bmad-create-architecture

Create software architecture decision documents through collaborative step-by-step workflow.

Updated Aug 25, 2024
One-click install
npx skills add https://github.com/LeGeRyChEeSe/LanPlay-DiscordBot --skill bmad-create-architecture-legerycheese
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: bmad-create-architecture
Source: https://github.com/LeGeRyChEeSe/LanPlay-DiscordBot/tree/main/bmad/src/bmm-skills/3-solutioning/bmad-create-architecture
Command: npx skills add https://github.com/LeGeRyChEeSe/LanPlay-DiscordBot --skill bmad-create-architecture-legerycheese

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? It prevents inconsistent AI agent implementations by producing a complete, validated architecture decision document that records technology choices, patterns, and project structure before coding begins. ## Core Features & Use Cases - Guided Architecture Workflow: Walks through eight sequential steps covering context analysis, starter template evaluation, core decisions, implementation patterns, project structure, and validation. - Consistency Rules for AI Agents: Defines naming, structure, format, and communication patterns so multiple AI agents write compatible code. - Resumable State Tracking: Tracks progress in document frontmatter so interrupted architecture sessions can resume at the correct step. - Use Case: After finishing a PRD, a team runs this workflow to decide their stack (e.g., Next.js with PostgreSQL), verify current versions via web search, and produce an architecture.md that guides all subsequent implementation. ## Quick Start Ask the AI to create a technical architecture for your project based on your PRD and UX design documents.

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 software architecture document with an AI agent?▼

Run this workflow after completing a PRD. It guides you through eight steps: initialization, context analysis, starter template selection, core decisions, implementation patterns, project structure, validation, and completion, producing an architecture.md file.

What inputs does the architecture workflow require?▼

A PRD is mandatory; the workflow will not proceed without one. It also optionally loads UX designs, research documents, product briefs, and project context files discovered in the planning artifacts or docs folders.

Can I resume an interrupted architecture workflow session?▼

Yes. Progress is tracked in the document frontmatter via stepsCompleted and inputDocuments fields. On reactivation, a continuation handler analyzes existing work and offers options to resume, continue, review remaining steps, or start over.

How does the workflow ensure AI agents write consistent code?▼

It defines implementation patterns covering naming conventions, file structure, API response formats, event naming, and error handling. These consistency rules prevent different agents from making conflicting choices during implementation.

Does the architecture workflow verify technology versions?▼

Yes. The starter template and decision steps mandate web searches to verify current versions and maintenance status of frameworks and libraries, explicitly forbidding reliance on hardcoded version numbers.