ARE

Tenets

The foundational elements of Agent Responsibility Engineering.

The tenets define the posture of ARE: build governance into the architecture, run agents through scoped authority, demand legible proof, and refuse unknown action paths.

Foundational

How you build it

I

Governance is architectural, not operational.

II

The spawn chain is the authority chain.

III

The ledger is ground truth.

Operational

How you run it

IV

Intelligence never grants authority.

V

Scope is a contract, not a suggestion.

VI

Trust has a half-life.

Epistemological

How you know it

VII

Every agent must be provably legible.

VIII

Proof requires falsification.

Posture

How you hold the line

IX

An unknown agent is an ungovernable agent.

X

The system must explain itself.

Reading The Tenets

They are not slogans. They are constraints.

Each tenet is meant to narrow the design space. If a proposed agent architecture cannot preserve scope, legibility, denial, and evidence, it is not yet governed in the ARE sense.

Architecture before operations

ARE starts in the system design. A policy that only appears in a runbook is too late to govern autonomous action.

Authority before intelligence

The model's capability can inform a decision, but it cannot grant the right to execute that decision.

Proof before trust

Legibility, falsification, and ledgered evidence make governance inspectable rather than ceremonial.