What problem does it solve? Designing a well-structured AI agent skill requires balancing persona design, capability prompts, memory architecture, and quality standards, which is difficult to do consistently by hand. This Skill guides the creation, editing, and quality analysis of BMad agents through structured conversational discovery. ## Core Features & Use Cases - Three agent archetypes: Produces stateless agents (single SKILL.md), memory agents (lean bootloader plus a persistent sanctum), and autonomous agents (memory plus PULSE scheduled wake behavior), with the type emerging from discovery questions rather than upfront selection. - Create, Edit, and Analyze intents: Routes to a build loop for new agents, surgical edit guidance that preserves existing design, or a quality-analysis report based on the prompt-quality canon. - Deterministic scaffolding: Emits agents from templates (SKILL.md, customize.toml, sanctum files, wake.py, init-sanctum.py) via a template-processing script, with lint gates and optional eval runs before handoff. - Use Case: Ask to create a personal writing coach agent; the builder discovers it needs memory, drafts the persona and capabilities, scaffolds the sanctum templates, and hands off an agent that wakes with continuity across sessions. ## Quick Start Ask the assistant to create a new agent that acts as a personal coding coach with memory between sessions.
