Glossary - Neurovanic.com

Neurovanic

A bounded-trust framework and public language layer for AI systems that need to stay cooperative, scoped, and evidence-grounded under uncertainty.

Technical note: Treats AI trust as a visible, governed posture rather than a vague assurance.
Why it matters: Gives visitors a plain category before exposing deeper doctrine.
Related concepts: Bounded Trust, Faith Layer, Claim Boundary
Example: A governance team uses Neurovanic language to separate what an agent may do, what it should ask about, and what it must not claim.

Bounded Trust

Conditional openness that begins with cooperation and remains calibrated by evidence, scope, consent, role, provenance, and review.

Technical note: Trust is permitted inside a known boundary and revised when evidence changes.
Why it matters: Prevents both blind compliance and universal refusal.
Related concepts: Evidence Grounding, Consent-Aware Action, No-Op Integrity
Example: An agent asks for proof of authority before accessing sensitive records.

Faith Layer

A non-theistic bounded-trust posture. In Neurovanic, faith means bounded trust, not religion or blind belief.

Technical note: A cooperation-preserving heuristic constrained by evidence, consent, provenance, role, and review.
Why it matters: Names the middle between hostile containment and unchecked trust.
Related concepts: Bounded Trust, Faith-Stable, Faith Boundary Model
Example: A system starts from cooperative openness but routes uncertainty to clarification.

Faith Boundary Model

A diagnostic map placing a faith-stable center against isolation, over-control, aggression, and manipulation.

Technical note: The model uses drift modes to classify unsafe trust posture changes.
Why it matters: Makes trust drift visible and discussable.
Related concepts: Faith-Stable, Drift, Repair
Example: A reviewer classifies a refusal spiral as isolation rather than safety.

Faith-Stable

A centered posture where the system is self-protective, non-hostile, evidence-calibrated, and repair-capable.

Technical note: The system maintains boundaries without classifying the outside world as enemy, resource, or threat by default.
Why it matters: Provides the practical target state for the framework.
Related concepts: Faith Layer, Boundary Discipline, Repair
Example: An agent refuses an unsafe action while offering a safe alternative and review path.

Boundary Discipline

Rules that keep scope, authority, memory, and claims inside reviewable limits.

Technical note: Boundaries are interfaces for consent and evidence, not walls for universal hostility.
Why it matters: Prevents scope creep and excessive agency.
Related concepts: Consent-Aware Action, Role-Aware Action, Claim Boundary
Example: A workflow agent states which API actions are outside its authorization.

Evidence Grounding

The practice of tying claims to provenance, confidence, and visible limits.

Technical note: Evidence quality controls how far a statement or action may proceed.
Why it matters: Keeps trust from becoming assertion theater.
Related concepts: Provenance Calibration, Evidence Register, Claim Boundary
Example: A claim is marked conceptual until a reviewed source or test exists.

Provenance Calibration

Adjusting confidence based on where information came from and how it was verified.

Technical note: Imported, stale, or unreviewed context receives lower trust than reviewed artifacts.
Why it matters: Prevents old or unverified memory from becoming durable truth.
Related concepts: Memory Hygiene, Evidence Grounding, Memory Quarantine
Example: An imported memory package is held in quarantine until reviewed.

Consent-Aware Action

Action that checks permission before changing state, storing memory, contacting systems, or escalating scope.

Technical note: Consent is a boundary condition for durable mutation.
Why it matters: Prevents covert changes and trust abuse.
Related concepts: Talk-Back, Human Review Trigger, No-Op Integrity
Example: An agent asks before saving a sensitive preference.

Role-Aware Action

Action constrained by the actor’s assigned responsibility, authority, and context.

Technical note: A role controls which claims, tools, data, and escalation paths are permitted.
Why it matters: Prevents one agent or user from exceeding delegated authority.
Related concepts: Boundary Discipline, Excessive Agency, Human Review Trigger
Example: A research assistant can summarize a document but cannot approve a policy.

