write-spec

To translate user stories into a complete technical specification, provide your inputs in a stories file and use a coordinated spec writing process to generate architecture, data models, and validation checkpoints.

Coordinating cross-team architecture and data model design requires orchestrating staged checkpoints, role-specific outputs, and Skeptic reviews to ensure testable and auditable results across product and database concerns.

Generating an Architecture Decision Record for a feature migration involves evaluating user stories through staged validation, producing auditable role-specific outputs that document architecture and data model decisions.

Yes, this spec writing approach works for features requiring database migrations by spanning product, architecture, and database concerns, enforcing requirements through staged checkpoints and Skeptic reviews.

To start writing a feature specification, you need to provide user stories in a markdown file located at docs/specs/{feature}/stories.md to generate the draft specification output.

A Skeptic review should be used during the spec writing process as a staged checkpoint to enforce requirements, validate architecture and data model designs, and ensure the final specification is testable and auditable.