I · Failure Cascade

Problem Framing: Govern Decision Entropy

The system begins with what capital is allowed to do, not with which asset looks most attractive. Financial markets fail when available intelligence exceeds the allocator's ability to govern action. Crypto intensifies this through continuous trading, fragmented liquidity, reflexive leverage, narrative contagion, protocol risk, and rapid capital migration.

M.A.I.C. compresses conflicting evidence into a bounded capital posture. A prediction is not an authorization. A ranking is not an authorization. A favorable narrative is not an authorization. Constraint is infrastructure: the system must be able to wait, reduce, rotate, preserve, or deny deployment.

II · Separation of Powers

Governance First: Observation Is Not Authority

M.A.I.C. separates market observation, analytical interpretation, conflict arbitration, and capital authorization. Analytical components may disagree and remain visible as disagreement. Final portfolio authority is logically separate from those views.

Invariant: asset attractiveness cannot create its own upstream portfolio-risk permission. Observation measures evidence. Analytical views interpret it. Constraints preserve disagreement and prior-state restrictions. Final Mode Authority owns governed capital posture. PAS only qualifies identities inside approved crypto buckets.

I / CONSTITUTIONAL CONTROLSeparation of powers is the architecture

The order is a governance constraint, not an exposed model algorithm.

RANKING≠PERMISSIONNARRATIVE≠PERMISSION
NON-NEGOTIABLE BOUNDARY

Evidence, including rankings and attractive narratives, cannot give itself capital permission. A downstream selector cannot enlarge the authorized portfolio.

Inspect the source-owned Authority ↗
III · Capital Location

Final Mode Authority: Multi-Asset Capital Routing

Final Mode Authority is the current portfolio-performance and capital-routing owner. M.A.I.C. remains crypto-native, but its model portfolio is not crypto-only.

Authority may route model capital among portfolio-eligible Cash, Gold, Silver, Equities, and Crypto when governed conditions permit. Treasury and rates are observation-only macro evidence: they may affect authority but Treasury is not a current portfolio destination.

Routing law: Crypto Universe is one downstream branch, not the destination of every capital-flow path.

IV · Crypto-Only Handoff

Crypto Branch: Qualification After Permission

Only capital authorized to the Crypto sleeve enters the crypto qualification pipeline. Cash, Gold, Silver, and Equities can remain valid portfolio destinations without passing through Crypto Universe.

  1. Crypto Universe
  2. Liquidity Filter
  3. Narrative Filter
  4. Survivability Filter
  5. Portfolio Bucket
  6. Protected Selection / PAS
II / CAPITAL LOCATIONEligible destinations are not the same as the Crypto branch

Four distinct destination categories are shown schematically; Gold and Silver are individually eligible metals. Only permitted Crypto continues to Crypto Universe.

MACRO EVIDENCETREASURY / RATESObservation only · never a current portfolio destination
FINAL MODE AUTHORITYPERMITTED CAPITAL DESTINATIONSBranches indicate eligibility, not current allocation or capital-flow volume.
01CASHEligible destinationNo Crypto Universe handoff
02GOLD / SILVEREligible destinationsNo Crypto Universe handoff
03EQUITIESEligible destinationNo Crypto Universe handoff
04CRYPTOConditional qualification route
CRYPTO UNIVERSEQUALIFICATIONPAS / PROTECTED
NON-NEGOTIABLE BOUNDARY

Treasury and rates inform risk interpretation but are not live portfolio destinations. Selecting a research asset never adds a fifth approval channel.

Inspect qualification under the existing mandate ↗
V · PAS Current State

Protected Asset Selection: Operational, Downstream, Frozen

PAS is implemented and operational. It is not a future or pending model. Its role is deliberately narrower than Authority: PAS selects identities inside Authority-approved crypto buckets and preserves identity continuity.

  • First governed ALT OPEN / grant: run PAS and freeze the entry basket.
  • TAKE CAP: reduce proportionally; do not replace frozen identities merely because the market moved.
  • Next governed REBUY: PAS may replace the basket under active model law.
  • Observation or Governance views cannot trigger PAS; the canonical Authority lifecycle owns activation.
  • Exact selected identities, rankings, rejected candidates, weights, and formulas remain protected.

Ownership: Final Mode Authority decides whether capital may enter the route. PAS decides which eligible identities occupy the authorized crypto buckets.

VI · Human-Facing Directives

Public State Semantics

OFFENSE, CORE, STANDBY, and DEFENSE are public presentation labels for governed states. They summarize operational consequence; they do not replace the backend machine contract or expose allocation law.

  • OFFENSE: broad governed deployment is open across eligible sleeves; crypto qualification expands only if Crypto is authorized.
  • CORE: core / bluechip exposure is prioritized under the active multi-asset route; effective identity can differ by relative leadership.
  • STANDBY: no new deployment; existing governed structure may remain.
  • DEFENSE: higher-beta risk is reduced; capital routes toward defensive eligible sleeves and cash.
  • DEGRADED: normal authority is constrained because evidence is stale, incomplete, unavailable, or malformed.

Fail closed: missing data is not zero. Degraded evidence must not be promoted into fresh authority.

VII · Time Ownership

Evidence Cadence: Observation vs Confirmation

M.A.I.C. separates the publication clocks. A synchronized 4H evidence boundary can update early observations and protected interpretation. Weekly confirmation belongs to the BTC-first Governance path and must not be mistaken for multi-asset Final Mode Authority. Final Mode Authority alone owns portfolio-capital permission, according to its separate canonical source and release policy.

Time law: early evidence is not the confirmed public state.

