Code Injection in Claude Code - CVE-2025-58764
Published: April 30, 2026
Claude Code
Anthropic
Description
The vulnerability allows a remote attacker to execute untrusted commands.
The vulnerability exists due to improper command parsing in the rg command when processing untrusted content in a Claude Code context window. A remote attacker can add crafted untrusted content to bypass the user approval prompt and execute untrusted commands.
Reliably exploiting this issue requires the ability to add untrusted content into a Claude Code context window.