I ran my Claude Code session history through Anthropic's AI Fluency Index taxonomy — 11 observable behaviors from their February 2026 study of 9,830 conversations — and found I'd never once touched three of the behaviors in the bottom tier of the skill radar. I knew I was fast. I didn't know I was narrow.
skill-tree classifies those 11 behaviors across your own sessions, then assigns one of seven archetype cards rendered as tarot cards with curated museum art. The archetypes (Illuminator, Navigator, Alchemist, and four others) aren't cosmetic — each maps to a specific cluster of behaviors across Dakan & Feller's three visible fluency axes: Description, Discernment, Delegation.
The live fixture at skill-tree-ai.fly.dev/fixture/illuminator shows what the Illuminator card looks like: the skill radar, the archetype narrative, the growth quest. Your own card comes back as a stable URL after a 7-step orchestration that runs in roughly 30–60 seconds.
The growth quest is the part I actually use. After each analysis, skill-tree picks one behavior you haven't triggered and pins it as a quest via the SessionStart hook. It persists in ~/.skill-tree/ between sessions so Claude Code surfaces it at the top of your next conversation.
Install in Claude Code:
claude plugin marketplace add robertnowell/ai-fluency-skill-cards
claude plugin install skill-tree-ai@ai-fluency-skill-cards
Also available as an MCP server (npm install skill-tree-ai) for Cursor, VS Code, and Windsurf, and as a zip for Claude Cowork.
github.com/robertnowell/skill-tree
Top comments (0)