A production-ready starter template for AI-Driven Development projects.

🇷🇺 Русская версия | 📖 Full Guide | 📜 Manifesto

This is a boilerplate repository for quickly starting projects using the AI-Driven Development (AIDD) methodology. It includes:

• ✅ Pre-configured project structure for AIDD workflow
• ✅ AI agent configurations for Claude Code
• ✅ Context management templates
• ✅ Specification examples
• ✅ TypeScript + Jest setup
• ✅ Best practices documentation

# Clone to your project
cp -r ~/AI/boiler ~/projects/my-new-project
cd ~/projects/my-new-project

# Remove git history
rm -rf .git
git init

npm install

Update CLAUDE.md with your project context:

# Project: [Your Project Name]

## Description
[What your project does]

## Tech Stack
- TypeScript 5.x
- Node.js 20.x
- [Your database]
- [Your frameworks]

claude-code .

# Start with a specification
"Let's follow AIDD methodology. Create a spec for [your feature]"

# Plan implementation
"Use the planner agent to create an implementation plan"

# Develop
"Use the backend-developer agent to implement the plan"

# Test
"Use the test-runner agent to create and run tests"

.
├── CLAUDE.md                 # 🧠 Main AI context (auto-loaded)
├── .claude/                  # 🤖 AI configuration
│   ├── agents/              # Specialized AI agents
│   │   ├── planner.md       # Planning & architecture
│   │   ├── backend-developer.md  # Implementation
│   │   ├── test-runner.md   # Testing & quality
│   │   └── code-reviewer.md # Code review
│   └── commands/            # Custom commands
├── ai-docs/                 # 📚 AI knowledge base
│   ├── context.md          # Detailed project context
│   ├── conventions.md      # Coding standards
│   ├── current-task.md     # Current work tracking
│   └── decisions/          # Architectural decisions
├── specs/                   # 📋 Specifications (write first!)
│   ├── features/           # Feature specifications
│   └── api/                # API specifications
├── src/                     # 💻 Source code (AI-generated)
│   ├── domain/             # Business logic (DDD)
│   ├── api/                # API endpoints
│   └── infrastructure/     # External services
├── tests/                   # 🧪 Tests (AI-generated)
└── docs/                    # 📖 Human documentation

• TypeScript with strict mode
• Jest for testing
• ESLint & Prettier for code quality
• Node.js 20+ runtime

• Planner - Breaks down specs into tasks
• Backend Developer - Implements features
• Test Runner - Creates and runs tests
• Code Reviewer - Reviews code quality

• CLAUDE.md - Main context template
• ai-docs/context.md - Extended context
• ai-docs/conventions.md - Coding standards
• specs/examples/ - Specification examples

graph LR
    A[Write Spec] --> B[AI Plans]
    B --> C[AI Implements]
    C --> D[AI Tests]
    D --> E[AI Reviews]
    E --> F[Human Validates]
    F --> G[Deploy]

# specs/features/user-auth.md

## Feature: User Authentication
- JWT-based authentication
- Email/password login
- Session management
- Password reset flow

"Read specs/features/user-auth.md and create an implementation plan"

"Use the backend-developer agent to implement authentication"
"Use the test-runner agent to test the auth module"

• Keep CLAUDE.md under 200 lines
• Update current-task.md daily
• Document decisions in ai-docs/decisions/

# Planning
"Use the planner agent to break down the payment feature"

# Development
"Use the backend-developer to implement the user service"

# Testing
"Use test-runner to achieve 90% test coverage"

# Review
"Use code-reviewer to check for security issues"

# Documentation
"Update the API documentation with the new endpoints"

Create .claude/agents/your-agent.md:

---
name: your-agent
tools: Read, Write, Bash
---

You are a specialized agent for [purpose].
Your responsibilities:
- [Responsibility 1]
- [Responsibility 2]

Create .claude/commands/your-command.md:

---
name: your-command
description: Does something specific
---

Steps:
1. [Step 1]
2. [Step 2]

• Node.js 20+
• npm or yarn
• Claude Code subscription
• Git

• 📖 Complete AIDD Guide - Comprehensive methodology guide
• 📜 AIDD Manifesto - Philosophy and principles
• 🔗 Claude Code Docs
• 💬 AIDD Community

This boilerplate is open for improvements. Feel free to:

• Add new agent configurations
• Improve project structure
• Share specification templates
• Report issues

MIT - Use freely in your projects

Start your AIDD journey today! 🚀

# Your first AIDD command:
"Let's build something amazing using AIDD methodology"