NIDUS — Neural Intelligent Distributed Unified System: Architecture Specification, Version 1.0

NIDUS (Neural Intelligent Distributed Unified System) is a self-learning, cell-based architecture for distributed quality control of code and knowledge. A single cell template is instantiated as many independent Spores across separate projects; the Spores share a common Brain (a cumulative learning dataset) and may communicate as a Mesh. Each Spore performs read-only analysis and proposes — it never mutates the host project autonomously. This document specifies the architectural model, terminology, functional taxonomy, seven invariant design canons, and the feedback-primary learning loop. It is the public reference for the NIDUS name and architecture. The reference implementation, models, training data, and operational topology are NOT included and remain the exclusive work of the author. Licensed CC BY-NC-ND 4.0: may be shared with attribution, but not used commercially and not modified. All other rights reserved by the author.