A Claude Code plugin that shows what's happening - context usage, active tools, running agents, and todo progress
Terminal score: 0–1000 raw, weighted across 4 dimensions. Public score: 0–10 normalized (shown in the 30-day stars chart above).