#55th of 5 in Security

pentestgpt

Autonomous pentest agent that drives Claude Code or Codex against a target

Stars: 15.9k
License: MIT
Last commit: Jul 2026
Last release: Dec 2025
Language: Python

Missed the latest GitHub check; these facts may be older.

Overview

PentestGPT runs a staged pipeline against a target IP or URL, using Claude Code or Codex as the backend to run tools and reason. CTF mode goes recon, exploit, walkthrough; pentest mode goes asset discovery, vulnerability identification, report. A separate pentestgpt-legacy mode keeps the interactive three-session design from the USENIX Security 2024 paper and supports several providers plus Ollama.

Who it is for: Pentesters and CTF players running authorized tests with LLM agents

Strengths

Staged pipelines for CTF and pentest modes, with saved and resumable sessions
Legacy interactive mode supports OpenAI, Anthropic, Gemini, DeepSeek, xAI, Qwen, Kimi and Ollama
Docker image bundles Claude Code and Codex CLIs, with logins persisted in volumes
Backed by a USENIX Security 2024 paper; reports 90/104 on an XBOW suite run in Dec 2025

Weaknesses

Autonomous pipeline supports only Claude Code and Codex backends
Needs an authenticated Claude Code or Codex CLI, so a paid account or API access
Sends anonymous usage telemetry to Langfuse by default; opt out with a flag
XBOW result is historical; the shipped CLI has no XBOW runner

What it needs

no GPU
Docker + Compose
Needs Python 3.12+, uv, Claude Code CLI, Codex CLI, Langfuse (telemetry, can be disabled)
Models: Claude, OpenAI, Google Gemini, DeepSeek, xAI Grok

Also in Security

See all 5

Also in Security
Rank	Project	Score
1	shannonAutonomous pentester that exploits web apps and APIs using source code analysis48.9k starsAGPL-3.0	76 out of 100
2	StrixAutonomous AI pentesting agents that validate vulnerabilities with working exploits68.6k starsApache-2.0	75 out of 100
3	pentagiAutonomous multi-agent penetration testing system with sandboxed security tools25.3k starsMIT	71 out of 100
#4	ifixaiPython CLI that audits AI agents for policy, access and reliability failures24.2k starsApache-2.0	59 out of 100

How to choose

Run them only against targets you own or are authorized to test; check the tool's scope controls.
Check which model backends are supported and whether findings stay on your host.
Look at how exploitation is sandboxed (containers, VMs) before pointing it at production.

Health

Releases: Not checked
Top contributor: Not checked
Security advisories: Not checked
Package advisories: Not checked
OpenSSF Scorecard: Not checked

Not read for this project yet: each refresh reads the facts of a few projects. Bots and coding-agent accounts are left out of the commit counts. Not part of the score: how we check.

License in plain words

MIT Permissive

Use it at work: Yes
Change it and rebrand it: Yes
Sell it or host it for others: Yes

Your part: Keep the copyright and license notice in copies you ship.

A plain summary, not legal advice. Read the license on GitHub ↗ (opens in a new tab)

Where to run it

RepoCloud ↗ (opens in a new tab)One-click hosting for about 890 open-source apps, each on its own VPS.

Referral links: if you sign up through one, we get a commission. They never change a rank.

#5 of 5 in Security on Archestack, as of 15 Oct 2026.

In your README

The badge updates twice a day.

[![Archestack](https://archestack.com/badge/greydgl/pentestgpt.svg)](https://archestack.com/best-of-ai/projects/pentestgpt/)

How badges work

On your site or blog

A card with the rank, score and signals, in light or dark to match the reader.

<iframe src="https://archestack.com/embed/best-of-ai/pentestgpt/" title="pentestgpt on Archestack" width="400" height="270" style="border:0;max-width:100%" loading="lazy"></iframe>

Preview the card

Best of Open-Source AI on GitHub

Each review comes from the project's README and facts we check on GitHub. The fix form opens with pentestgpt filled in.