6.2 KiB

Raw Blame History

Getting Started

Welcome to the BMAD-Spec Orchestrator - your enterprise-grade AI development team in Claude Code.

What You'll Build

In the next 10 minutes, you'll:

Set up the orchestrator system
Create your first AI-orchestrated application
Experience the full agent workflow from idea to deployment

Prerequisites

Claude Code installed - Get it at claude.ai/code
Basic project idea - Something simple like "task management app" or "blog platform"

Step 1: Installation (2 minutes)

Clone and Setup

# Clone the repository
git clone https://github.com/oimiragieo/BMAD-SPEC-KIT.git
cd BMAD-SPEC-KIT

# Copy to your project directory
mkdir my-first-bmad-project
cp -r bmad-spec-orchestrator-kit/.claude my-first-bmad-project/
cd my-first-bmad-project

Open in Claude Code

# Open your project in Claude Code
code .

Claude Code automatically detects the .claude folder and loads the orchestrator system.

Step 2: Activate the System (30 seconds)

In Claude Code, type:

I need to create a simple task management web application with user authentication and project collaboration features.

What happens next:

Claude detects this as a development request
The BMAD Orchestrator activates
You'll see a complexity assessment and recommended workflow

Step 3: Watch the Agents Work (5 minutes)

The Analyst Agent Starts

Claude will likely activate the Analyst first:

Uses think hard to research task management market
Creates competitive analysis
Generates project brief with market insights

The PM Agent Takes Over

The Product Manager agent:

Uses think harder for comprehensive planning
Transforms the brief into a detailed Product Requirements Document (PRD)
Defines user stories and feature prioritization

The Architect Agent Designs

The Architect agent:

Uses ultrathink for system architecture decisions
Creates security-first technical architecture
Selects optimal technology stack with detailed rationale

Step 4: Experience the Quality (2 minutes)

Notice how each agent:

Follows Enterprise Rules

Professional writing without AI jargon
Security-first approach to all decisions
Accessibility compliance (WCAG 2.1 AA)
Performance optimization throughout

Uses Claude Thinking Optimization

ultrathink - Critical architecture and security decisions
think harder - Complex analysis and trade-offs
think hard - Important research and design work
think - Standard reasoning and validation

Maintains Context

Each agent builds on previous work
No information loss between handoffs
Comprehensive validation at each step

Step 5: Move to Implementation (Optional)

If you want to continue to development:

I approve the architecture. Let's move to implementation with the Developer agent.

The Developer agent will:

Create project structure
Implement core functionality
Write comprehensive tests
Follow all enterprise coding standards

What You Just Experienced

🎯 Precision Orchestration

10 specialized agents each with distinct expertise
Automatic complexity scoring (1-10 scale) determines workflow
Context-aware handoffs between agents

🛡️ Enterprise Security

No dangerous permission bypasses
Security-first development practices
Comprehensive audit trails
Zero-trust architecture principles

⚡ Claude Optimization

Thinking triggers optimized for Claude's capabilities
Maximum decision quality through strategic thinking deployment
Context preservation across long workflows

📊 Quality Gates

Cross-agent validation ensures consistency
Automated quality thresholds
Error recovery with improved context
Continuous learning from patterns

Common First-Time Patterns

Simple Projects (Complexity 1-3)

"Create a personal blog with markdown support"

Analyst → PM → Developer workflow
Fast execution, focused features

Medium Projects (Complexity 4-6)

"Build a customer relationship management system"

Full planning phase with Architect
UX Expert creates detailed user flows
QA agent provides comprehensive testing

Complex Projects (Complexity 7-10)

"Design a multi-tenant SaaS platform with real-time collaboration"

All 10 agents participate
Multiple quality gate validations
Extensive documentation and testing

Next Steps

Learn More About Agents

Agent Reference - Deep dive into all 10 agents
Workflow Guide - Understanding complexity scoring

Customize Your System

Templates Guide - Using the 11 document templates
Configuration - Advanced customization

Try Advanced Examples

Basic Projects - More starter examples
Enterprise Projects - Complex workflows

Quick Tips

Trigger Agent Activation

Use clear, project-oriented language: ✅ "I need to create a web application for..." ✅ "Help me build a system that..." ✅ "Design and implement a solution for..."

Get Specific Agent Help

✅ "Activate the Architect to design a microservices system" ✅ "I need the UX Expert to create user flows" ✅ "Have the QA agent review this implementation"

Maximize Claude Performance

The agents automatically use optimal thinking triggers, but you can request specific analysis: ✅ "Use ultrathink to evaluate this architecture decision" ✅ "Think harder about the security implications"

Troubleshooting

Agents Don't Activate

Verify .claude folder is in project root
Check that Claude Code detects the folder
Try restarting Claude Code

Missing Context Between Agents

Ensure all agent folders are present in .claude/agents/
Verify CLAUDE.md is in the root .claude/ folder

Need Help

Welcome to the future of AI-orchestrated development

6.2 KiB Raw Blame History