Generate both files in current directory:
`bash
aidottxt
`

Generate in a specific directory:
`bash
aidottxt --out ./public
`

Preview without writing files:
`bash
aidottxt --dry-run
`

Generate only ai.txt:
`bash
aidottxt --ai-only --out ./dist
`

Generate only llms.txt:
`bash
aidottxt --llms-only
`

Generate only robots.txt and humans.txt:
`bash
aidottxt --robots-only --humans-only
`

Output Files

$3

Located at the root of your output directory. Format:

`


ai.txt for My Site

Training: allow
Retention: allow
Commercial-Use: allow
Rate-Limit-RPS: 10
`

$3

Located at .well-known/llms.txt. Format:

`json
{
"version": "1.0",
"site_name": "My Site",
"contact": "mailto:admin@example.com",
"models": ["*"],
"capabilities": ["chat", "embed", "fine_tune", "crawl", "train"],
"policy": {
"allow": ["/*"],
"disallow": []
},
"training": true,
"retention": true,
"commercial_use": true,
"rate_limit_rps": 10
}
`

$3

Located at the root of your output directory. Format:

`
User-agent: *
Allow: /api/*
Allow: /docs/*
Disallow: /admin/*
Disallow: /private/*
Crawl-delay: 1
Sitemap: https://example.com/sitemap.xml
`

$3

Located at the root of your output directory. Format:

`
/ TEAM /
Developer: John Doe
Contact: https://github.com/johndoe

Designer: Jane Smith
Contact: https://twitter.com/janesmith

/ THANKS /
Open Source Community
Coffee

/ SITE /
Last update: 2024/01/01
Language: English
Standards: HTML5
Components: Node.js, JavaScript
`

Programmatic Usage

You can also use this package programmatically:

`javascript
import { buildAiTxt, buildLlmsJson } from "@profullstack/ai-dot-txt";

const config = {
siteName: "My Site",
baseUrl: "https://example.com",
contact: "mailto:admin@example.com",
models: ["gpt-4", "claude-3"],
capabilities: ["chat", "embed"],
allowPaths: ["/api/", "/docs/"],
disallowPaths: ["/admin/*"],
training: "allow",
retention: "allow",
commercialUse: "disallow",
rateLimitRps: 10,
};

// Generate ai.txt content
const aiTxtContent = buildAiTxt(config);

// Generate llms.txt content
const llmsTxtContent = buildLlmsJson(config);
`

Configuration Options

When running interactively, you'll be prompted for:

- Site Name: Your website or application name
- Base URL: Your public base URL
- Contact: Contact information (email, URL, etc.)
- Models: Comma-separated list of allowed AI models (use * for all)
- Capabilities: Select from: chat, embed, fine_tune, crawl, train
- Allow Paths: Comma-separated paths to allow (e.g., /, /api/)
- Disallow Paths: Comma-separated paths to disallow (e.g., /admin/*)
- Training Permission: Allow or disallow AI training on your content
- Retention Permission: Allow or disallow data retention
- Commercial Use: Allow or disallow commercial use of your content
- Rate Limit: Requests per second limit

Development

$3

`bash
pnpm install
`

$3

`bash
pnpm test
`

$3

`bash
pnpm lint
`

$3

`bash
pnpm lint:fix
`

$3

`bash
pnpm format
`

Project Structure

`
.
├── src/
│ ├── args.js # CLI argument parser
│ ├── generators.js # File content generators
│ ├── cli.js # Main CLI entry point
│ └── index.js # Programmatic API exports
├── tests/
│ ├── args.test.js # Argument parser tests
│ └── generators.test.js # Generator tests
├── package.json
├── eslint.config.js
├── .prettierrc
└── README.md
`

Requirements

- Node.js >= 20.0.0
- pnpm (recommended) or npm

License

MIT

Contributing

Contributions are welcome! Please follow these steps:

1. Fork the repository
2. Create a feature branch
3. Write tests for your changes
4. Ensure all tests pass (pnpm test)
5. Ensure code is linted (pnpm lint`)
6. Submit a pull request

Related Standards

This project follows several web standards and specifications:

- humans.txt - See humans.txt for team and site information
- robots.txt - Standard for web robots
- llms.txt - LLM-friendly documentation format
- ai.txt - AI agent interaction specification

Author

Profullstack, Inc.