🎭 Phantom AI

Adaptive AI-powered web crawler with behavioral mutation engine.

Powered by Claude Sonnet 4.5+ - the crawler evolves its behavior to match what targets expect.

✨ Features

- 🧠 AI-Driven: Uses Claude Sonnet 4.5 for behavioral analysis
- 🔄 Self-Mutating: DNA evolves based on target responses
- 🚦 Green Light System: Trust-based progression from RED → GREEN
- 📊 Real-time Dashboard: Web UI for monitoring and control
- 🎯 Multi-target: Crawl multiple sites simultaneously
- 🔒 Stealth: Adapts to avoid detection

🚀 Quick Start

$3

``bash


Via npx (no install)

npx phantom-ai-crawler

Or install globally

npm install -g phantom-ai-crawler
phantom-ai

Or local install

npm install phantom-ai-crawler
npx phantom-ai
`

$3

`bash


1. Setup (configure API key)

phantom-ai setup

2. Start server

phantom-ai start

3. Open dashboard

http://localhost:8081

`

🛠️ Commands

`bash
phantom-ai # Start server (default)
phantom-ai start # Start backend + dashboard
phantom-ai setup # Run configuration wizard
phantom-ai status # Check configuration
`

$3

`bash
phantom-ai start -p 3000 # Backend on port 3000
phantom-ai start -u 8080 # Dashboard on port 8080
phantom-ai start --setup # Force reconfiguration
`

🔧 Configuration

The setup wizard will ask for:

1. Anthropic API Key - Get from https://console.anthropic.com
2. Claude Model - Recommended: claude-4-5-sonnet-20250929
3. Backend Port - Default: 4000
4. Dashboard Port - Default: 8081

Config is saved to .env in your working directory.

📁 Project Structure

`
phantom-ai/
├── .env # Your configuration
├── phantom.db # SQLite database
├── src/
│ ├── cli.ts # CLI entry point
│ └── server/ # Backend API
├── dashboard/ # Web UI
└── prisma/ # Database schema
`

🔌 API Endpoints

| Endpoint | Method | Description |
|----------|--------|-------------|
| /health | GET | Server status |
| /api/targets | GET | List targets |
| /api/targets | POST | Create target |
| /api/targets/:id | GET | Target details |
| /api/dna/:id/current | GET | Current DNA |
| /api/mcp/analyze/:id | POST | Run MCP analysis |

🧬 How It Works

1. Discovery (🔴 RED): Initial reconnaissance phase
2. Learning (🟡 YELLOW): Testing behavioral patterns
3. Established (🟢 GREEN): Trusted access achieved
4. Maintenance: Continuous adaptation

The MCP (Model Context Protocol) analyzes each interaction and suggests DNA mutations to improve trust scores.

📝 Requirements

- Node.js 18+
- Anthropic API key
- Claude Sonnet 4.5+ access

🐛 Troubleshooting

Port already in use:
`bash
phantom-ai start -p 3000 -u 8080 # Use different ports
`

Database issues:
`bash
rm phantom.db # Reset database
phantom-ai setup
`

API key not working:
`bash
phantom-ai status # Check configuration
phantom-ai setup # Reconfigure
`

🏗️ Development

`bash
git clone https://github.com/0x2e8-gc/phantom-ai-crawler.git
cd phantom-ai
npm install
npm run dev # Development mode
npm run build # Build for production
`

📦 Publishing to NPM

`bash


1. Update version

npm version patch

2. Build

npm run build

3. Publish

npm publish

Or dry run first

npm publish --dry-run
``

📄 License

MIT

---

Built with 🎭 by 0x2e8