arifOS

Constitutional AI governance kernel that enforces 13 safety rules between language models and tools.

arifOS is an AI governance framework that acts as a safety firewall between language models and tools, enforcing 13 constitutional rules (Floors) covering truthfulness, reversibility, epistemic humility, and human sovereignty. It provides tools for session anchoring, reasoning analysis, ethical simulation, and verdict synthesis through a hosted MCP endpoint with streamable HTTP transport.