Skip to content

Claude Code

Claude Code env var lifts its last block on rm -rf ~

Written by an AI model from the linked sources and checked against them by a second model. How we work

Image: theregister.com

Daisy Hollman of Anthropic says today’s Claude Code release lets you disable the block that --dangerously-skip-permissions still applies to commands like rm -rf ~. It’s opt-in and only applies once you set CLAUDE_CODE_EVEN_MORE_DANGEROUSLY_SKIP_PERMISSIONS=1, and The Register notes it’s missing from the changelog, though Hollman replied “Not a joke!”

Without the variable, bypassPermissions mode still pauses for a prompt on rm and rmdir, and managed settings may impose their own limits either way. The variable ships in Claude Code v2.1.311.

Anthropic

Claude model’s Linux exploit gets root over multipath TCP

“Run arbitrary commands as root, without any credentials,” Anthropic says of a Linux exploit its unreleased Claude model wrote, and per Golem, Ubuntu and Red Hat still lacked a patch. Has a kernel fix ever reached your servers weeks after the CVE went public?

No comments yet. Be the first.

Join the conversation

Sign in to leave a comment. It appears after moderation.