In February 2026, Anthropic published an AI Fluency Index built from 9,830 Claude conversations. They measured 11 observable collaboration behaviors — things like whether users verify outputs, delegate judgment, or iterate on prompts — and classified each session across a population. I read it and immediately wanted to know: which of those 11 behaviors show up in my sessions, and which ones I've never touched once.
So I built skill-tree: it reads your Claude Code or Cowork session history, runs the same 11-behavior classification against Anthropic's taxonomy, and maps your distribution against that population baseline.
The classifier runs remotely on Claude Haiku (hosted on Fly.io, ~30–60 seconds end-to-end) and returns a stable URL with a rendered visualization. The taxonomy comes from Dakan & Feller's 4D AI Fluency Framework — three axes are visible in chat logs (Description, Discernment, Delegation), and the tool uses those to assign one of seven archetype cards, rendered as tarot cards with curated museum art. Live example: skill-tree-ai.fly.dev/fixture/illuminator.
The part I use most: it picks one behavior from the 11 that you haven't tried and sets it as a growth quest for your next session. That quest persists via a SessionStart hook, so Claude Code surfaces it at the start of the next conversation without you having to remember to check.
Install in Claude Code:
claude plugin marketplace add robertnowell/ai-fluency-skill-cards
claude plugin install skill-tree-ai@ai-fluency-skill-cards
Also works in Cowork via skill-tree-ai.zip, and available as an MCP server (npm: skill-tree-ai) for Cursor, VS Code, and Windsurf.
Top comments (0)