Planned semantic intent layer

Logos Lexicon

Bounded intent tokens for policy resolution

Logos Lexicon is the advanced layer after VINAC-FM proves the physical authorization path. It gives policies a compact vocabulary for what an agent wants to do without passing arbitrary instructions.

Four Foundational Principles

Logos Lexicon is not a command language. It is a compact vocabulary of pre-agreed semantic tokens resolved locally without instruction parsing.

Bounded Intent, Not Instructions

Agents transmit compact intent references. The receiving system resolves meaning locally from a versioned logic library.

Transport-Agnostic

The same state-frame model can ride over VINAC-FM, optical, UWB, NFC, wired local buses, or conventional network transport.

Policy Before Execution

Every token is checked against version, context, parameter bounds, nonce or session reference, authorization level, and local policy.

Local Logic Libraries

Logic libraries are pre-agreed and locally stored. No remote code execution. No prompt parsing. Meaning is deterministic and auditable.

The Semantic Coordination Stack

Logos Lexicon sits between cloud policy and local execution. The transport proves delivery or presence; Logos defines meaning.

Cloud Policy Layer

Projects, protected endpoints, policies, receipts

Logos Lexicon 2.0

Intent IDs, context bits, parameters, nonce refs

This layer

Local Logic Library

Versioned handlers and safety constraints

Transport Abstraction

VINAC-FM · Optical · UWB · NFC · Wired · Network

How It Works

From policy-bound intent to deterministic local resolution

01

State-Frame Token Generation

Vocabulary version, intent ID, context bits, parameter indexes, and nonce or session references become a compact state frame.

02

Transport Validation

VINAC-FM, NFC, QR, optical, wired, or network transport supplies delivery and proof metadata. The token itself stays transport-independent.

03

Local Logic Library Resolution

The receiving machine maps the token to a pre-approved handler and enforces context, parameter ranges, risk level, required authorization, and audit labels.

Logos 2.0 Seed Vocabulary

Eight starter intent references for approvals, sessions, agent execution, receipts, and policy bundle handoff.

Token IDNamespaceRisk TierDescription
LL_APPROVE_001approval.corehighApprove a pre-registered action hash
LL_DENY_001approval.coremediumDeny a pending request with reason code
LL_CONFIRM_001approval.corehighRequest confirmation for a bounded action
LL_EXEC_001agent.executionhighExecute a pre-approved local action
LL_HALT_001agent.executioncriticalStop or quarantine an active agent flow
LL_SESSION_001session.controlmediumStart or renew a constrained session
LL_ACK_001receipt.controllowAcknowledge receipt of a signed approval
LL_POLICY_001policy.controlcriticalApply a versioned policy bundle reference

Developer SDK

@silentauth/logos-lexicon for local token generation, policy checks, and deterministic resolution

Token Generator

Create state-frame tokens from intent, context, parameters, and nonce references with a single call

Policy Evaluator

Evaluate any Logos intent against your local logic library and return allow, deny, or required proof

Namespace Resolver

Resolve namespaced actions such as approval.core or agent.execution to local execution policy

Schema Validation

Validate intent declarations against the Logos 2.0 state-frame schema before transmission

Vocabulary Client

Fetch and cache the active vocabulary from the SilentAuth API with full TypeScript types

Logic Library SDK

Bootstrap a local logic library in any Node.js or browser environment with zero dependencies

Integration Example

Token Generation: Agent A

import { LogosLexicon } from '@silentauth/logos-lexicon';

const lexicon = new LogosLexicon({
  publicKey: 'pk_live_…',
  version: '2.0.0',
});

// Generate a state-frame token for a pre-approved action
const token = await lexicon.generate({
  intent: 'LL_EXEC_001',
  context: {
    session_id: 'sess_abc123',
    agent_id:   'agent_local',
    timestamp:  Date.now(),
  },
  parameters: {
    action_hash: 'sha256:...',
    policy_hash: 'sha256:...',
  },
});

// token.id -> 'LL_EXEC_001'
// token.state_frame -> compact bounded reference
// token.version -> '2.0.0'

Policy Resolution: Agent B (Receiver)

import { LogosLexicon } from '@silentauth/logos-lexicon';

const lexicon = new LogosLexicon({
  publicKey: 'pk_live_…',
  version: '2.0.0',
});

// Resolve incoming token against local logic library
const result = await lexicon.resolve(receivedToken);

if (result.allowed) {
  console.log(result.action);
  // {
  //   label:          'authorize_agent_action',
  //   namespace:      'agent.execution',
  //   required_layer: 4,
  //   risk_tier:      'high',
  //   execution_policy: {
  //     max_uses: 1,
  //     rate_limit_per_hour: 20,
  //     approval_window_seconds: 300
  //   }
  // }
}

Policy Evaluation: standalone, no network required

import { evaluatePolicy } from '@silentauth/logos-lexicon';

// Evaluate offline, logic library is locally bundled
const decision = evaluatePolicy({
  token_id: 'LL_APPROVE_001',
  context:  { action_hash: 'sha256:...', vinac_level: 3 },
});

// { allowed: true, required_presence: 'VINAC_FM_LEVEL_3' }

Use Cases

AI Agent Pipelines

Replace natural language instruction passing between agents with typed semantic tokens. Every action is pre-agreed, bounded, and auditable, with no prompt injection path.

Cross-System Automation

Coordinate microservices, IoT devices, and ML models with a common semantic vocabulary. A token emitted by a sensor is resolved identically in cloud infrastructure.

Physical-Digital Bridging

Bind digital actions to physical presence using VINAC-FM as the transport. The Logos Lexicon token carries the semantic intent; the acoustic channel proves it is legitimate.

Zero-Trust M2M Auth

Remove implicit trust from machine-to-machine calls. Every action is an explicit token resolved against a pre-shared, locally-auditable logic library.

Traditional Messaging vs. Logos Lexicon

Traditional M2M

Text to packet to parse to interpret to execute
Instructions transmitted over the wire
Remote code execution surface
Variable interpretation per implementation
Presence proof bolted on later
Susceptible to prompt injection

Logos Lexicon

Semantic token -> resolve -> execute
No instructions transmitted, ever
Locally resolved logic library
Deterministic, bounded execution
Explicit presence and policy requirements
No prompt parsing path by design

Included In Open Preview

Logos Lexicon is part of the shared SilentAuth preview workspace. Pricing tiers are intentionally disabled while the vocabulary, VINAC-FM binding, and local gateway path are validated.

Logos 2.0 seed vocabulary
JavaScript and TypeScript SDK examples
Local policy evaluation model
VINAC-FM transport binding docs
Receipt and audit export model
Custom namespace design notes

The grammar of machine coordination

Stop passing instructions. Start exchanging pre-agreed intent references that resolve locally under versioned policy.