About Product OS

Product OS is a structured repository that holds everything a team needs to know about its product: vision, features, backlog, architecture decisions, research, and metrics. It is designed so that both humans and AI agents can read from it and write to it.

The pages in this section explain why Product OS exists, how different roles interact with it, and the mechanics of keeping it current.

Pages

Why Product OS — The problem of fragmented product knowledge, the design principles behind Product OS, and how it fits alongside tools like Jira and Linear.

Developer Workflow — Concepts (repository as contract, data/content separation, agent-agnostic design), setup, daily workflow for engineers, PR conventions, and team onboarding checklist.

Product Manager Workflow — How product managers use Product OS to define work, track progress, write specifications, and delegate trivial tasks to agents.

Agent Workflow — Reference for AI agents: the principles they follow, what to read before implementing, what to update after shipping, and the data files they operate on.

Agent Rules for Product Repos — The rule content to add to each code repository so agents treat Product OS as the source of truth. Includes placement instructions for Cursor, Claude Code, Windsurf, Copilot, and others.

Pet Store Demo — How the included pet store demo works: domain-based features, data flow, and fork/agent instructions in YAML.