Anthropology's February 2026 AI Fluency Index studied 9,830 Claude conversations and classified 11 observable collaboration behaviors across all of them. I read it and immediately wanted to know where my sessions landed on that same taxonomy — not as a score, but as a distribution. Which behaviors do I reach for constantly? Which ones haven't shown up in my logs once?
That question is what skill-tree answers.
The tool installs into Claude Code via the plugin marketplace:
claude plugin marketplace add robertnowell/ai-fluency-skill-cards && claude plugin install skill-tree-ai@ai-fluency-skill-cards
On run, it walks a 7-step pipeline: find your session files, extract your messages, send them to a remote classifier (Claude Haiku on Fly.io), assign an archetype, synthesize a narrative, render, and return a stable URL. End-to-end it takes 30–60 seconds.
The 11 behaviors and the three visible axes — Description, Discernment, Delegation — come from Dakan & Feller's 4D AI Fluency Framework. The fourth axis, Diligence, doesn't surface in chat logs so it's excluded. The archetype assignment maps your behavior distribution to one of seven cards, rendered as tarot cards with curated museum art. A live fixture is at skill-tree-ai.fly.dev/fixture/illuminator if you want to see the output format before running it on your own history.
The part I found most useful wasn't the archetype — it was the growth quest. skill-tree picks a behavior from the 11 that hasn't appeared in your sessions and pins it as a prompt for your next session via a SessionStart hook. That hook persists across sessions in ~/.skill-tree/, so the suggestion carries forward rather than disappearing the moment you close the window.
Also works in Claude Cowork via skill-tree-ai.zip, and there's an MCP server (npm install skill-tree-ai) for Cursor, VS Code, and Windsurf.
Top comments (0)