careful

Preventing destructive shell commands requires a safety layer that pattern-matches risky inputs. This Skill detects recursive deletes, force pushes, and hard resets before execution, issuing overrideable warnings to stop irreversible actions.

Yes, you can add guardrails to Kubernetes cleanup and Docker maintenance. The Skill detects resource deletion and Docker pruning patterns in your shell session, warning you about high-risk infrastructure operations before they execute.

Protecting shared environments during production debugging requires safe operational guardrails. This Skill intercepts high-risk commands before execution, prompting you to review and choose whether to proceed or cancel to reduce accidental damage.

Yes, destructive command detection works with database administration. The Skill pattern-matches irreversible SQL actions like drops and truncates, warning you and providing an overrideable prompt before the high-risk operation runs.

Overriding a safety prompt for an intentional high-risk Git command uses the user override workflow. After the Skill warns you about the destructive action, you can review the warning and choose to proceed with the execution or cancel it.

Automated guardrails for shell work rely on pattern matching for known destructive syntax. You should anticipate limitations with complex, obfuscated commands or aliases that bypass standard recursive delete, force push, or hard reset patterns.