Skill Detail
Block destructive Terraform, database, Kubernetes, cloud, and Git commands before Claude Code can execute them with Agent Guardrails
Add hard pre-execution guardrails to Claude Code so destructive shell commands are blocked before an agent can run them.
Security & VerificationClaude Code
Security & Verification
Claude Code
Security Reviewed
⭐ 2 GitHub stars
INSTALL WITH ANY AGENT
npx skills add agentskillexchange/skills --skill block-destructive-terraform-database-kubernetes-cloud-and-git-commands-before-claude-code-can-execute-them-with-agent-guardrails
Works best when you want a reusable capability, not another fragile one-off prompt.
At a glance
Tools required
Claude Code, jq, shell access to project or user Claude config, repository clone of the guardrail scripts
Install & setup
Clone the repository, copy the `claude-code` settings, hooks, and scripts into either your project `.claude/` directory or `~/.claude/`, make the shell scripts executable, then dry-run one of the guard scripts to verify blocking behavior.
Author
Robotic Force
Publisher
Organization
Last updated
Apr 22, 2026
Quick brief
Use Agent Guardrails when you want Claude Code to help with infrastructure or repository work but you need hard blocks on destructive commands first. Invoke this instead of using Claude Code normally when the missing step is policy enforcement on shell actions such as `terraform destroy`, `DROP DATABASE`, dangerous `kubectl delete`, or force pushes. The scope boundary is narrow and skill-shaped: pre-tool command interception and blocking for Claude Code, not a generic security platform or plain config bundle listing.