This post introduces an open corpus describing a formal invariant decision framework.
The framework is intentionally:
- non-normative
- non-prescriptive
- non-authoritative
It does not claim governance power, compliance status, or alignment authority.
It is designed to either work or fail — nothing else.
What it focuses on
- invariant filtering of inputs
- explicit decision traceability
- Δ-based outcome evaluation
- reproducibility over persuasion
No rhetoric, no behavioral nudging, no policy recommendations.
What it is not
- Not an AI alignment manifesto
- Not a governance proposal
- Not a standard, norm, or guideline
This corpus does not ask to be adopted.
It asks to be audited, reused, forked, or rejected on technical grounds only.
Open corpus
Full materials are publicly available here:
https://osf.io/ub5f4/
Adversarial review and technical critique are welcome.
Top comments (0)