Research should make the architecture easier to attack.
NULLWORKS publishes working research when a recurring operational problem needs a sharper model. These are not certifications or universal laws. They are versioned arguments, methods, and experiments intended to survive criticism, failure, revision, and succession.
Continuity Calculus
How a system changes without losing the intent, invariants, provenance, and reasoning required for valid succession.
Architecture Lineage
Trace how operating ideas, controls, systems, and implementations inherit from prior work instead of appearing without provenance.
Operational Relativity
Examine how the meaning of state, authority, evidence, and success depends on the operational frame in which a system is acting.
Outcome-First Operational Fuzzing
Challenge workflows from consequential outcomes and failure conditions rather than merely exercising software inputs.
Software Recency Bias
Investigate the assumption that newer software, interfaces, models, or practices are inherently more valid than older working mechanisms.
Model-Agnostic Transplant
Test whether useful operating architecture survives movement across models, providers, agents, and implementation environments.
Control Coverage
Map consequential claims and capabilities to the controls, evidence, tests, authority, and recovery paths that actually constrain them.
Assurance Index
Separate source status, commercial model, public metadata, evidence depth, and bounded NULLWORKS assurance conclusions.