bmad-create-architecture

To create architecture decisions for AI agents, use a collaborative, stepwise workflow that guides elicitation and pattern establishment. This approach tracks progress using YAML frontmatter to ensure uniform implementation across multiple agents.

The best way to align multiple AI agents is through workflow-driven architecture leadership that uses collaborative elicitation and decision making. This establishes uniform patterns and tracks step-by-step progress to maintain consistency across implementations.

Frontmatter-driven state tracking uses a unique identifier in YAML frontmatter to monitor progress across pre-defined steps. This mechanism ensures safe step execution and maintains state throughout the collaborative architecture decision process.

Yes, you can include optional resources under scripts, references, and assets directories within your architecture decision workflow. These components support the stepwise process while maintaining safe step execution and frontmatter-tracked progress.

Yes, YAML frontmatter with a unique identifier is required to track progress accurately in the architecture workflow. This ensures safe step execution and maintains state across collaborative elicitation and decision-making steps.

Use a stepwise workflow when a team needs to align multiple AI agents on complex architectural decisions. The structured process with collaborative prompts and frontmatter-tracked state ensures consistent implementation across all agents.