AutoRedTeam-Orchestrator is an AI-driven automation framework that exposes 101 MCP tools within a single orchestrated workflow. It integrates with MCP-capable editors and assistants, enabling natural language prompts to select, configure, and execute security tooling in a coordinated attack simulation or red-team exercise. Use it to plan reconnaissance, vulnerability discovery, exploitation, post-exploitation, and reporting steps, all while leveraging a knowledge graph and Monte Carlo Tree Search guided attack planning. The server aggregates tool results, performs multi-method validation to reduce false positives, and can generate JSON/HTML/Markdown professional reports automatically.

Prerequisites:

• Python 3.10+ and pip
• Git
• Optional: Python virtual environment tools (venv)

Installation steps:

1. Clone the repository git clone https://github.com/Coff0xc/AutoRedTeam-Orchestrator.git cd AutoRedTeam-Orchestrator

2. Create and activate a virtual environment python -m venv venv

   On Windows

   venv\Scripts\activate.bat

   On macOS/Linux

   source venv/bin/activate

3. Install dependencies pip install -r requirements.txt

4. Prepare configuration (optional)

   • Copy and modify config.yaml if provided, or rely on defaults.

5. Run the MCP server python -m autoredteam_orchestrator

Note: If the package is installed via pip in editable mode, you can run the module equivalent as configured by your environment. Ensure any required external services (e.g., databases, caches) are accessible as configured in your environment.

Tips and common issues:

• Ensure Python 3.10+ is used to match the project’s dependencies.
• If you encounter missing native extensions, install system dependencies (e.g., build tools for cryptography or zstandard).
• Set ARTE_API_KEY if you are using cloud-backed MCP integrations; otherwise, you can run in a local, offline mode.
• Use ARTE_LOG_LEVEL=DEBUG during troubleshooting to get verbose logs.
• If using a YAML/JSON config, validate syntax before starting the server to avoid startup failures.
• For MCP editors to interact correctly, ensure the MCP endpoints are reachable (network/firewall adjustments may be required).
• Review the knowledge graph and MCTS configuration to tailor attack planning to your environment.