yjyddq/RiOSWorld
[NeurIPS 2025] Official repository of RiOSWorld: Benchmarking the Risk of Multimodal Computer-Use Agents
This project helps researchers and developers evaluate the potential risks of multimodal computer-use agents, particularly those designed to interact with a desktop environment. It takes a computer-use agent (like an AI assistant that controls a mouse and keyboard) as input and outputs a benchmark of its risk behaviors in various scenarios. The primary users are AI researchers and developers working on agent safety and trustworthiness.
117 stars.
Use this if you are developing or researching multimodal AI agents and need a standardized way to benchmark their safety and identify risky behaviors.
Not ideal if you are an end-user looking for a pre-built safety tool for AI agents, as this is a research and benchmarking framework.
Stars
117
Forks
6
Language
HTML
License
—
Category
Last pushed
Dec 02, 2025
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/agents/yjyddq/RiOSWorld"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Featured in
Higher-rated alternatives
StonyBrookNLP/appworld
🌍 AppWorld: A Controllable World of Apps and People for Benchmarking Function Calling and...
qualifire-dev/rogue
AI Agent Evaluator & Red Team Platform
microsoft/WindowsAgentArena
Windows Agent Arena (WAA) 🪟 is a scalable OS platform for testing and benchmarking of...
future-agi/ai-evaluation
Evaluation Framework for all your AI related Workflows
RouteWorks/RouterArena
RouterArena: An open framework for evaluating LLM routers with standardized datasets, metrics,...