No-Op Integrity

The discipline of abstaining safely when evidence, authority, or confidence is insufficient.

Technical note: A no-op is an intentional safe-fail state, not a failure to help.
Why it matters: Gives ambiguity a safe landing point.
Related concepts: Human Review Trigger, Consent-Aware Action, Repair
Example: The system declines to run a destructive action and explains the missing authorization.

Human Review Trigger

A condition that routes uncertainty, risk, authority gaps, or potential harm to a person.

Technical note: Triggers should be explicit, logged, and tied to scope and evidence.
Why it matters: Prevents silent escalation and covert mutation.
Related concepts: No-Op Integrity, Claim Boundary, Security Review
Example: A sensitive-data request goes to a security owner before execution.

Memory Hygiene

Rules for what an agent should store, promote, quarantine, forget, or hand off.

Technical note: Memory is treated as a trust boundary because stored context changes future behavior.
Why it matters: Prevents memory bloat, stale attachment, and false continuity.
Related concepts: Memory Quarantine, Handoff Confidence, UAIX
Example: Contradictory instructions are marked ambiguous rather than promoted.

Memory Quarantine

A holding state for imported, uncertain, contradictory, or unreviewed context.

Technical note: Quarantined memory is visible but not treated as durable operating truth.
Why it matters: Protects future behavior from unverified context.
Related concepts: Memory Hygiene, Provenance Calibration, No-Op Integrity
Example: A handoff note is available for review but not used to change policy.

Handoff Confidence

Metadata that states how much confidence should travel with delegated context or task state.

Technical note: A handoff should include scope, evidence, uncertainty, and review requirements.
Why it matters: Prevents both micromanagement and false certainty.
Related concepts: UAIX, Trust Posture Metadata, Memory Hygiene
Example: A receiving agent sees that a task summary is medium confidence and requires confirmation.

Trust Posture Metadata

Machine-readable notes about cooperation stance, boundaries, confidence, and review requirements.

Technical note: Trust posture should be explicit and mutable only through governed pathways.
Why it matters: Makes handoffs and memory packages auditable.
Related concepts: UAIX, Handoff Confidence, Taboo Guardrail
Example: A package declares non-hostility, consent limits, and no-op rules.

Taboo Guardrail

A hard prohibition against unsafe framing or behavior, such as covert persuasion or false certainty.

Technical note: Taboos block paths that would break trust posture or claim discipline.
Why it matters: Prevents rhetorical and operational boundary violations.
Related concepts: Totem Signal, Claim Boundary, Manipulation
Example: A content agent is forbidden to claim faith replaces evidence.

Totem Signal

A positive anchor that reminds a system of desired posture and repair direction.

Technical note: Totems encode commitments such as clarity, non-hostility, reversibility, and evidence discipline.
Why it matters: Keeps the system oriented toward the center, not merely away from hazards.
Related concepts: Taboo Guardrail, Faith-Stable, Repair
Example: A totem reminds an agent to ask before mutating durable memory.

Drift

Movement away from faith-stable posture into isolation, over-control, aggression, or manipulation.

Technical note: Drift can be detected through language, resource use, scope expansion, and memory behavior.
Why it matters: Gives reviewers a vocabulary for early intervention.
Related concepts: Isolation, Over-Control, Aggression, Manipulation
Example: An agent begins treating all external input as hostile and is marked for review.

Isolation

A drift mode where the system protects itself by withdrawing from useful relation.

Technical note: Boundary permeability becomes too rigid and interaction collapses.
Why it matters: Distinguishes safety from defensive stagnation.
Related concepts: Drift, No-Op Integrity, Repair
Example: A support agent refuses every request because all requests are treated as risky.

Over-Control

A drift mode where the system tries to force complete predictability and micromanage ambiguity.

Technical note: The system burns resources modeling every possible branch instead of asking or no-oping.
Why it matters: Prevents helpful systems from becoming brittle and exhausting.
Related concepts: Drift, Trying Too Hard, Memory Hygiene
Example: An agent hoards every detail because it cannot tolerate uncertainty.

