# pentagi

> Autonomous multi-agent penetration testing system

PentAGI is an open-source autonomous security testing platform powered by multi-agent AI workflows. It automates penetration testing tasks and vulnerability assessments using a sandboxed Docker execution environment equipped with over 20 professional security tools. Users manage and monitor agent tasks, reports, and execution logs through a web UI or programmatically via GraphQL and REST APIs.

- Source: https://github.com/vxcontrol/pentagi
- Homepage: https://pentagi.com
- License: MIT
- Language: Go
- Stars: 25227
- Forks: 3242
- Contributors: 29
- Last commit: 2026-10-03
- Latest release: v2.1.0 (2026-05-29)
- Purpose: AI & LLM tooling, Automation, Security & identity, Agents & frameworks
- Runs on: Web, Docker / self-host, CLI
- For: Personal, Enterprise
- Enterprise signals: sso

## Worth score: 76/100

How much DigGitHub recommends it, from activity, adoption, docs, license and security signals.

- Popularity: 22/25
- Momentum: 10/20
- Maintenance: 21/25
- Community: 11/15
- Readiness: 12/15

## Alternatives

- [Shannon](https://diggithub.com/KeygraphHQ/shannon.md): AI pentester for your web apps and APIs
- [Strix](https://diggithub.com/usestrix/strix.md): AI agents that pentest your own apps
- [AutoGPT](https://diggithub.com/Significant-Gravitas/AutoGPT.md): Platform for building and running autonomous AI agents
- [Sim](https://diggithub.com/simstudioai/sim.md): Build and deploy AI agent workflows
- [Skyvern](https://diggithub.com/Skyvern-AI/skyvern.md): Automate browser workflows with AI
- [OpenHands](https://diggithub.com/OpenHands/OpenHands.md): AI software engineering agent platform
- [n8n](https://diggithub.com/n8n-io/n8n.md): Workflow automation with native AI capabilities
- [Ruflo](https://diggithub.com/ruvnet/ruflo.md): Orchestrate swarms of AI agents

---

Source page: https://diggithub.com/vxcontrol/pentagi
Updated: 2026-10-04