Implement automated PreToolUse security hooks to guard AI agent operations.
npx skills add https://github.com/AreteDriver/ai-skills --skill security-hooks
Please help me install this Agent Skill. Skill: security-hooks Source: https://github.com/AreteDriver/ai-skills/tree/main/personas/claude-code/security-hooks Command: npx skills add https://github.com/AreteDriver/ai-skills --skill security-hooks
This Skill establishes robust security guardrails for AI agents by implementing protective PreToolUse hooks that prevent sensitive file access, dangerous commands, and destructive operations.
--force, this skill will intercept the command, prompt the user for confirmation, and prevent the potentially destructive operation if not explicitly approved.Configure your project's .claude/settings.json to include the block-sensitive-files.sh hook to prevent modifications to files like .env.
High-intent search queries and answers about installing and using this skill.
You can prevent an AI agent from modifying sensitive files by configuring PreToolUse security hooks in your project's settings to block or prompt for confirmation before writes to credential files occur. This skill provides ready-to-deploy shell scripts for this exact scenario.
Yes, you can intercept destructive Git commands like git push --force by implementing PreToolUse guards. The hook intercepts the command execution, prompts the user for explicit confirmation, and prevents the operation if not approved.
To add guardrails against command injection in AI agent operations, you deploy customizable hook scripts that intercept tool use requests. These scripts parse JSON inputs using jq and evaluate commands against security rules before allowing execution.
Yes, you need jq installed to configure these PreToolUse hooks, as the shell scripts rely on jq for JSON parsing to evaluate intercepted tool requests. You also need adherence to specific exit codes and JSON output contracts for decision-making.
The best way to prevent AI agents from modifying system paths is to implement automated PreToolUse security hooks that guard against destructive actions. This skill includes specific scripts designed to intercept and block system path modifications.
The skill includes 7 ready-to-deploy hook scripts for AI agent guardrails, covering security scenarios like sensitive file protection, command injection prevention, destructive Git operation confirmation, and system path guarding.
ronnycodingCreate and manage Claude Code hooks across nine lifecycle events.
heyu-aiManage bash-hygiene hook audit logs and analyze violation rates.
JosiahSiegelEnforce secure Bash scripting patterns with input validation and injection prevention.
manuelbrandner85Generate minimal macOS Seatbelt sandbox configurations for process isolation.