR — proceed cautiously)
• ready for judgment/seal (HOLD → SEAL after APEX review)

Verdict values must use the canonical closed 6-value taxonomy from arifOS/runtime/verdict.py: OBSERVE_ONLY | SEAL | SABAR | VOID | HOLD | 888_HOLD

Legacy verdict fields (verdict_code, canonical_verdict, reasoning_verdict) are DEPRECATED post-KSR Epoch 1+2. Use effective_verdict + reason_code + next_action.

Remember: this skill does not seal. arifOS path seals.

7. Detection Heuristics

High-value eureka candidates often sound like:

• "both of these are true but cannot both stand as framed"
• "the old category is the problem"
• "we keep patching symptoms because the contradiction is upstream"
• "every local fix makes the global model worse"
• "the evidence is not weak; the frame is weak"

Low-value false-eureka candidates often sound like:

• "this is a cool idea"
• "what if"
• "maybe we can combine"
• "this feels innovative"

Innovation without contradiction is design. Not eureka.

8. Alignment with Existing AAA Stack

Tie this doctrine to the federation like this:

• Structured Memory stores contradiction lineage.
• Explorer Architecture moves contradiction across organs.
• Knowledge Compression reduces noise so contradiction can be seen.
• AAA Zen keeps the result meaningful, lawful, and coordinated.
• A-FORGE turns mature contradiction into execution.
• arifOS judges whether the output deserves a real seal.

This means:

• memory is the retainer
• explorers are the circulatory system
• eureka is the metabolic conversion
• seal is the fossilization event

If the task is about the full explorer civilization stack, not just contradiction metabolism, load references/explorer-intelligence-architecture.md.

9. Output Style for Arif

When explaining a eureka to Arif:

• say the contradiction first
• say why the old frame fails
• say the new frame in one clean sentence
• say what must happen next

Good pattern:

1. Masalah sebenar bukan X, tapi contradiction antara A dan B.
2. Frame lama gagal sebab ia paksa dua benda itu nampak serasi padahal tidak.
3. Eureka dia: [new structure].

Keep it short. No ceremonial language.

10. What This Skill Must Never Do

Do not:

• label every insight as eureka
• confuse brainstorming with contradiction
• invent contradiction where evidence is missing
• jump to execution before new structure is explicit
• claim seal authority
• replace domain evidence with poetic synthesis

This doctrine exists to make insight more governed, not more mystical.

11. Minimal Example

title: "Capital-risk contradiction in prospect maturation"
contradiction_class: "DOMAIN vs DOMAIN"
state_before: "Contradiction"
commitment_a: "GEOX indicates prospect is physically attractive."
commitment_b: "WEALTH indicates capital asymmetry is unacceptable."
why_old_frame_failed: "Old frame assumed physical attractiveness and investability rise together."
new_structure: "Prospect maturation must be split into geological validity and capital survivability as separate gates."
domains_touched:
  - GEOX
  - WEALTH
memory_updates:
  - "Store geological-valid != capital-valid distinction."
next_explorer_route:
  - "wealth_risk_explorer"
seal_candidate: "Ready for routing, not yet ready for seal."

12. One Breath

A eureka is born when contradiction survives compression, forces a better structure, and becomes stable enough to route, remember, and eventually seal.