bmad-create-architecture

Create AI agent architecture designs through collaborative decision-making workflows.

Updated Apr 25, 2026
One-click install
npx skills add https://github.com/petrkohut/bmad-todo-app --skill bmad-create-architecture-petrkohut
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: bmad-create-architecture
Source: https://github.com/petrkohut/bmad-todo-app/tree/main/.github/skills/bmad-create-architecture
Command: npx skills add https://github.com/petrkohut/bmad-todo-app --skill bmad-create-architecture-petrkohut

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires bmad-advanced-elicitation, bmad-party-mode, and includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve?

This Skill facilitates the creation of a comprehensive architecture solution design for AI agents, ensuring consistency in implementation.

Core Features & Use Cases

  • Architecture Workflow: Guides through a step-by-step process to collaboratively define architecture decisions.
  • Technical Preference Discovery: Identifies user technical preferences for language, frameworks, databases, etc.
  • Starter Template Evaluation: Analyzes and selects the most suitable starter template based on project requirements.
  • Core Architectural Decisions: Facilitates decision-making on data architecture, security, communication, frontend, and infrastructure.
  • Implementation Patterns: Defines naming, structure, format, communication, and process patterns to prevent conflicts.
  • Project Structure: Defines the complete project directory structure and architectural boundaries.
  • Validation: Validates architectural coherence, completeness, and readiness for implementation.
  • Handoff: Provides clear implementation guidance and next steps.

Quick Start

Initialize the architecture workflow by running the 'step-01-init' script.

Frequently Asked Questions about bmad-create-architecture

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I create a consistent architecture solution design for AI agents?

To create an AI architecture solution design, use a guided workflow for technical preference discovery, starter template evaluation, and core architectural decisions. This ensures implementation consistency by defining naming, structure, and communication patterns.

What is the process for defining implementation patterns in AI solution design?

Defining implementation patterns in AI solution design involves establishing naming, structure, format, communication, and process rules. These patterns prevent conflicts and ensure architectural coherence before validating the project structure and handing off for implementation.

How do I select a starter template for my AI project architecture?

Selecting a starter template for AI project architecture requires evaluating project requirements against available templates. The architecture workflow analyzes your technical preferences for languages and frameworks to recommend the most suitable foundation for your solution.

Can I use this architecture workflow without predefined technical preferences?

You can use the architecture workflow without predefined technical preferences because it includes a dedicated discovery phase. This step collaboratively identifies your preferred languages, frameworks, and databases during the initial architecture process.

What architectural decisions are covered when designing an AI solution?

Designing an AI solution covers core architectural decisions regarding data architecture, security, communication, frontend, and infrastructure. The workflow also defines the complete project directory structure and establishes clear architectural boundaries.

Do I need stakeholder input to define project structure and architectural boundaries?

You need stakeholder input to define the project structure and architectural boundaries because the workflow requires project context documentation and technical preferences. This collaborative input ensures the final architecture is validated for coherence and implementation readiness.