bmad-create-architecture

To create an architecture decision document, analyze your project repository and product docs to capture requirements, evaluate starters, and define project structure. This process yields a structured plan ensuring AI agents follow consistent technical choices across the full-stack implementation.

The best way to ensure architecture consistency is to generate a collaborative document defining naming, format, communication, and process rules. This enforcement prevents conflicting technical choices and incomplete implementation guidance when multiple agents build compatible code simultaneously.

Analyze PRDs and UX notes by reviewing them to summarize scope, constraints, and complexity. This context analysis facilitates collaborative choices for data, security, API, frontend, and infrastructure, resulting in a validated project structure definition.

Yes, you can use this architecture planning workflow for multi-step software design. It supports requirements analysis, starter evaluation, decision capture, project structure definition, and validation before implementation, applying to complex workflows needing consistent AI agent guidance.

Yes, the architecture workflow supports YAML frontmatter state tracking and document appending. It also features menu-driven user confirmations, ensuring consistent guidance and state management for AI agents across the full multi-step software design process.

You need a structured architecture plan when building an end-to-end full-stack SaaS project with AI agents. It is required to prevent conflicting technical choices and incomplete implementation guidance, ensuring all agents build compatible code from a validated decision document.