Playbooks
How I work, written down. I use each playbook the same way on every project, and each one can be handed straight to an AI agent.
AI Project Setup
The six things I set up before an AI works on a project, in order: rules for every project, a written brief, rules for this project, memory for corrections, checks that stop a bad deploy, and commands for long sessions. Each one is there because something went wrong without it.
Read playbookAI Operating System
How I work with AI once a project is running: the seven ways AI work drifts, the rules that keep it on track, the phases I work through, and the checks every output has to pass.
Read playbookPrompt Library
The prompts I keep for design work, in the order the work happens: a project brief the model reads first, then prompts for discovery, defining the problem, designing, and checking the result. Each one says when to use it and what to check before you trust the answer.
Read playbookDiscovery to Scope
How I turn a vague request into a plan a team can build. I use Jobs to be Done to find what the customer is trying to get done, the Double Diamond to explore and then decide, and a scope document that names the problem, the bet, how success is measured, and what is left out.
Read playbookFigma Operating System
How I run Figma with an AI agent working in the file: naming, status glyphs, tokens, versions, handoff, and the context the agent needs so its work matches the design system.
Read playbookInterface Content System
How I write the words on every screen. Write from the person's side, name buttons for what they do and keep that name, and make errors say what to do next. Every rule is shown as a before and an after.
Read playbookDesign System Migration
How I move a product to a new design system while feature work keeps shipping. Take stock of what is in use, keep old names working through the token library, let Storybook say when each component is done, track it every sprint, and retire the old system on a published date.
Read playbook