III / TIME AND EVIDENCE OWNERSHIPDifferent clocks; separate evidence populations

A faster observation cannot change the meaning of a confirmed governance posture or silently create new Final Authority.

014H
Early evidence boundaryFaster evidence and protected interpretation
02WEEKLY
Governance confirmationBTC-first governed posture; not Final Authority
03FINAL RELEASE
FINAL MODE AUTHORITYSeparate multi-asset capital permission and publication owner
POINT-IN-TIME OOSCausal historical research record
≠
FROZEN LIVE FORWARDForward-only after the published freeze
NON-NEGOTIABLE BOUNDARY

The weekly Governance signal is not the multi-asset Final Mode Authority. OOS and frozen live-forward remain separate populations and must not be joined into realized investor returns.

Inspect population methodology ↗
VIII · Evidence Populations

Performance Evidence: OOS and Frozen Live Forward

Performance ownership belongs to Final Mode Authority with downstream PAS identities. Historical research replay is multi-asset and causal: a decision may affect only a subsequent interval and future information is excluded from the decision set.

  • Point-in-Time OOS: causal multi-asset Authority + PAS research replay beginning 2023-01-01; dataset-limited disclosure retained.
  • Frozen Live Forward: forward-only append-only model record after the Generation-1 freeze on 2026-09-30T14:35:19Z; no pre-freeze live-return backfill.
  • Frontend role: render canonical backend publication; do not rerun selectors or invent missing values.

OOS and Frozen Live Forward are separate evidence populations. Historical research evidence is not audited fund NAV, realized investor P&L, or a guarantee of future performance.

IX · Discoverability Boundary

Access Architecture: Capability Is Not Entitlement

M.A.I.C. distinguishes whether a capability exists from whether a user is entitled to access it. PAS can be operational while exact PAS output remains protected. Describing a higher access class does not itself grant access.

  • Class 01 · Public Research: confirmed posture and public evidence; no protected identities or allocator routing.
  • Class 02 · Tactical Portfolio: confirmed multi-asset portfolio route; earlier model inference and PAS identities are not public.
  • Class 03 · Allocation Intelligence: earlier/deeper protected intelligence and operational PAS access; server authorization required.
  • Class 04 · Institutional Mandate: same intelligence ceiling plus private mandate review; mandate execution/custom controls are not generally active.

Cross-user Intelligence Relay remains inactive until real identity and persistent quota/release records exist.

X · Control Boundary

Non-Custodial Boundary: No External Execution Authority

M.A.I.C. governs model state and route classification inside the software. It does not custody funds, hold user assets, place orders, operate as an exchange or broker, or control external accounts. Capital authorization inside M.A.I.C. is software governance, not authority over a user or external account.

IV / INDEPENDENT RIGHTS BOUNDARIESFour questions. No automatic delegation.

Visibility, model permission and external account control are different legal and technical questions.

01 / CAPABILITYDoes the function exist?Operational PAS can exist without public identities.
02 / ENTITLEMENTCan this user inspect it?Class, source release and server access govern visibility.
03 / MODEL PERMISSIONFINAL AUTHORITYA market decision does not come from the access class.
04 / EXTERNAL CONTROLNO ORDER / NO CUSTODYPersonal execution remains outside the service.
NON-NEGOTIABLE BOUNDARY

Purchasing or receiving protected access neither changes the market state nor grants M.A.I.C. custody, discretionary trading rights or external account authority.

Inspect access and operational limitations ↗
XI · Dual-Layer System

M.A.I.C. and A.N.I.M.A.: Separate Jobs

M.A.I.C. answers the governed decision problem: what, when, and which governed route. A.N.I.M.A. answers the interpretation problem: why, where, who, and how. Open-ended interpretation does not silently become capital authority, and capital authority does not need to collapse the market into one narrative.

XII · Public Claim Discipline

Category Precedence, Versioning, and Limitations

M.A.I.C. is publicly positioned as the world's first crypto-native capital operating environment under versioned Category Evidence Method V1.1. The category claim is defined by explicit criteria and maintained as a challengeable precedence record, not treated as proven merely by branding.

Public Doctrine 2026.10 supersedes 2026.08 for current architecture while older documents remain historical evidence. Exact formulas, weights, thresholds, rankings, identities, and protected route logic remain proprietary. Operational capability does not imply universal entitlement, billing availability, custody, or execution.

Appendix · Public Invariants

Canonical Public Invariants

  • Observation is not authority.
  • Final Mode Authority owns portfolio permission and performance identity.
  • Treasury is observation-only; Cash, Gold, Silver, Equities, and Crypto are portfolio-eligible according to governed conditions.
  • Only the Crypto route enters Crypto Universe.
  • PAS is operational and downstream; PAS never creates upstream permission.
  • OOS research and Frozen Live Forward remain separate evidence populations.
  • Missing data is not zero; degraded evidence fails closed.
  • M.A.I.C. is non-custodial and has no external execution authority.
  • Protected implementation is not a substitute for public evidence.
  • Capability existence does not equal user entitlement.

Current owner: FINAL_MODE_AUTHORITY_WITH_DOWNSTREAM_PAS_IDENTITIES

Four disciplines, one permission boundary. Probability Engine evaluates distributions rather than prophecy; Risk Architecture governs exposure; Intelligence Topology compresses relationships rather than turning rankings into orders; Defense Doctrine stops at authorization, outside custody and execution. A.N.I.M.A. expands interpretation, while M.A.I.C. retains capital permission. Conflicting evidence may delay or deny authorization; historical replay cannot authorize new live risk.