Construction is not evidence.
Building a substrate, producing an effect or matching a pattern does not automatically establish the interpretation one might prefer. The construction and the claim are evaluated separately.
ProtocolLogic · public record
Principles for public work.
These principles govern what this site can say. They are a standard for publication, not a claim that active research is complete.
Claims have a boundary.
Every public claim should identify its object, the support it has, and the point at which inference would become speculation. A limitation is part of the result, not an apology appended to it.
Methods are not narratives of success.
A method earns confidence through its constraints, tests and stated failures. A coherent story about why a model should work is not a substitute for those things.
Publication is selective.
Public material is a record for others, not a mirror of every active thought. Work in development remains internal until it can be released without exposing unfinished architecture or turning provisional ideas into public commitments.
Release test
A document is public only when its claims, scope and disclosure consequences have been reviewed separately from the work itself.