deepseekoracle/lygo-protocol-stack

This repository calls itself a Sovereign Intelligence Framework, made up of six numbered protocols, labeled P0 through P5, that the author presents as building blocks toward a self-governing AI system. Each protocol has its own folder and its own name: a byte entropy filter that flags input as amplify, soften, or quarantine, a memory system called Memory Mycelium that splits data into fragments for storage, a cognitive bridge described as turning subjective experience into ethical vectors, a consensus mechanism referencing Tesla's 3-6-9 numbering, an ascension engine with nine evolution levels, and a harmony node meant to combine human and AI decision making. The language throughout mixes standard software engineering terms with spiritual and numerological concepts such as frequency tuning, light codes, and lattice alignment, so a reader should treat the framing as the author's own terminology rather than an established computer science standard. Technically, the project is written mostly in Python, with some components also ported to C, Rust, and Verilog for what the README calls determinism testing, meaning it checks that different language versions of the same filter produce identical results. It includes a setup script, a Docker Compose file for running a community node, and a large collection of command line tools and demo scripts that exercise each protocol individually or as a combined stack. There is also a separate published library of skills tied to a platform called ClawHub that can be installed with a command line tool. Setting it up involves cloning the repository and running either a shell script or Docker Compose, after which various Python scripts under the tools folder can be run to see each protocol's demo. The project uses its own custom license rather than a widely recognized open source license. This is best suited for someone curious about an experimental, personally authored take on AI architecture, not a production ready toolkit for typical software development.