Was just talking with a coworker yesterday how we both don't let Cursor automatically run commands without permission.
Case in point.
Case in point.
"Bash(ceedling:*)",
"Bash(find:*)",
"Bash(grep:*)",
"Bash(ls:*)",
"Bash(rg:*)",
"WebFetch(domain:docs.anthropic.com)",
"Bash(git checkout:*)",
"Bash(gcc:*)",
"Bash(git add:*)",
"Bash(mkdir:*)"
I'm OK for now!``` Based on the Claude Code documentation, it appears that command restrictions are handled internally by Claude Code's security system. However, I can help you understand how to work with this limitation:
Current Claude Code Security:
- Claude Code already has built-in protections against dangerous commands
- Commands like find / -exec rm would likely be blocked by the existing security measures
- The system sanitizes inputs to prevent command injection
```Seems that they already thought of that.