bmad-workflow-builder

You can build an AI workflow through conversational discovery by following guided phases for intent discovery, classification, requirements gathering, drafting, building, and testing. This process iteratively refines the workflow structure using LLM guidance and Python scripts.

The best way to optimize existing AI skills is by applying comprehensive validation using lint scripts and LLM scanner subagents. This quality optimization process identifies structural inefficiencies and applies targeted improvements or fixes to enhance performance.

Yes, you can convert non-compliant AI skill formats into standardized workflows. The skill builder supports format conversion alongside editing existing workflows, ensuring your AI skills meet structural requirements and operate efficiently.

Yes, you need Python installed to run the workflow builder scripts. While an LLM provides the conversational guidance for discovery and drafting phases, Python executes the underlying scripts required for building and validating the workflows.

Conversational discovery guides multi-stage workflow creation by leading users through six distinct phases: intent discovery, skill type classification, requirements gathering, drafting, building, and testing. This ensures complex workflows with human-in-the-loop interactions are well-structured and efficient.