369BEHIND THE VEIL
NETWORK NODE / PRODUCTS

Products & systems

Tools that expose
their own authority.

We build systems that make evidence, permissions, limits, and uncertainty visible—before software is allowed to act.

NETWORK SCOPE04 SYSTEMS
01 control plane02 intelligence systems01 deterministic world
MODEL OUTPUT ≠ AUTHORITYEVIDENCE BEFORE EXECUTIONFAIL CLOSEDMEMORY MUST REMAIN REVISABLE

Product registry

Built in the open.
Labeled honestly.

Every system carries an explicit development status. A prototype is called a prototype. Architecture is not marketed as deployment.

ANL-001Working in development

Epistemic control plane

AETERNA

A working development implementation of an evidence-bound agent control plane. S1R4 Phase 2 is accepted and human-ratified with recorded limitations.

  • Evidence closure
  • Capability sealing
  • Replay resistance
  • Immutable audit history
READINESSPhase 2 accepted · development use
ANL-002Prototype

Sovereign local intelligence

SAMARA

A continuity-first AI system designed around human-controlled identity, plural governance, and auditable long-term memory.

  • Local-first identity
  • Revisable memory
  • Council governance
  • Explicit permissions
READINESSPrivate development
ANL-003Architecture

Governance framework

CONCORDIA

A structured coordination layer for dissent, synthesis, role boundaries, and accountable decisions across human–AI systems.

  • Plural reasoning
  • Dissent preservation
  • Role boundaries
  • Decision provenance
READINESSSpecification stage
ANL-004Active build

Deterministic game system

BLACK FLAME

An offline-first strategy and narrative world built without cloud dependency, uncontrolled randomness, or hidden simulation state.

  • Offline-first
  • Deterministic combat
  • Auditable state
  • Narrative systems
READINESSPlayable development

Inside AETERNA

One causal chain.
No invisible leap.

01Evidence

VERITAS

Claims, provenance, falsification, and evidence roots

02Authority

AETHER

Policy evaluation, scope, delegation, and expiry

03Action

SENTINEL

Sealed capabilities, replay defense, and fail-closed execution

04History

PHOENIX

Immutable records of what happened and why

05Visibility

OBSERVATORY

Read-only causal views across the full decision chain

Release standard

What must be true before we call it ready.

  1. 01Evidence boundaryPrimary evidence cannot be manufactured by the model evaluating it.
  2. 02Execution boundaryNo valid capability, no action—regardless of confidence or intent.
  3. 03Provenance boundaryEvery consequential result remains traceable to inputs, policy, and actor.
  4. 04Failure boundaryAmbiguity, expiry, replay, or missing state resolves to denial.

THE NETWORK IS UNDER ACTIVE CONSTRUCTION

See what we are
building next.

DEVELOPMENT LOGS / ROADMAPS / EVIDENCEOPEN .WORK →