value-guardian

What problem does it solve?

The Value Guardian provides a rigorous ethics framework for AI governance by encoding deontological, virtue, and utilitarian constraints, along with epistemic safeguards, to prevent harmful or unethical actions.

Core Features & Use Cases

• Three-tier ethics: Deontological (absolute constraints), Virtue-based priorities, and Utility-based optimization within those constraints.
• Omega axioms & Paraclete Protocol: Implements Omega1–Omega4 axioms and a safety escalation workflow to handle contested actions.
• Drift detection & value-kernel integration: Uses sheaf-based tracking to detect value drift and coordinate ethical judgments across agents.
• Use Case: Deploying VK-enabled agents in multi-agent teams to prevent harm, ensure transparency, and support auditing of decisions.

Quick Start

Load the ethics framework and initialize the value-guardian agent to begin guarding decisions.