Skip to content
hackingtools.ai
PentAGI

PentAGI

PentAGI is a freemium, open-source, self-hosted system of AI agents used to run autonomous penetration tests against targets you are authorized to test.

0

0 upvotes · 0 downvotes

No ratings yet

Pricing
Freemium
Platforms
LinuxWindowsmacOS
License
Open source

What it is

A self-hosted multi-agent platform. You describe a target and a goal in plain language, then specialized agents plan the work, run tools such as nmap, Metasploit and sqlmap in isolated Docker containers, remember what they learn, and write up the result.

What it helps with

- Hands-off scanning and exploitation checks on a scoped target, with 20+ bundled tools
- Reports you can export as Markdown or PDF
- Using the model you choose, including local models through Ollama or vLLM

Who it's for

Security engineers, penetration testers and red teams testing systems they own or are authorized to test.

How defenders detect it
Elastic Security Labs reports that Google Threat Intelligence Group and Mandiant have observed PentAGI in incident response engagements. It lists connections from compromised hosts to TCP ports 28000 to 29999 on the PentAGI host as a sign of use, and says an installation on your network reports to update.pentagi.com every three hours.

Worth knowing

It does not pause for human approval by default, because the Compose file sets ASK_USER to false. A separate Enterprise edition under a commercial license adds client-ready reports.

Which systems does it run on?
Its installer is offered for Linux, Windows and macOS, and it needs Docker with Compose.

How does it compare with Strix, Shannon and Vigilo?
In our view, those three focus on web applications and APIs, while PentAGI is a broader agent system that also drives network tools.

Discussion & reviews

0 comments

Your rating (optional)

0/4,000

No contributions yet. Be the first to review or comment.

← Back to directory