PentAGI
Autonomous AI agents executing penetration testing tasks in isolated Docker environments.
Run specialized AI security agents with built-in professional pentesting tools and sandboxing.
PentAGI is an autonomous AI agent system designed to execute complex penetration testing tasks inside isolated Docker environments. It suits security engineers and information security professionals who want to automate security testing using specialized multi-agent architectures.
What it is
The system features a team of specialized AI agents equipped with over 20 professional security tools including nmap, metasploit, and sqlmap. It integrates a smart memory system, Graphiti knowledge graph via Neo4j, and external search APIs to gather intelligence and perform testing.
Who it's for
It suits security researchers and ethical hackers with Docker and LLM management experience who need automated penetration testing workflows. It does not suit teams looking for a simple point-and-click scanner or predefined breach and attack simulation campaigns.
Key features
Secure Sandboxed Execution
All penetration testing operations run inside an isolated Docker environment.
Professional Tool Suite
Built-in integration with over 20 security utilities including nmap and metasploit.
Specialized Agent Team
Task delegation across research, development, and infrastructure AI agents.
Knowledge Graph Tracking
Graphiti-powered knowledge graph using Neo4j for semantic relationship tracking.
Up and running in 3 steps
Our honest verdict
What's great
- Supports over 10 different LLM providers and custom endpoints like Ollama and OpenAI
- Includes Grafana, Prometheus, and Langfuse integrations for deep observability
- Exposes full-featured REST and GraphQL APIs with Bearer token authentication
- Maintains long-term memory of research results using PostgreSQL with pgvector
Where it falls short
- Requires managing multiple dependent services including PostgreSQL, Neo4j, and Docker containers
- Consumes significant external API usage and costs when driving large LLMs continuously
- Not an out-of-the-box BAS platform with predefined campaigns; requires active configuration
Pricing & license
The software is free under its license, but running it requires provisioning your own server infrastructure, database instances, and paying for underlying LLM API usage if using commercial models.
Open remote AI-training roles
873 open remote AI-training roles, updated November 15, 2026; most pay $20–$170 an hour. Zekai lists them free; the application goes to the employer’s platform.
Alternatives
These reviewed AI tools do the same job instantly — from our scored directory.
Use it, or have us set it up
FAQ
What penetration testing tools are included?
The system includes over 20 tools such as nmap, metasploit, and sqlmap running inside isolated Docker containers.
Can I use local LLMs with PentAGI?
Yes, it supports Ollama and custom endpoints alongside major cloud providers like OpenAI, Anthropic, and AWS Bedrock.
