Breaking

PentAGI Framework Reaches Version 2.2.0 as Open-Source AI Penetration Testing Tool Hits 29k GitHub Stars

October 14, 2026 at 13:34 EDT

AI Agents
Security
Open Source

VXControl released version 2.2.0 of PentAGI, an open-source framework that deploys autonomous AI agents to perform infrastructure penetration testing, marking significant maturation of a tool that has attracted nearly 30,000 GitHub stars since its January 2025 debut.

October 2026

PentAGI 2.2.0: AI-Driven Security Testing Matures

Open-source autonomous penetration testing framework reaches 29,000 GitHub stars with production-grade features and enhanced security hardening.

How PentAGI Conducts Automated Testing

User Input

Describe target & goal

Natural language instructions

Planning

Multi-agent orchestration

OSINT & reconnaissance

Execution

Sandboxed testing

20+ security tools

Results

Findings & reports

1–5 hours to completion

Key Capabilities

Agent Teams

Orchestrator, Researcher, Developer, Executor work autonomously

LLM Support

13 providers including Claude, GPT, Gemini, local Ollama

Integration

REST & GraphQL APIs for CI/CD & automation

Memory

Vector DB & knowledge graph maintain cross-assessment findings

Strengths vs. Limitations

What It Does Well

Rapid reconnaissance & OSINT
Basic exploitation automation
Continuous 24/7 testing
Reduced cost vs. traditional pentesting
Self-hosted data privacy

Gaps Requiring Humans

10–30% false positive rate
Business logic flaw detection
Complex multi-step attacks
Strategic reasoning
Exploit validation

Industry Assessment

Security teams view PentAGI as a force multiplier that handles routine reconnaissance and basic exploitation, freeing human experts for complex analysis and strategic planning. The consensus: professionals who integrate AI-driven assessment into their workflows will remain valuable; those relying only on manual pentesting face disruption.

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.