Becoming an AI Power User

A principle-first guide to building an AI practice that compounds. Power users build compounding loops: they experiment at the frontier, reject weak output, diagnose failures, meta-prompt improvements, save reusable assets, evaluate with clear standards, and retest as models change. This book teaches the durable principles through practical projects: quality ratcheting, meta-prompting, error-to-spec transformation, context engineering, experimentation as R&D, evaluation loops, separation of concerns, source verification, portable infrastructure, creative building, judgment preservation, and workflow redesign.

Skill Guide: Turn Repeated Work Into Reusable Skills

AI memory and personalization keep getting better. This guide shows when that is enough, when important context still needs explicit encoding, and how reusable Skills preserve the procedures modern AI should follow every time.

For anyone turning repeated Claude and Codex workflows into reusable skills