sandbox-shell

Execute bounded shell commands with explicit timeouts and captured output.

81|14|Updated Jul 8, 2026
One-click install
npx skills add https://github.com/guaidao2/XuanMu-RedTeam-Agent --skill sandbox-shell-guaidao2
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: sandbox-shell
Source: https://github.com/guaidao2/XuanMu-RedTeam-Agent/tree/main/sandbox/.agents/skills/sandbox-shell
Command: npx skills add https://github.com/guaidao2/XuanMu-RedTeam-Agent --skill sandbox-shell-guaidao2

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill provides a controlled, consistent way to perform shell-level work inside an authorized sandbox without relying on unmanaged execution or unclear output handling.

Core Features & Use Cases

  • Command Execution: Run bounded synchronous commands or dispatch longer-running work asynchronously with explicit timeouts.
  • Output Management: Retrieve compact command metadata and read captured output in controlled line ranges.
  • Task Support: Handle environment setup, script execution, downloads, package workflows, scanning, browser CLIs, and security tooling within the sandbox.
  • Use Case: Use it to inspect a repository, execute a validation script, capture its output, and report meaningful results while keeping generated files scoped to the task.

Quick Start

Use the sandbox-shell skill to inspect the selected sandbox, run the required authorized command with an explicit timeout, and summarize the meaningful results.

Frequently Asked Questions about sandbox-shell

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I execute shell commands safely in a sandbox environment?

Shell commands execute safely in an authorized sandbox by enforcing bounded execution with explicit timeouts and capturing output to scoped files, ensuring controlled and consistent task performance without unmanaged execution risks.

Can I run long-running security scanning tasks asynchronously with output management?

Long-running security scanning tasks run asynchronously with explicit timeouts, while output management retrieves compact command metadata and reads captured output in controlled line ranges to prevent blocking the main workflow.

What is the best way to inspect a repository and capture validation script output?

The best way to inspect a repository is dispatching bounded command execution within a sandbox, running the validation script, and reading the captured output files to report meaningful results while keeping generated artifacts scoped to the task.

Does sandbox shell execution support package workflows and browser CLI automation?

Sandbox shell execution supports package workflows, browser CLI automation, environment setup, downloads, and security tooling by applying bounded command execution with scoped task artifacts and asynchronous job handling for longer-running work.

Why do I need explicit timeouts for command automation in a sandbox?

Explicit timeouts are required for command automation to ensure bounded execution, preventing long-running or unresponsive scripts from blocking the sandbox environment and ensuring captured output remains controlled and scoped to the task.