LimFlex
Evidence

Three reasons to take LimFlex seriously.

The public site keeps this simple. Full methods, metrics, scope, and limitations are in the current Zenodo manuscript.

1. Judgment can stay stable even when AI process paths differ.

The architecture is designed around the qualified judgment boundary, not identical reasoning traces.

2. AI output does not need to inherit authority.

LimFlex keeps judgment, authority, and execution as separate operational concerns.

3. External records can become reusable judgment structures.

Public research includes fresh external records and selective reuse / requalification behavior.

Technical evidence

Want the numbers, methods, and claim boundaries?

Read the current public manuscript on Zenodo.

Engineering the Path from Question to Judgment — v1.5
The site sells the value.
The paper carries the proof.

That separation keeps the public page readable while preserving full technical traceability.