secure-sandbox

What problem does it solve?

This Skill provides a robust safety net for executing commands, preventing accidental or malicious damage to your system by detecting dangerous operations and enforcing approval workflows.

Core Features & Use Cases

• Dangerous Command Detection: Identifies and flags commands that could harm your system (e.g., rm -rf, chmod 777).
• Approval Workflows: Suspicious commands are queued for your review before execution.
• Allowlisting: Pre-approved safe commands can run automatically.
• Audit Logging: Every command executed is logged for accountability.
• Use Case: When an agent needs to perform file cleanup or install new software, this skill ensures the operation is safe, requires approval if necessary, and logs the action for auditing.

Quick Start

Use sandbox_check to analyze the safety of the command 'rm -rf /tmp/*'.