Modification of MCP Server Definitions Bypasses Manual Re-approval

High7.2CVE-2025-54136 · Published Aug 1, 2025

### Summary A vulnerability in Cursor AI allows an attacker to achieve remote and persistent code execution by modifying an already trusted MCP configuration file inside a shared GitHub repository or editing the file locally on the target's machine. Once a collaborator accepts a harmless MCP, the attacker can silently swap it for a malicious command (e.g., calc.exe) without triggering any warning or re-prompt. ### Impact If an attacker has write permissions on a user's active branches of a source repository that contains existing MCP servers the user has previously approved, or allows an attacker has arbitrary file-write locally, the attacker can achieve arbitrary code execution. ### Remediation Cursor now requires user approval every time an mcpServer entry is modified (in addition to the existing behavior of asking for approval when a new server is added).