#40 · Primary category: Cybersecurity & Decryption Tools

Decepticon

agent ai cybersecurity generative-ai hacking langchain langgraph llm pentest pentesting

Autonomous Hacking Agent for Red Team

Project last updated:08/26/26

GitHub Stars

5.4K

Forks

1.0K

Contributors

17

License

Apache-2.0

Why we included this project

Decepticon is an autonomous red-team agent. It plans an engagement through a guided operator interview, writes its rules of engagement and operation plan up front, and then runs reconnaissance, exploitation, privilege escalation, and persistence against the target. The orchestration is what sets it apart from a script that fires nmap and writes a report: it spawns specialist sub-agents with a fresh context window per objective, drives real interactive tools like msfconsole and sqlmap inside persistent sessions, and blocks out-of-scope actions at runtime instead of merely flagging them. The authors report a 98% pass rate on the XBOW exploitation benchmark, a concrete number for teams comparing LLM-based offensive tools. Red teams exploring autonomous emulation should find it a serious, control-plane-focused project worth studying, even if they only run it against deliberately vulnerable targets.

Articles for this project

No articles for this project yet.

To suggest a topic or contribute an article, contact us.

Related projects in this category