A new LLM-based architecture detects and quantifies the intensity of human values within text. It bypasses the need for complex prompt engineering or rigid value theories. This framework allows developers to assess how autonomous decisions align with moral considerations. Practitioners can now identify implicit ethical signals without relying on static, theory-bound datasets.