A smarter, self-hosted AI assistant — multi-user, multi-agent.
Terminal score: 0–1000 raw, weighted across 4 dimensions. Public score: 0–10 normalized (shown in the 30-day stars chart above).