Spec Kitty: Structured AI Coding for Real Teams

​Workshop with Robert Douglas, founder and creator of Spec Kitty.

​AI coding agents can write real code today. That's not in question anymore. What breaks is everything around the code: work that drifts from what was actually needed, no record of why the agent made the choices it made, and results that are hard to hand off or review.

​Spec Kitty is a structure for working with AI coding agents — Claude Code, Codex, Cursor, Gemini CLI, and others — that replaces "vague prompt, hope for the best" with a clear spec, a plan, and a checklist your agent actually works from.

​In this hands-on session, you'll see:

• ​Why agentic coding breaks down without structure — and why that's a workflow problem, not a model-quality problem

• ​How a Spec Kitty mission actually works, live — from spec, to plan, to reviewed and merged code

• ​How it fits into a real, messy, existing codebase — this isn't a greenfield-only tool

• ​A live testimony from a real customer on what adopting it actually looked like — the good, the friction, and what they'd tell you before you start

​Whether you're evaluating agentic coding tools for the first time or already using one without a system behind it, you'll leave with a concrete picture of what "structured" actually looks like day to day.