Proposal and permission are separate
A model, rule or researcher supplies a candidate. The executor evaluates permission independently; a stronger prediction does not enlarge the authority envelope.
Architectural principles
Prediction is not the hard problem.
Authority is.
Execution authority is an architectural boundary. Neurobrick separates the candidate a research process proposes, the permission an owner grants and the action an engine can submit. The selected KIS path makes that separation concrete: KisOrderExecutor consumes envelope precheck and authorization, while portfolio controls and Control Plane observation retain separate responsibilities.
The design problem is what remains bounded when a prediction is wrong, a process restarts or a broker response is ambiguous. Permission, attempt identity and observation scope therefore have explicit representations and separate consumers.
Authority is consumed at a boundary
The selected KIS executor reads envelope precheck and final authorization. Owner, scope and usage bounds belong to the permission that the consuming path evaluates.
Stopping and opening are asymmetric
The design permits bounded automation to tighten controls or reduce risk. Opening or expanding live-capital access remains a human authority decision; a stop does not grant permission to resume.
Evidence follows the consuming path
Source integration, an isolated test and a natural operating event establish different facts. Each result retains its engine, surface, instrument, source revision and observation window.
Asymmetric authority
Human authority defines the execution envelope.
The human operator defines the live-capital envelope. Automation can act in the safe direction only inside its bounded authority; it cannot turn a proposal, approval record or healthy endpoint into wider permission. The dated system review identifies the observed permissions and preserves the open questions about actor attribution, surface coverage and effective state.
Scope
Governance is an engineering discipline within this independent financial infrastructure. Returns, balances, strategy parameters and investment performance are outside the public surface. The question here is whether authority is bounded and whether the evidence supports that claim.