Skip to content

Repository files navigation

👁️ Erebus: AI-Powered Dark Web OSINT Tool

Erebus Banner

Erebus is an advanced OSINT tool for investigating dark web threats. It combines Tor-based crawling with local LLMs (Ollama) to find, filter, and visualize hidden service content.

Features

  • Tor Integration: Automatically handles Tor connections, circuit rotation, and port detection (9050/9150).
  • Search Engines: Crawls multiple engines like Ahmia, Torch, and Haystak.
  • Direct Scraping: Scrape specific .onion URLs directly with liveness checks and form detection.
  • AI Analysis: Uses Ollama (LLM) to refine queries and summarize content.
  • Visual Graph: Interactive node-link diagram showing connections between findings.
  • Artifact Detection: Auto-extracts Emails, Crypto Addresses, SSNs, and more.
  • Reporting: Generates professional HTML/Markdown reports.

Installation

Method 1: Docker (Recommended)

This method handles the Tor connection for you automatically.

  1. Prerequisites: Docker Desktop, Ollama (running on host).
  2. Run:
    docker-compose up --build
  3. Access: Open http://localhost:8501.

Method 2: Local Python

  1. Prerequisites:

    • Python 3.10+
    • Tor Service:
      • Linux/Mac: Run tor (listens on 9050).
      • Windows: Run Tor Browser (listens on 9150) OR install Tor service.
    • Ollama running (ollama serve).
  2. Install:

    pip install -r requirements.txt
  3. Run:

    • Windows: Double-click run_windows.bat
    • Linux/Mac: streamlit run app.py

    Or use the CLI:

    python cli.py --query "leaked database" --report

Configuration

Edit .env (copy from .env.example if needed) or config.py to change:

  • TOR_PROXY_URL: if using a different proxy.
  • OLLAMA_BASE_URL: if Ollama is remote.
  • TELEGRAM_BOT_TOKEN: for alerts.

Disclaimer

This tool is for educational and authorized testing purposes only. Accessing the dark web carries risks. Use responsibly.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

Packages

Contributors

Languages