Invariant Filtering and Δ>0 Logic: A Minimal Decision-Making Protocol

This post introduces a formal, non-normative decision-making protocol designed to operate independently of authority, ethics layers, or narrative framing.

The protocol is based on three core principles:

1. Invariant filtering
   
   Only what survives context changes, projection, and reinterpretation is retained.

2. A calculable Δ (delta)
   
   Every candidate action must yield a measurable delta.

3. A minimal propagation rule
   
   Only actions where Δ > 0 are allowed to propagate.
   Δ = 0 results in silence.
   Δ < 0 is discarded.

The system does not attempt to replace existing standards or frameworks (ISO, NIST, EU AI Act, etc.).
It does not claim alignment, morality, or intent.

It either functions or it does not.

The full corpus is:

• public,
• auditable,
• reproducible,
• and open to criticism and counterexamples.

Repository:
https://osf.io/ub5f4

Feedback is welcome, especially from readers familiar with formal methods, systems engineering, or decision theory.