bmad-create-architecture

To create an architecture document for AI-assisted development, provide your project requirements and supporting context like PRDs and UX specs to generate a single source of truth for technical decisions.

The best way to structure a project for multiple AI agents is to define implementation patterns, naming rules, and directory boundaries that ensure consistent collaboration and reliable execution across the codebase.

Make technical decisions for a new software project by collaboratively evaluating starter templates and selecting current technology versions for data, security, APIs, frontend, and infrastructure based on analyzed requirements.

Yes, you can use existing PRDs and epics to generate your project architecture by analyzing the scope, constraints, and cross-cutting concerns defined in those project documents.

You need to validate architecture readiness for implementation when starting a new BMAD project to ensure boundary definitions and end-to-end handoff preparation are complete before coding begins.

Inconsistent implementation happens with AI agents due to ambiguous requirements; prevent it by building a complete architecture document that establishes consistency patterns, naming rules, and validation checks.