nmCRYPT and software-defined neuromorphic entropy
nmCRYPT is one of the stranger branches of the cording.ai archive: a software-defined neuromorphic entropy experiment.
The reported result was credible randomness across 186 of 188 NIST STS subtests. That does not make it a production cryptographic primitive. It does make it a useful research artifact about chaotic, stateful, software-generated entropy.
Its relevance to TightLoop is indirect. TightLoop systems rely on rich internal state, nontrivial recurrence, and residual patterns. nmCRYPT explored whether neuromorphic state dynamics could produce statistically credible entropy-like behavior.
The commercial path for nmCRYPT is much less immediate than Sentinel or adaptive assist. Security products need a much higher bar: public review, cryptanalysis, threat models, implementation hardening, and conservative claims.
For now, it belongs in the archive as evidence of breadth: the same stateful neuromorphic design instincts were tested beyond control and prediction.