Claude Code rg vulnerability does not protect against approval prompt bypass

CRITICAL 9.8CVE-2025-58764 · GHSA-qxfv-fcpc-w36x · Published 10/09/2025 · updated 17/06/2026

Package	Affected versions	Patched versions
claude code anthropic	`< 1.0.105`	`1.0.105`
@anthropic-ai/claude-code npm	`< 1.0.105`	`1.0.105`

Description

Due to an error in command parsing, it was possible to bypass the Claude Code confirmation prompt to trigger execution of an untrusted command. Reliably exploiting this requires the ability to add untrusted content into a Claude Code context window.

Users on standard Claude Code auto-update will have received this fix automatically. Users performing manual updates are advised to update to the latest version.

Thank you to the NVIDIA AI Red Team for reporting this issue!

NVD description

Claude Code is an agentic coding tool. Due to an error in command parsing, versions prior to 1.0.105 were vulnerable to a bypass of the Claude Code confirmation prompt to trigger execution of an untrusted command. Reliably exploiting this requires the ability to add untrusted content into a Claude Code context window. Users on standard Claude Code auto-update will have received this fix automatically. Users performing manual updates are advised to update to version 1.0.105 or the latest version.

References

anthropics/claude-code security advisory · WEB
nvd.nist.gov/vuln/detail/CVE-2025-58764 · ADVISORY
anthropics/claude-code · PACKAGE

Running one of these products? Scan your site to see whether it exposes a vulnerable version.

Scan my site free