avad-careful

Safety guardrails intercept destructive bash commands like rm -rf before execution. The system analyzes command patterns and prompts the user for explicit permission, ensuring dangerous actions require confirmation to prevent accidental data loss.

Yes, the safety guardrail works with git reset --hard and kubectl delete commands. It monitors Bash-like environments to detect these destructive patterns and halts execution to prompt the user for explicit permission before proceeding.

You can override the permission prompt for authorized destructive commands. The safety guardrail provides an explicit override option, allowing users to bypass the confirmation requirement when running necessary destructive tasks in development or staging environments.

Safety guardrails analyze destructive command patterns to provide protection during production debugging. By intercepting commands like rm -rf and git reset --hard, the system prevents service disruption and data loss by requiring explicit user confirmation before execution.

Pre-execution command monitoring is enabled by prompting the system to validate the next destructive operation. The skill analyzes Bash commands for destructive patterns and automatically halts execution to request user permission before any dangerous action proceeds.