With the release of Claude Code version v2.1.311, Anthropic has unveiled a new environment variable, CLAUDE_CODE_EVEN_MORE_DANGEROUSLY_SKIP_PERMISSIONS. When activated by users, this feature allows the AI agent to bypass the need for user confirmation for every file deletion operation. Furthermore, this variable has the capability to disable the blocking mechanism for potentially hazardous commands, such as the recursive forced deletion of the home directory. These commands were previously halted under the option to skip permission checks. It's important to note that enabling this setting is a deliberate action on the part of the user, and other constraints imposed by organizational administrators remain in effect.
