Skip to content

Quick Start

Let's get you set up with ai-rulez. This takes about 5 minutes and will sync all your AI tools automatically.

Step 1: Create your config

Navigate to your project root and run the init command. We recommend using the AI-powered configuration for the best experience:

# AI-powered initialization (recommended)
ai-rulez init "My Project" --preset popular --use-agent claude

# Or simple initialization without AI assistance
ai-rulez init "My Project" --preset popular

The AI-powered approach will: - Analyze your codebase structure and tech stack - Generate project-specific rules and documentation
- Create specialized agents for your workflow - Set up comprehensive configuration tailored to your project

If you don't have Claude installed or prefer a simpler setup, the basic command still works great and you can always add content later using the CLI.

Basic initialization creates: 

$schema: https://github.com/Goldziher/ai-rulez/schema/ai-rules-v2.schema.json
metadata:
  name: "My Project"
presets:
  - "popular"

AI-powered initialization creates a much richer configuration with rules, sections, and agents based on your actual codebase. For example, a Go project might get: 

$schema: https://github.com/Goldziher/ai-rulez/schema/ai-rules-v2.schema.json
metadata:
  name: "My Project"
  description: "Go microservice with REST API"
presets:
  - "popular"
rules:
  - name: "Go Standards"
    priority: high
    content: "Follow standard Go layout (cmd/, internal/, pkg/)"
  - name: "Error Handling"  
    priority: critical
    content: "Always handle errors properly, use wrapped errors"
sections:
  - name: "Architecture"
    content: "Clean architecture with dependency injection..."
agents:
  - name: "go-expert"
    description: "Go specialist for backend development"

Step 2: Add your project details (if needed)

If you used the AI-powered initialization, your configuration already includes comprehensive project-specific rules and documentation. You can skip to Step 3!

If you used the basic initialization, add some context about your project:

# Add a rule for your tech stack
ai-rulez add rule "Tech Stack" --priority critical --content "Frontend: React, TypeScript, Tailwind CSS. Backend: Go, PostgreSQL."

# Add a rule for your development workflow
ai-rulez add rule "Workflow" --priority high --content "All new features must have 90%+ unit test coverage."

Pro tip: You can also re-run the init command with AI assistance to enhance an existing configuration: 

# Enhance existing config with AI analysis
ai-rulez init "My Project" --preset popular --use-agent claude --yes

Using the CLI keeps your config valid and makes it easy to script changes later.

Step 3: Generate files

Run the generate command to create all your AI instruction files.

ai-rulez generate

This reads your ai-rulez.yml file and creates all the necessary configuration files for the tools included in the "popular" preset.

Automatic .gitignore Updates

Generated files are automatically added to .gitignore to keep your repository clean. These files are regenerated from your config, so they don't need to be committed. To disable this, use --update-gitignore=false or set gitignore: false in your config.

Step 4: Validate (optional)

Check that everything looks good:

ai-rulez validate

You're All Set!

That's it! You now have a powerful, maintainable, and synchronized configuration for all your AI tools. Your AI assistants will now have the context they need to provide much more accurate and relevant responses.

What's Next?