An interactive 3D map of 907 agent skills curated in
VoltAgent/awesome-agent-skills.
Skills that describe similar capabilities end up near each other in space.
Pipeline
Parse ~900 one-liners from the source README into (name, team, description).
Embed each name: description with all-MiniLM-L6-v2 (384-d).
Project to 3D with UMAP (cosine metric).
Cluster in the 384-d space with KMeans (k=10) — projection is lossy, so we cluster on the embedding, not on the 3D coords.
Label each cluster by asking gemma4:e2b (via Ollama) for a 3–6 word human title over its 30 centroid-nearest members.
Render with Three.js — glow sprites, neighbor-web, fog, orbit controls.
Colors
Color by Topic — each point is tinted by its KMeans cluster, labeled via Gemma (e.g. "Authentication and Identity Management").
Color by Team — each point is tinted by the GitHub-org prefix of its skill (anthropics/, microsoft/, getsentry/, …). 161 distinct teams.
Controls
Drag to rotate · scroll to zoom · right-drag to pan.
Hover a point → tooltip · click a point → detail panel (with link to the skill).
Sidebar row: click to isolate, shift-click to toggle.
Search: full-text over name, description, and team.