Critical Anthropic’s MCP Vulnerability Enables Remote Code Execution Attacks

OX Security disclosed a critical architectural vulnerability in Anthropic's Model Context Protocol (MCP) SDKs across multiple languages that can enable arbitrary command execution (RCE) and full system takeover; researchers demonstrated multiple exploitation families, confirmed execution on several live platforms, identified around 10 CVEs (some patched, some still reported), and provided mitigation guidance including restricting public access, treating MCP inputs as untrusted, sandboxing, and updating affected services.