Skip to main content

Welcome to PentAGI

PentAGI is an innovative tool for automated security testing that leverages cutting-edge artificial intelligence technologies. The project is designed for information security professionals, researchers, and enthusiasts who need a powerful and flexible solution for conducting penetration tests. All operations are performed in a sandboxed Docker environment with complete isolation, ensuring your security testing is both powerful and safe.

Quick Start

Get PentAGI running in under 5 minutes with Docker Compose

Installation Guide

Comprehensive setup instructions and system requirements

Architecture

Understand PentAGI’s microservices and agent system

Configuration

Configure LLM providers, search engines, and integrations

Watch PentAGI in Action

See how PentAGI autonomously conducts penetration testing: PentAGI Overview Video

Key Features

Secure & Isolated

All operations performed in sandboxed Docker environment with complete isolation

Fully Autonomous

AI-powered agent that automatically determines and executes penetration testing steps

Professional Tools

Built-in suite of 20+ professional security tools including nmap, metasploit, sqlmap, and more

Smart Memory

Long-term storage of research results and successful approaches for future use

Knowledge Graph

Graphiti-powered knowledge graph using Neo4j for semantic relationship tracking

Web Intelligence

Built-in browser for gathering latest information from web sources

Advanced Search

Integration with Tavily, Traversaal, Perplexity, DuckDuckGo, Google, and Searxng

Team of Specialists

Delegation system with specialized AI agents for research, development, and infrastructure

Comprehensive Monitoring

Detailed logging with Grafana/Prometheus integration for real-time observation

Detailed Reporting

Generation of thorough vulnerability reports with exploitation guides

Smart Containers

Automatic Docker image selection based on specific task requirements

Modern Interface

Clean and intuitive web UI for system management and monitoring

API Integration

Support for REST and GraphQL APIs for seamless external system integration

Persistent Storage

All commands and outputs stored in PostgreSQL with pgvector extension

Scalable Architecture

Microservices-based design supporting horizontal scaling

Self-Hosted

Complete control over your deployment and data

Flexible Authentication

Support for OpenAI, Anthropic, Ollama, AWS Bedrock, Google AI, and custom configurations

Quick Deployment

Easy setup through Docker Compose with comprehensive environment configuration

Core Capabilities

AI-Powered Multi-Agent System

PentAGI employs a sophisticated multi-agent architecture where specialized AI agents collaborate to conduct penetration tests:
  • Orchestrator: Coordinates overall testing strategy and delegates tasks
  • Researcher: Analyzes targets and gathers intelligence
  • Developer: Plans attack strategies and exploit chains
  • Executor: Performs actual security testing and tool execution
  • Pentester: Specialized agent for vulnerability assessment
  • Coder: Generates and analyzes code for exploits

Advanced Memory System

PentAGI maintains three types of memory for intelligent decision-making:
  • Long-term Memory: Persistent storage of knowledge, techniques, and tool capabilities in vector database
  • Working Memory: Active context including current goals, system state, and task progress
  • Episodic Memory: Historical actions, results, and learned success patterns

Professional Security Toolkit

Built-in integration with industry-standard penetration testing tools:
  • Network Scanning: nmap, masscan
  • Vulnerability Assessment: OpenVAS, Nessus
  • Web Application Testing: sqlmap, Burp Suite, OWASP ZAP
  • Exploitation: Metasploit Framework
  • Password Cracking: John the Ripper, Hashcat
  • And 15+ more professional tools

Supported LLM Providers

PentAGI supports multiple AI providers for maximum flexibility:

OpenAI

GPT-4, GPT-5, o-series reasoning models

Anthropic

Claude 3.5, Claude 4 with extended thinking

Google AI

Gemini 2.0, 2.5 with multimodal support

AWS Bedrock

Multi-provider enterprise foundation models

Ollama

Local inference with zero-cost operation

Custom

OpenRouter, DeepInfra, DeepSeek, Moonshot, and more

System Requirements

Before installing PentAGI, ensure your system meets these minimum requirements:
  • Docker and Docker Compose installed
  • 2 vCPU minimum (4+ recommended)
  • 4GB RAM minimum (8GB+ recommended)
  • 20GB free disk space
  • Internet access for downloading images and updates
  • At least one LLM provider API key (OpenAI, Anthropic, Gemini, or AWS Bedrock)

Community & Support

Discord Community

Connect with security researchers and AI enthusiasts

Telegram Group

Get support and share insights with fellow ethical hackers

Next Steps

Ready to get started? Choose your path:

Quick Start

Launch PentAGI in 5 minutes with Docker Compose

Full Installation

Comprehensive setup with all configuration options
Ethical Use Only: PentAGI is designed for authorized security testing only. Always obtain proper authorization before testing any systems you do not own. Unauthorized access to computer systems is illegal.

Build docs developers (and LLMs) love