trailofbits/pajaMAS
Multi-agent system (MAS) hijacking demos
This project demonstrates various ways that multi-agent systems (MAS) can be exploited through 'MAS hijacking,' where an attacker manipulates the system's control flow. It takes different MAS configurations as input and shows how specific vulnerabilities can lead to unintended actions or code execution. Security researchers, AI developers, and system architects can use these examples to understand and mitigate potential attack vectors in agentic AI.
Use this if you are developing or securing multi-agent AI systems and need to understand practical examples of how these systems can be maliciously manipulated.
Not ideal if you are a non-technical user looking for a general overview of AI security or a tool to audit your production systems without deep technical understanding.
Stars
42
Forks
3
Language
Python
License
Apache-2.0
Category
Last pushed
Mar 08, 2026
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/agents/trailofbits/pajaMAS"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
4thfever/cultivation-world-simulator
基于 AI Agent 工作流的修仙世界模拟器，旨在还原智能、开放的仙侠世界。| An open-source Cultivation World Simulator using...
nikmcfly/MiroFish-Offline
Offline multi-agent simulation & prediction engine. English fork of MiroFish with Neo4j + Ollama...
oil-oil/wolfcha
AI-powered Werewolf (Mafia) social deduction game where every player is controlled by top LLMs...
yasserfarouk/negmas
Negotiation Multi-Agent System (A negotiation library designed for situated negotiations within...
KsanaDock/Microverse
A god-simulation sandbox game built on Godot 4 as a multi-agent AI social simulation system. In...