PRD to Design

This skill is the second step in the build pipeline. It sits between the PRD (what and why) and the task list (build), and answers how:

idea-to-prd  →  prd-to-design  →  design-to-tasks  →  tasks-to-code
  (what/why)      (architecture)     (task list)         (build)

The PRD deliberately excludes schemas, API shapes, and component structure — those are design decisions. This skill makes them, records the rationale, and hands design-to-tasks a concrete architecture to decompose. The PRD stays authoritative for requirements; the design is authoritative for structure.

When to Use This — and When to Skip It

This step is optional. Use it when the architecture is not obvious and getting it wrong is expensive:

• More than two or three components, or a new boundary between existing ones
• A new or changed data model, public API, or integration contract
• A decision that is hard to reverse later (storage engine, sync model, auth model)
• Multiple reasonable approaches that need to be weighed before committing