careful

You can add a warning prompt by using a Bash hook that detects destructive patterns like rm -rf and requires explicit confirmation before execution. This guardrail approach intercepts risky commands and enforces an interactive safety check.

Common destructive commands to guard against include rm -rf, git reset --hard, kubectl delete, and DROP TABLE operations. A safety hook can detect these dangerous patterns in user input and warn before execution to prevent catastrophic data loss or service disruption.

Yes, you can override the safety warning by providing explicit confirmation after the prompt appears. The override feature is designed for experienced users who need to proceed with confirmed risky operations in development, staging, or production environments.

Yes, the bash safety guardrail works for git and kubectl commands by detecting dangerous patterns such as git reset --hard, force-push, and kubectl delete. It uses a Bash hook to intercept these commands and enforce confirmation before execution.

You set up a bash hook that intercepts user input and matches it against known dangerous command patterns like rm -rf and git reset --hard. Once a match is detected, the hook pauses execution and requires explicit user confirmation before proceeding.

You should not rely on a destructive command guardrail as your only safety measure in fully automated pipelines where interactive confirmation is not possible. The hook requires explicit user confirmation, making it best suited for interactive development, staging, and production sessions.