An invariant, auditable decision framework (formal methods, non-narrative)

I’m sharing a public OSF corpus describing a closed, invariant, non-narrative decision framework.

The system is based on:

• fixed sequential states,
• invariant vector axes,
• deterministic scalar evaluation,
• a sealed priority-stop mechanism.

It does not generate actions, prescribe targets, or automate decisions.
It strictly constrains reasoning for auditability, stability, and error prevention.

The repository includes an A3 comparative table positioning the framework against
ISO/IEC, NIST AI RMF, EU AI Act, IEEE EAD, and OECD AI — limited to structural properties
(auditability, invariance, decision structure, reproducibility).

Non-normative, non-prescriptive.

OSF repository:
https://osf.io/ub5f4/