DCL Evaluator

About

DCL Evaluator provides cryptographic verification of all AI agent decisions. You receive irrefutable proof of each LLM's choices, ensuring deterministic, tamperproof, and exactly reproducible results. Every output undergoes policy evaluation, resulting in a COMMIT or NO_COMMIT verdict. Each decision is secured with a SHA-256 hash, linked sequentially. Compatible with Ollama, Claude, GPT-4, Grok, and Gemini. Operates entirely offline and is designed primarily for desktop use.