Ideas & Frameworks

This section contains public versions of reusable analytical structures. A framework page explains the claim, use, boundary, and public applications without publishing every private scoring rule, threshold, or working component.

CIEA — Core-Inference Evaluation Architecture

Tests whether an apparent improvement in a metric reflects progress toward the underlying core or moves cost, verification, rollback, and responsibility outside the decision table.

Open the architecture →

Feasible-Set Governance

Examines how candidate sets and decision environments are structured before choice begins.

Open the framework →

O3M

A human-centered framework for understanding how repeated interaction makes older categories for AI unstable.

Open the framework →
Public-note principle: framework pages expose a usable analytical surface and clear boundaries. They do not imply that every internal mechanism, draft lineage, or scoring rule is public.