VALLIGNUS

Documentation

Documentation

Architecture, policy model, enforcement semantics, and operational guarantees for execution-layer governance.

Request AccessRead the Whitepaper

What This Covers

Runtime governance concepts

Enforcement boundary and decision flow

Policy model and evaluation semantics

Audit trail and chain of custody

Operational guarantees and failure handling

Core Concepts

Runtime Supervisor

The enforcement layer that intercepts and evaluates every agent action before execution.

Enforcement Boundary

The point at which policy decisions are made and authority is granted or denied.

Deterministic Policy Evaluation

Policies are evaluated synchronously with consistent, reproducible outcomes.

Audit Chain

Cryptographic record of all decisions, actions, and state changes for review and attestation.

Deny-by-Default

No implicit permissions. Every action requires explicit policy authorization.

Policy Model

Illustrative policy example. This is not installation or setup guidance.

policy "agent_network_access" {
  default = "deny"

  identity {
    require_attestation = true
    allowed_issuers     = ["internal-ca.vallignus.io"]
  }

  network {
    allow = [
      "api.internal.corp:443",
      "telemetry.vallignus.io:443"
    ]
  }

  execution {
    max_runtime_seconds          = 300
    terminate_on_policy_violation = true
    audit_all_actions            = true
  }
}

Guarantees

Deny-by-default posture

Synchronous enforcement at execution time

Forced termination on violation

Cryptographic audit trail suitable for review and attestation

No implicit trust inheritance across tasks

Evaluating Vallignus for enterprise or government use?

Request AccessWhitepaper