# vxcontrol/pentagi

Fully autonomous AI Agents system capable of performing complex penetration testing tasks

Repository: https://github.com/vxcontrol/pentagi
Canonical: https://ross.abutalabs.com/products/pentagi
Homepage: https://pentagi.com
Language: Go
License: MIT
License Family: permissive
Topics: ai-agents, ai-security-tool, autonomous-agents, golang, graphql, multi-agent-system, penetration-testing-tools, react, security-automation, security-testing, security-tools, anthropic, gpt, offensive-security, open-source, openai, penetration-testing, self-hosted
Last push: 2026-08-06T11:10:55+00:00

## Health v2 (maintenance only)
Score: 78/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 96, release rhythm 74, longevity 43
- inputs: {"age_days": 604, "days_push": 27, "days_rel": 96, "gap_med": 41.5, "n_releases_24m": 7}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 22033, forks 2924 (observed 2026-08-28T04:11:32.295685+00:00)

## What it is
PentAGI is a self-hosted, fully autonomous multi-agent AI system for performing complex penetration testing tasks, built in Go with a React frontend. It orchestrates LLM-powered agents (OpenAI, Anthropic, Ollama, and others) with Docker-isolated tool access to automate offensive security workflows.

## Use cases
- automate penetration testing with ai agents
- run autonomous security assessments of my infrastructure
- self-hosted ai pentesting platform
- multi-agent system for offensive security tasks
- llm-powered vulnerability discovery and exploitation
- automate red team reconnaissance and testing

## When to choose
- you want autonomous, end-to-end automated penetration testing rather than single-purpose scanners
- you need self-hosted control with Docker-isolated agent execution
- you want to plug in your own LLM provider including local models via Ollama
- you are a security team or researcher exploring agentic offensive security

## When to avoid
- you need a traditional rule-based vulnerability scanner with deterministic results
- you cannot provide LLM API access or run local models
- you lack authorization to test the target systems - unauthorized use is illegal
- you need a lightweight CLI-only tool without a web interface

## Facets
- artifact type: application
- maturity: active
- function: agent-framework, security, penetration-testing, llm-inference, chat-interface, self-hosted
- domain: security, penetration-testing, large-language-models, developer-tools
- platform: self-hosted, go, cross-platform
- tags: autonomous-agents, multi-agent-system, offensive-security, security-automation, pentesting, llm-agents, red-team, ai-agents, docker, web-server

## Member repositories
- vxcontrol/pentagi (main) score 78

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:11:32.295685+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T16:57:44.130722+00:00, confidence not recorded.
  - readme: https://github.com/vxcontrol/pentagi (fetched 2026-08-28T04:11:32.295685+00:00, sha 05dd256816ba)
  - homepage: https://pentagi.com (fetched 2026-08-29T07:55:54.514077+00:00, sha 9ffbd466132c)
- Data as of 2026-08-30T08:39:29.467469+00:00.