Enterprise audit agent workspace with Agentic RAG, governed tool use, evaluation harness, memory, and human-review delivery workflows.
Terminal score: 0–1000 raw, weighted across 4 dimensions. Public score: 0–10 normalized (shown in the 30-day stars chart above).