Problem Addressed

BMAD-METHOD

Overview

BMAD-METHOD (Breakthrough Method of Agile AI-Driven Development) is an AI-driven agile development framework providing 21 specialized agents, 50+ guided workflows, and scale-adaptive intelligence that adjusts methodology depth based on project complexity. Unlike traditional AI tools that generate solutions directly, BMAD agents act as expert collaborators who guide developers through structured processes grounded in agile best practices across analysis, planning, architecture, and implementation phases.

Problem Addressed

Key Statistics

Key Features

Core Framework

• 21 Specialized Agents: Domain experts including PM, Analyst, Architect, Developer, UX Designer, Scrum Master, Quinn (QA), Tech Writer
• 50+ Guided Workflows: Structured processes organized into 4 development phases
• Scale-Domain-Adaptive: Automatically adjusts planning depth based on project complexity and domain
• AI Intelligent Help: /bmad-help provides contextual guidance, answers questions, and suggests next steps
• Party Mode: Multi-agent collaborative sessions for planning, troubleshooting, and discussion

Development Phases

Quick Flow (Simple Path)

For bug fixes, small features, clear scope:

1. /quick-spec - Analyze codebase, produce tech-spec with stories
2. /dev-story - Implement each story
3. /code-review - Validate quality

Full Planning Path (BMad Method)

For products, platforms, complex features:

1. /product-brief - Define problem, users, MVP scope
2. /create-prd - Full requirements with personas, metrics, risks
3. /create-architecture - Technical decisions and system design
4. /create-epics-and-stories - Break work into prioritized stories
5. /sprint-planning - Initialize sprint tracking
6. Per story: /create-story -> /dev-story -> /code-review

Agents in BMM Module

Extension Modules

Technical Architecture

Project Structure

BMAD-METHOD/
  src/
    bmm/                    # BMad Method module
      agents/               # 8 agent YAML definitions
      workflows/            # 9 workflow categories
      teams/                # Multi-agent team configurations
      data/                 # Supporting data files
      module.yaml           # Module manifest
      module-help.csv       # AI help knowledge base
    core/                   # Core framework utilities
    utility/                # CLI and helper utilities
  tools/
    cli/                    # npx bmad-method install
  docs/                     # Documentation (Astro/Starlight)
  website/                  # Marketing site

Agent Definition Format

Agents are defined in YAML files (*.agent.yaml) with:

• Role and expertise definition
• Available workflows and commands
• Knowledge and data file references
• Menu-driven interaction patterns

Workflow Structure

Each workflow directory contains:

• Workflow definition files
• Templates and prompts
• Output document specifications
• Validation rules

Installation Flow

# Install to any project
npx bmad-method install

# Follow prompts to select:
# - Core module (BMM)
# - Optional modules (TEA, BMGD, CIS)
# - IDE integration (Claude Code, Cursor, Windsurf)

IDE Integration

• Installs slash commands to .claude/commands/ or equivalent
• Agents accessible via /agent-name syntax
• Workflows triggered via /workflow-name commands
• AI help via /bmad-help with contextual awareness

Installation and Usage

Prerequisites

• Node.js v20.0.0 or higher
• AI IDE (Claude Code, Cursor, Windsurf, or similar)

Installation

npx bmad-method install

Using AI Help

/bmad-help
/bmad-help How should I build a web app for my TShirt Business?
/bmad-help I just finished the architecture, what's next?

Quick Flow Example

# Step 1: Analyze and create tech spec
/quick-spec

# Step 2: Implement stories
/dev-story

# Step 3: Code review
/code-review

Full Method Example

# Planning phase
/product-brief
/create-prd
/create-architecture

# Story breakdown
/create-epics-and-stories
/sprint-planning

# Per-story development
/create-story
/dev-story
/code-review

Relevance to Claude Code Development

Direct Applications

1. Structured Workflow Patterns: 50+ workflows provide templates for Claude Code skill workflows, showing how to break complex tasks into guided steps.

2. Agent Definition Format: YAML-based agent definitions demonstrate a portable, readable format for agent persona and capability specification.

3. Scale-Adaptive Intelligence: Methodology that adjusts based on project complexity shows patterns for context-aware skill activation.

4. AI-Assisted Help System: /bmad-help with contextual awareness demonstrates patterns for self-documenting, adaptive skill systems.

5. Multi-Agent Collaboration: Party Mode shows patterns for coordinating multiple specialized agents in collaborative sessions.

Patterns Worth Adopting

1. Phase-Based Organization: Grouping workflows into analysis, planning, solutioning, and implementation phases provides clear navigation.

2. Module System: Extension modules (BMB, TEA, BMGD, CIS) show how to package domain-specific agent collections.

3. Progressive Disclosure: Quick Flow (3 commands) vs Full Method (many commands) allows users to choose depth.

4. Knowledge-Driven Help: CSV-based help knowledge base enables AI-assisted guidance that evolves with installed modules.

5. Dual-Path Methodology: Simple path for quick tasks, full path for complex projects addresses different use cases.

Integration Opportunities

1. Workflow Import: BMAD workflows could inform Claude Code skill design for product development tasks.

2. Agent Personas: Agent definitions could be adapted as Claude Code agent configurations.

3. Help System Pattern: AI-assisted help could inform Claude Code skill documentation approaches.

4. Module Marketplace: BMAD's module system and upcoming community marketplace could inform Claude Code plugin ecosystem design.

5. IDE Neutral Design: Installation to multiple AI IDEs shows patterns for portable agent systems.

Comparison with Claude Code Ecosystem

References

Research Method: Information gathered from GitHub repository README, GitHub API (repository metadata, directory structure, package.json), and official documentation links. Statistics verified via direct GitHub API calls on 2026-02-01.

Freshness Tracking

Review Triggers:

• Version 6.0.0 stable release (exit beta)
• GitHub stars milestone (40K, 50K)
• New module releases
• Breaking changes to agent/workflow format
• Community marketplace launch
• New IDE integrations