The AI Developer's Field Guide: Classes, Monsters, and Anti-Patterns in Coding

Previously published as The AI Developer's Field Guide: Classes, Monsters, and Anti-Patterns in AI Coding by Timothy O'Brien. VOLUME I - SECOND EDITION Generative AI changed how software gets built. It also changed how teams fail. Code arrives faster. Pull requests get bigger. Dashboards stay green. Somewhere between the confident answer and the production incident, a team stops asking the questions that used to keep it out of trouble. The AI Developer's Field Guide gives developers, engineering managers, and technical leaders a practical vocabulary for recognizing those moments. Timothy O'Brien uses the classes, artifacts, spells, and monsters of role-playing games to name the habits of real people building software with AI. Meet the Fighter, Wizard, Rogue, and Cleric: four recurring developer archetypes, each with familiar strengths and tempting shortcuts. Then enter the bestiary: the Scope Creep Kraken, Temporal Tangle Worm, Congealing Slop, Stagnant Miasma, Phantom Intern, and Probability Pixie. These are the failures that appear when plausible output outruns understanding, coordination, and judgment. Inside you'll find: - Recognizable patterns of overconfident merging, ornamental architecture, fragile fixes, and misplaced trust in automation. - A bestiary of failures that spread through projects and teams, including how those failures reinforce one another. - Review questions and practical interventions: who should act, what to ask, and which habit needs to change. >This second edition updates the discussion for agentic coding workflows, connected tools, persistent context, and AI code review. It adds practical attention to agent boundaries and version-aware context while keeping the focus on human decisions and engineering responsibility. Written for developers and the people who lead them, this illustrated field guide is sharp, funny, and grounded in the everyday work of building software. No tabletop gaming experience required. If you've ever looked at an AI-generated change and thought, "This seems right, but something is wrong," this book gives you names for what you're seeing-and a place to start.