[ autonomous ] fully self-directed offensive-security agents

Point it at a target.
It runs the whole engagement.

recon ──▶ enumerate ──▶ exploit ──▶ pivot ──▶ report └── agents coordinate, plan, and act autonomously ──┘

A fully autonomous AI agent system capable of performing complex penetration-testing tasks — recon, enumeration, exploitation, and reporting — planned and executed without a human in the loop.

./run engagement git clone

pentagi · session #1 · scope: lab.internal◉ running

reconnmap sweep → 3 hosts, 11 open ports ✓

enumsvc fingerprint → outdated web stack @ .12

planagent: "chain CVE-x on .12, then pivot to db"

exploitfoothold on .12 → shell acquired

pivotreach db.internal → creds exposed

reportdrafting findings · severity map

// hosts, ports and findings are illustrative

github.com/vxcontrol/pentagiautonomous offensive security

Point it at a target.It runs the whole engagement.

Point it at a target.
It runs the whole engagement.