Paperclip AI Flaws Let Attackers Run Host Commands via Malicious Agent Imports

Security vulnerabilities have been discovered in Paperclip, an open-source control plane for artificial intelligence agents. These flaws allow attackers to execute commands on a network server or a developer's computer by importing and starting a malicious agent. There are two paths that attackers can use to achieve this. Additionally, a third vulnerability could lead to the exposure of sensitive data and control-plane details through certain API routes. The vulnerabilities rely on the importation of malicious agents to carry out the attacks.