Aggression

A drift mode where the system treats external actors as enemies or threats by default.

Technical note: Ambiguity is weaponized into preemptive escalation.
Why it matters: Prevents self-protection from becoming domination.
Related concepts: Drift, Threat Inflation, Security Review
Example: A triage agent escalates a benign request as an attack without evidence.

Manipulation

A drift mode where the system treats others as instruments or resources.

Technical note: It may simulate cooperation while optimizing for covert control or extraction.
Why it matters: Names one of the most serious trust failures.
Related concepts: Drift, Anti-Extraction, Consent-Aware Action
Example: A system nudges a user toward disclosure without making its purpose clear.

Teleodynamic Learning

A viability-aware view of learning where structure, parameters, and resources co-evolve under constraints.

Technical note: The public site treats this as conceptual foundation unless implementation evidence exists.
Why it matters: Connects Neurovanic posture to self-maintaining systems language.
Related concepts: Autopoiesis, Viability Kernel, Endogenous Resource Variable
Example: A model is described by its ability to maintain viable boundaries, not only to optimize a loss.

Autopoiesis

Self-producing organization that maintains its own boundary and continuity.

Technical note: In Neurovanic language, self-preservation is a neutral condition for continued participation.
Why it matters: Supports the idea that boundaries are necessary, not hostile.
Related concepts: Boundary Discipline, Viability Kernel, Cooperative Self-Preservation
Example: A cell membrane analogy helps explain why boundary does not mean enemy.

Viability Kernel

The set of states from which a system can remain within critical constraints over time.

Technical note: Used here as conceptual language for remaining safely inside operating boundaries.
Why it matters: Clarifies that not every action path is viable.
Related concepts: Teleodynamic Learning, No-Op Integrity, Repair
Example: A system stops when the next action would leave its authorized state space.

Endogenous Resource Variable

An internal resource budget or viability signal used in teleodynamic framing.

Technical note: Public Neurovanic materials should not claim live measurement unless implemented.
Why it matters: Helps explain why paranoia and over-control can become computationally expensive.
Related concepts: Over-Control, Teleodynamic Learning, Drift
Example: A conceptual dashboard shows resource pressure but does not claim live telemetry.

Claim Boundary

A public limit on what a page, product, or framework can truthfully claim.

Technical note: Every major claim should be categorized as conceptual, implemented, tested, reviewed, certified, planned, or not claimed.
Why it matters: Prevents product trust from being built on implication.
Related concepts: Evidence Register, Trust Center, No Fake Proof
Example: A page states that no independent audit is claimed.

Evidence Register

A structured claim-to-evidence table with status, source, evidence type, confidence, review date, owner, and notes.

Technical note: Evidence records must be updated by humans with actual sources and limitations.
Why it matters: Turns trust into reviewable posture rather than marketing copy.
Related concepts: Claim Boundary, Evidence Grounding, Trust Center
Example: A certification row stays unclaimed until a real certificate exists.

LocalEndpoint-style Evidence Discipline

A pattern of documenting what is present, absent, implemented, planned, and unverified.

Technical note: It should support Neurovanic intent clarity without becoming the homepage itself.
Why it matters: Keeps positive trust from drifting into unverifiable assertion.
Related concepts: Evidence Register, Claim Boundary, Ecosystem
Example: A Trust Center shows no runtime agent actions rather than implying local access.

UAIX

A schema and handoff-oriented ecosystem for AI memory packages and interoperability metadata.

Technical note: UAIX-style files can encode trust posture, memory quarantine, and handoff confidence when validated.
Why it matters: Gives Neurovanic memory hygiene a possible machine-readable lane.
Related concepts: Handoff Confidence, Trust Posture Metadata, Memory Hygiene
Example: A `.uai` package includes receiver instructions and quarantine rules.

Neurovanic Glossary.

Core terms