Protective computing, minimization-first systems, and degraded-first design. Notes written for low-trust reality: where incentives are misaligned and “normal conditions” are optional.
A reproducible trust case for an offline-first encrypted evidence app, with a deliberately bounded guarantee surface tied to an exact hosted-green release.
Why degraded first architecture is the only viable path for high threat environments
Building a chronic pain journal that reduces collection by default without breaking the product
A method for modeling legal, operational, and digital threat surfaces
Protective Computing, now pinned to a stable canonical reference.
Why reducing collection by default is a product boundary, not a preference.
A layered discipline for systems built under instability: theory → operations → measurement.