capability
Representation and evidence gap analysis
CAP-2status · offered with declared uncertainty2026-07-13
Systematic comparison of what an organization actually does against what its representations, human and machine-legible, allow an evaluator to verify: which requirements go unsubstantiated, which claims carry no checkable evidence, and which capabilities are invisible to a machine reader.
Why this is marked experimental: the analysis framework derives from the research program's untested hypotheses (H-2 most directly) and no published method yet backs it. It is diagnostic work performed carefully and labeled honestly.
What this cannot tell you: that closing the gaps will change any particular evaluator's recommendation. We measure and diagnose; we do not promise evaluator behavior.
Authors
Upstream Zero
Machine rendering