The Polycentric Autonomy-Audit System has two production implementations here. The framework, theory, and ongoing research live on oumo.space — this page covers what's actually running.
Full-stack production implementation. Cryptographic hash-chain audit trail, blind review isolation, built-in agent simulation.
Simplified two-container variant — the same governance pattern without the infrastructure.
Orbsys's agent simulation tooling exists and runs, but hasn't yet been through rigorous, published live-simulation trials under real adversarial conditions. Efficiency multipliers, crisis-survival rates, and similar figures in the framework documentation are design projections, not measured outcomes from a live deployment. Treat them as the hypothesis the architecture was built to test, not yet a verified result.