Skip to main content

axag init

Initialize AXAG configuration in your project with interactive setup prompts.

Usage​

axag init [options]

Examples​

# Interactive setup
axag init

# Overwrite existing config
axag init --force

Options​

FlagDefaultDescription
--forcefalseOverwrite existing configuration file

What It Creates​

Running axag init creates an axag.config.json in your project root:

{
  "$schema": "https://axag.dev/schema/v1/config.json",
  "outputDir": ".axag",
  "domain": "ecommerce",
  "ai": {
    "enabled": false,
    "provider": "openai",
    "model": "gpt-4o"
  },
  "scanner": {
    "maxPages": 10,
    "headless": true,
    "timeout": 30000,
    "excludePatterns": ["logout", "signout", "/admin"]
  },
  "validation": {
    "conformanceLevel": "AA",
    "strict": false
  }
}

Interactive Prompts​

The init wizard asks:

  1. Domain — What type of application? (e-commerce, CRM, analytics, enterprise, etc.)
  2. AI inference — Enable AI-powered inference? (requires API key)
  3. AI provider — OpenAI or Anthropic?
  4. Max pages — How many pages to crawl?
  5. Conformance level — Target validation level (A, AA, AAA)?

Config File Locations​

axag-cli looks for configuration in this order:

  1. axag.config.json
  2. .axagrc
  3. .axagrc.json
  4. .axagrc.yaml
  5. "axag" key in package.json