Curated skills, sub-agents, and config templates that supercharge Claude Code — research, image gen, GitHub automation & more.
Terminal score: 0–1000 raw, weighted across 4 dimensions. Public score: 0–10 normalized (shown in the 30-day stars chart above).