A critical vulnerability in the MS‑Agent framework’s Shell tool allows untrusted input to be executed as operating‑system commands, potentially giving attackers full control of affected systems. This short note summarizes the issue, its impact, and immediate mitigations, and points to the original advisory for technical details. Overview MS‑Agent exposes a Shell capability intended to let AI agents run OS commands
Tag: sandboxing
Claude Cowork finally lands on Windows
Anthropic’s Claude Cowork has arrived on Windows, closing a major platform gap and bringing its AI-powered desktop agent to a much larger share of enterprise users. The Windows release delivers local file access, multi-step automation, plugin integrations, and Model Context Protocol connectors, matching the macOS version and positioning Cowork as a mainstream productivity agent for knowledge work. What Cowork brings