skill-tree: finding the Claude behaviors you systematically avoid

After months of daily Claude Code use, I ran my session history through Anthropic's AI Fluency Index taxonomy and found I'd triggered 8 of the 11 observable behaviors — but three had zero occurrences. Not low. Zero. That's the problem skill-tree is built around. It's not hard to get faster at the th